UN Population Division, Department of Economic and Social Affairs,
with support from the UN Population Fund (UNFPA)
*************************************************************************** This document is being made available by the Population Information Network (POPIN) of the United Nations Population Division (DESIPA), in collaboration with the Population Programme Service, Women and Population Division of the Food and Agriculture Organization of the United Nations. For further information, please contact Mr. Jacques du Guerny, Chief of the Population Programme Service via email: firstname.lastname@example.org **************************************************************************** Population and the environment: a review of issues and concepts for population programmes staff II. POPULATION AND LAND DEGRADATION September 1995 ======================================================================== FOREWORD This paper is the second in a series designed to bring to the attention of Country Support Team Advisers (and staff of UNFPA programmes concerned) state-of-the-art information on major population-environment issues and methodological advice for dealing with such issues in the context of population policy work and population/development programmes. The purpose of each paper is to help fellow population specialists at the regional and country level carry out such tasks as: - promote awareness of population and environment linkages and related issues qua relevant elements in development policies; - help integrate environmental concerns and considerations in population policy analyses; - help design or carry out population-centred research in support of development policy studies; or - help design data collection and monitoring systems on population/environment issues. For this purpose, each paper provides factual information on the environmental issue(s) under review, tries to elucidate the role of population variables, proposes analytical tools and examines statistical information problems where appropriate. The first paper focused on water resources issues.1/ The present one deals with land degradation, a global problem of crucial importance in view of the vital functions of soils in the survival of the people. This paper is addressed to Country Support Team Directors, Advisers on Population and Development, and FAO Advisers. Suggestions for further distribution and requests from field projects are welcome. Alain Marcoux FAO/UNFPA TSS ======================================================================== CONTENTS 1. LAND DEGRADATION: TYPOLOGY OF ISSUES 4 1.1 Erosion 4 1.2 Chemical deterioration 5 1.3 Physical deterioration 5 1.4 On "desertification" 6 2. EXTENT AND IMPACT OF LAND DEGRADATION 6 2.1 Incidence by type of degradation 6 2.2 Impact of land degradation 7 3. CAUSES OF LAND DEGRADATION: THE ROLE OF POPULATION FACTORS 9 3.1 The causative factors 9 3.2 Population and land degradation processes 12 3.2.1 Deforestation and overexploitation of vegetation 12 3.2.2 Overgrazing 13 3.2.3 Improper agricultural management 15 3.3 Population and technological factors 17 3.4 Social and institutional factors 20 3.4.1 Agrarian structures and poverty 20 3.4.2 Land tenure 22 3.4.3 Markets and public policies 23 4. CONCLUDING REMARKS 24 4.1 Population in the chains of explanation of land degradation 24 4.2 Relevance for population programmes 25 REFERENCES 28 NOTES 31 ANNEXES ====================================================================== If the soil on which agriculture and all human life depends is wasted away then the battle to free mankind from want cannot be won.2/ Land, like water, is a vital resource to humankind (see Annex 1); but that resource is easily overrated. Only 11 percent of the world's land area presents no limitations for agricultural use; on some 28 percent the climate is too dry, and on 10 percent it is too humid; on 23 percent the soil has critical chemical imbalances, and on 22 percent it is too shallow; the remaining 6 percent is permanently frozen (FAO, 1980). In addition, various forms of degradation attack that resource as a result of various natural and human-made factors. This paper is devoted to a review of some related issues. 1. LAND DEGRADATION: TYPOLOGY OF ISSUES The concept of land degradation "refers to the deterioration or total loss of the productive capacity of the soils for present and future use" (FAO, 1980). Such loss occurs mainly because of various forms of erosion (by wind and water) and of chemical and physical deterioration. This typology, as established for the Global Assessment of Soil Degradation (GLASOD) project (ISRIC/ UNEP, 1991), is reviewed hereunder and will be used in the rest of this paper. 1.1 Erosion The most common form of erosion is the loss of topsoil under the action of water or wind. Water runoff carries the topsoil away; this occurs under most climatic and physical conditions. Displacement of topsoil by wind action is more widespread in arid and semi-arid climates than under more humid conditions. The loss of topsoil reduces fertility because [a] as the soil becomes denser and thinner, it is less penetrable by growing roots and may become too shallow for them; [b] the capacity of the soil to retain water and make it available to plants is reduced; and [c] plant nutrients wash away with soil particles. A more extreme form of erosion is terrain deformation. Water may cause the formation of rills (i.e. small channels, which can be ploughed over) and gullies (i.e. deeper channels, cut by larger water flows and difficult or impossible to level by ploughing). It may also cause the destruction of riverbanks, and mass movement (landslides). Wind action may create deflation hollows and dunes. Finally, the covering of the land surface by wind-carried particles (or overblowing) is also recognized as a specific form of degradation. Erosion risks depend both on natural conditions and on land use patterns. The climate (especially rain intensity), slopes, vegetation cover, and nature of the soil are important.3/ With regard to land use, any human activity which entails the removal of the protective vegetation cover (forest, shrubs, grass etc.) fosters erosion; so do improper measures such as ploughing along slopes. 1.2 Chemical deterioration Chemical deterioration may consist in: (a) The loss of soil nutrients (mainly nitrogen, phosphorus and potassium) or organic matter. In part, nutrients are lost through erosion: "in the humid tropics, many nutrients are leached during the intense rainstorms, especially on unprotected land"; in addition, they can be "depleted by the crops themselves, particularly if the same crops are grown on the same land year after year" (FAO, 1983). Depletion is widespread where "agriculture is practised on poor or moderately fertile soils, without sufficient application of manure or fertilizer" (ISRIC/UNEP, 1991). (b) Salinization, or the concentration of salts in the topsoil, which may occur because of: (i) poor management of irrigation schemes■high salt content of irrigation water or insufficient attention to drainage can easily lead to rapid salinization of the soils, especially in arid areas where high evaporation rates foster the process; (ii) the intrusion of seawater or saline groundwater in water reserves of good quality;4/ or (iii) human activities which increase evaporation in soils on salt-containing material or with saline groundwater (ISRIC/UNEP, 1991). Salinization has "a deleterious effect on soil productivity and crop yields" (FAO, 1994); in extreme cases, "damage from salinization is so great that it is technically unfeasible or totally uneconomic to reverse the process" (FAO, 1983). (c) Acidification, which may occur either because of excessive application of acidifying fertilizer or because of drainage in particular types of soil; and (d) Pollution of various origins (waste accumulation, excessive use of pesticides or manuring, oil spills etc.), can strongly reduce the agricultural potential of lands. 1.3 Physical deterioration Three types of physical deterioration are recognized: (a) Soil compaction, usually resulting from the use of heavy machines on unstable soils or from cattle trampling; sealing and crusting, usually caused by the impact of raindrops. These conditions make tillage more costly and impede seedling emergence. Also, by restricting water infiltration, they cause faster run-off and water erosion. (b) Waterlogging, i.e. the rise of the water table to the root zone of plants, caused by an excessive input of water with respect to drainage capacities. It is typical of irrigated areas, but may also occur through river flooding. Waterlogging also increases salinity (see above). As with salinization, the causes of waterlogging are in part physical and in part related to agricultural practices, namely inappropriate irrigation. (c) Subsidence (i.e. lowering of the land surface) of organic soils, which can be caused by drainage or oxidation. 1.4 On "desertification" A note about the concept of desertification is needed. The United Nations Conference on Desertification, which popularized the word, defined it as "the reduction or destruction of the land's potential, finally resulting in the appearance of desert conditions" (United Nations, 1977). Gorse and Steeds (1987) write about a process of decline in the biological productivity of land that results in "desert, or skeletal soil that is irrecuperable". Although in principle one should use the concept only in reference to processes which have resulted in desert conditions ■or will shortly and inevitably do so■it is often used in a broader sense. UNEP (1991) defines it as land degradation in arid, semi-arid and dry sub-humid areas resulting mainly from adverse human impact. The particular attention to dry climate settings owes much to the droughts of the 1970s and early 1980s and the alleged expansion of the Sahara.5/ The term "desertification" has some annoying aspects. It provides no information on the nature of the degradation, nor on the nature of possible corrective actions. It misstates problems: "The concept of expanding deserts and advancing sand dunes has become the dominant image in the public's eye rather than [...] less visible and much more serious problems" (Liamine, 1993), in particular "more subtle, more complex, pulsating deteriorations, sometimes with reversals, but at least with substantial periodic remissions, radiating out from centers of excessive population pressure" (Nelson, 1990). And it seems to designate an absolute evil, while the salinization of an irrigation area, although reversible, may be a greater loss than the washing away of the last inch of topsoil in a marginal area. For FAO (1986a), desertification "is only one extreme aspect of the widespread deterioration of ecosystems under the combined pressure of adverse climate and agricultural exploitation". The rest of this paper will implicitly cover "desertification" in this sense, but usually will not isolate it from land degradation phenomena in general. 2. EXTENT AND IMPACT OF LAND DEGRADATION 2.1 Incidence by type of degradation The GLASOD study, covering most of the world's land surface (ISRIC/UNEP, 1991), found that globally 15% of the land area was degraded as a result of human activities 6/ ; the respective impact of the various forms of degradation at the global level was estimated as reported in Table 1 (next page). Annex 2 provides details on patterns of degradation at the regional level. Table 1. Incidence of 10 forms of land degradation at the global level (percentage of total area degraded) Loss of topsoil....... 70.0% Pollution............. 1.1% Terrain deformation... 13.0% Overblowing........... 0.6% Loss of nutrients..... 6.9% Waterlogging.......... 0.5% Salinization.......... 3.9% Acidification......... 0.3% Compaction............ 3.5% Subsidence............ 0.2% The impact of salinization and waterlogging must be measured with reference to irrigated areas. FAO (1994) estimates that, out of some 240 million hectares currently irrigated, about 30 are severely affected by salinity and another 60 to 80 are affected to some extent. Of the four countries with the largest irrigated areas (together accounting for half of the global area) salinity affects 28 percent of irrigated land in the USA, 23 percent in China, 21 percent in Pakistan and 11 percent in India (Umale, 1993). The rate at which degraded areas expand is poorly known on a global scale, because there exist no preceding data with which to compare GLASOD results. Estimates vary commonly between 5 and 12 million hectares lost annually (out of a total 4.8 billion hectares of arable land and pastures). FAO warns that much progress still needs to be made on land use data collection before this and other important trends are adequately known. The extent of the threat is certainly considerable: on the basis of its classical study on potential population-supporting capacities of the lands, FAO (1984) estimated that without soil protection measures, close to 550 million hectares of rainfed cropland could be lost during 1975-2000, with percentage losses ranging from 10 percent in south America to 38 percent in Asia. In addition much of the remaining cropland would lose some fertility due to the degradation of topsoil, with an overall loss in production potential in the order of 30 percent.7/ 2.2 Impact of land degradation Some typical consequences of land degradation are illustrated in Annex 3. Let us review on- and off-farm effects, then look briefly into the question of aggregate effects at the sector level. The main on-farm effect of land degradation is a decline in yields■or an increased need for inputs to maintain those yields: since "subsoils generally contain fewer nutrients than topsoils, more fertilizer is needed to maintain crop yields. This, in turn, increases production costs. Moreover, the addition of fertilizer alone cannot compensate for all the nutrients lost when topsoil erodes" (FAO, 1983). Where degradation is serious, the plots may be either abandoned temporarily or permanently, or converted to inferior value uses, e.g. cropland being converted to grazing land, or grazing land left to shrubs. With salinization and waterlogging in irrigated areas, reductions in yields are even bigger because the starting point is higher: field studies indicate reductions in yields oscillating from 30 to more than 80 percent; values around 50 percent are the most common (Pinstrup-Andersen and Pandya-Lorch, 1994). The economic loss resulting from such reductions in the very raison d'etre of the heavy investments made is highly significant. A complete picture of erosion costs should include offsite effects. It has been argued that measurements of soil erosion from test plots "typically overestimate the consequences for productivity, since the eroded soil can remain for decades elsewhere in the farming landscape before it is delivered to the oceans. Thus, a portion of on-site erosion represents a transfer of assets rather than a complete loss from the standpoint of agricultural productivity" (World Resources Institute, 1993). This argument should not be carried too far. First, as the same source adds, geographic shifts in productivity have potentially important distributional consequences: it is not unimportant that topsoil washed from slopes held by the poor ends up in valley bottoms held by the better-off, or is lost by a mountainous country to the benefit of downstream countries. Also, the fine soil particles for the most part are carried to waterways and seas; along the way they may make water unsuitable for human consumption, silt up dams, irrigation systems or river transport channels. Eventually their nutrients are permanently lost for agriculture, but cause nutrient loading and eutrophication, damaging aquatic life systems and fisheries. Indeed, even the better soils of many parts of developing regions are gradually becoming less productive; some fertility declines affect the very areas which call for urgent and sharp production increases; one key reason is the "mining" of soil nutrients by cropping without adequate replenishment. Land is also taken out of production because of chemical pollution from all origins including industrial and urban waste. It is difficult to estimate the total losses caused by land degradation worldwide. According to FAO (1992) about 25 billion metric tons of soil (17 tons per cultivated hectare) erodes each year. Translating this into an estimate of foregone production is even more difficult, since the effect varies from place to place depending on the mix of productions, and the hypothetical reduction in yields. Only crude estimates are possible. According to Sfeir- Younis (1986), the food production of rainfed croplands could decline by 19 to 29 percent over the 1985-2010 period. Brown et al. (1990) estimated that land degradation worldwide causes the loss of roughly 14 million tonnes of grain annually, i.e. half the quantity needed to cover the needs of the additional global population for the same period. In addition, eroded land becomes more vulnerable to climatic variations; its fertility may collapse entirely after a year of drought. "When production conditions are adverse [...] the margin of productivity or of survival for a producer on degraded land is smaller than that of a producer on better managed land [...] Land degradation, as well as drought, has been partly responsible for the severity of famine in agricultural areas of Ethiopia and Sudan" (Blaikie and Brookfield, 1987). Whether land productivity declines, or steps are taken to restore productivity and prevent further losses, "the yield of labour [...] is adversely affected. Land degradation, therefore, directly consumes the product of labour, and also consumes capital inputs into production" (Blaikie and Brookfield, 1987). Wellbeing diminishes because it takes longer and harder work hours to obtain the same product, not to mention taking protective measures on the fields, fertilizing more intensively, or reaching more distant fields and pastures. This added burden usually falls disproportionately on women (except for land clearing work), because they are predominantly involved in food crop cultivation and activities connected with livestock. Also, decreases in the productivity of traditional agriculture often are decisive in triggering out-migration (usually of males), with consequent increases in the workloads of those left behind, including children. On less frequent, but dramatic occasions, land degradation forces population displacement. Hundreds of thousands of hectares have to be abandoned each year being too degraded for cultivation or even grazing. This may mean that the population which depended on those areas for subsistence must seek other lands to settle on. In India, for instance, wasteland accounts for about 38 percent of the land area (about the same as the total cropped area). Most of that area was cultivated at some time in the past, but cultivation was given up (mostly during this century) because of land degradation, and the people formerly cultivating it were displaced to distant areas (Maloney, 1991). Also in India, it was estimated in 1990-1991 that about 8 million hectares were damaged by waterlogging or salinity from irrigation and that as many as 1.5 million farmers had been displaced by those problems since Independence. In Pakistan "irrigated land [was] going out of production at the rate of 100 hectares a day"; many displaced farmers moved to the newly irrigated areas in Western Punjab■likely to face the same situation a few years later (Maloney, 1990). Others went to swell the numbers of urban slum dwellers. 3. CAUSES OF SOIL DEGRADATION: THE ROLE OF POPULATION FACTORS 3.1 The causative factors Figure 1 (page 10) illustrates the factors of land degradation as described in section 1, distinguishing natural conditions and human actions which facilitate or cause the different kinds of degradation. It is worth noting that those factors have different roles: some directly cause degradation; others merely enable the action of the former. For instance, in the case of erosion, the direct cause is the action of water or wind. That action is enabled by a series of conditions, both human-made (deforestation, ploughing slopes etc.) and natural (steepness, soil texture etc.). In the case of salinization, the direct cause can be the intrusion of saltwater in groundwater reserves, and overuse of freshwater the enabling factor; or, the direct cause can be the mix of excessive irrigation and insufficient drainage, with aridity an enabling or accelerating factor.8/ Figure 1. Causation links between human actions, soil degradation and natural conditions. Table 2 classifies human actions and natural conditions, as seen in Figure 1, into "direct factors" and "enabling factors" categories. With these distinctions in mind, let us try to discern the role of population changes in the causative factors of land degradation. Table 2. A classification of factors of land degradation ------------------------------------------------------------ Human actions: Natural conditions: ------------------------------------------------------------ Enabling Deforestation Topography factors: Allow overgrazing Soil texture Excessive use of Soil composition vegetation Aridity, drought Ploughing slopes Vegetal cover Removing grass for Hydrographic cultivating regimes ------------------------------------------------------------- Direct Use of machines Strong rains factors: Driving cattle Floods Shortening fallow Strong winds Excessive water input/insufficient drainage Excessive acid fertilization Excessive use of chemicals/manure Domestic/industrial waste disposal ----------------------------------------------------------------- The GLASOD project quantified the impact of damaging human activities, classifying these into five broad categories, as follows (ISRIC/UNEP, 1991): 1/ Deforestation and removal of natural vegetation for cropping or cattle raising, large scale commercial forestry, road construction, urban development etc. 2/ Overgrazing (destroys soil cover, causes compaction and fosters the encroachment of undesirable shrub species). 3/ Agricultural activities. Improper land management includes cultivation of fragile soils, reduced fallow, uncontrolled use of fire, "practices that result in the net export of soil nutrients", diversion of rivers for irrigation purposes, or irrigation of inadequate soils (FAO, 1993b). 4/ Overexploitation of vegetation for domestic use (use of the vegetation for fuelwood, fencing etc., where the remaining vegetation no longer provides sufficient protection from soil erosion). 5/ (Bio-)industrial activities, causing pollution. The GLASOD assessment then attributed degradation in each surveyed area to one or two of these causative factors; Table 3 shows their respective impacts. While at the global level deforestation, overgrazing and agricultural activities have relatively similar incidence, at the regional level patterns differ markedly. Overgrazing dominates in Australasia and Africa, agricultural activities in North America and deforestation in the other regions. Domestic overuse of vegetation is negligible in developed regions but quite significant in Africa. Pollution is marginal, except in Europe. Table 3. Incidence of the five causative factors of land degradation, by region (percentage of degraded area)9/ --------------------------------------------------------------- Defore- Over- Agric. Overex. (Bio-) station grazing activit. veget. indust. ---------------------------------------------------------------- Africa 14% 49% 24% 13% * N./C.America 11% 24% 57% 7% * South America 41% 28% 26% 5% - Asia 40% 26% 27% 6% * Australasia 12% 80% 8% * * Europe 38% 23% 29% * 9% WORLD 29% 35% 28% 7% 1% --------------------------------------------------------------- Although erosion may take place without human intervention, in practice it usually is started or/and accelerated by human actions which cause the disappearance of the protective cover of natural vegetation or damage soil structure. Let us examine the role of population factors in the occurrence of such practices. 3.2 Population and land degradation processes This section reviews the possible linkages between population factors (population size, geographic distribution, age/sex structures, and changes in these), on the one hand, and the main causative factors, i.e. the first four, on the other.10/ 3.2.1 Deforestation and overexploitation of vegetation The destruction of forests is caused for the most part by land clearance for agricultural purposes. "Both slash-and-burn agriculture, when land is not allowed to lie fallow as long as traditional practices dictated, and permanent clearing for modern farms, are taking a toll" (FAO, 1983). Shifting cultivation entails "cutting trees and shrubs and tall grasses, burning the litter, growing crops for 2 to 5 years on the cleared land, and then allowing the natural cover to return to regenerate the soil [...] the fallow period may last any time from 5 to 15 years, depending on the soil and type of vegetation" (FAO, 1983). Such operations are estimated to have contributed some 60 percent of the expansion of farmland between 1973 and 1988.11/ The removal of vegetation cover starts or accelerates soil erosion under rain and wind action, and "burning for weed control [...] encourages leaching and soil loss" (Cruz, 1994). Land clearing in shifting cultivation is largely driven by population growth, through the growth in requirements food and other agricultural products. Comparatively, forest clearing for pastures is a minor factor on a global scale (although it is important in certain countries). There also are examples of rapid deforestation for commercial agriculture. These seem of growing importance, particularly in Latin America and Asia. But, so far and globally, forest clearing has been more typical of situations of subsistence agriculture (in addition, population growth also is a factor in commercial agriculture). As for logging, it concerns smaller areas, does not destroy the whole vegetation, and does not involve the destruction of organic matter, roots, seeds etc. that forest burning does; it does play an enabling role by opening access roads, but it does not create the need for land clearing. The other cause of destruction of the vegetation cover is its overuse by households, mainly from fuelwood collection. To cover vital energy needs, most households in developing countries resort to "free" gathered biomass fuels, including crop residues and animal dung but, most of all, fuelwood. When the annual use of wood exceeds the sustainable yield of wooded areas, forests and woodlands are gradually destroyed. This in turn triggers or accelerates soil erosion. Around 1980, FAO estimated that about 2 billion people (or 3/4 of the population of developing countries at that time) depended on biomass for their daily energy consumption (FAO, 1983). But close to 1.4 billion of these could not meet their requirements without compromising future fuelwood supplies, and it was expected that the number would increase to 3 billion (2.4 billion in rural areas) by the year 2000. The impact of population growth on fuelwood consumption in the vast areas concerned is direct, since energy needs are essentially proportional to population. Another feature of population dynamics plays an important role, namely urbanization. A first effect arises from population concentration, which makes the impact on resources felt acutely over a peripheral zone which typically suffers disproportionately from deforestation. A second effect arises from changes in habits: urban dwellers frequently prefer charcoal to wood; this increases the impact on wood resources per consumption unit. Overall, population pressure is determinant in vegetation loss, especially in areas with limited land reserves and energy sources. In the high population density areas of West Africa, for instance, "concentrations of demand for arable land and fuelwood lie at the root of resource abuse. It is in these areas that patches of desertification are the most visible" (Gorse and Steeds, 1987). 3.2.2 Overgrazing Excessive pressure on the vegetal cover by animals can be a crucial problem, especially in developing countries where rangelands usually are much more crowded than in the developed world (FAO, 1983). While livestock does not necessarily cause environmental problems (see Annex 4), overgrazing can be a major factor in land degradation, causing half of the damage assessed in Africa and one- fourth in other developing regions. Cases such as the damage caused by goats in the Mediterranean area and elsewhere are well known. In Africa, the increase in cattle numbers and the decline in the quality of rangelands have been significant during the recent decades (FAO, 1986b). These two trends are obviously incompatible in the long run, and local crises are likely in the future. Nomadic grazing in semi-arid areas is an environmentally compatible, effective land use system developed over the centuries by pastoralists; but local collapses of such systems are being noted with increasing frequency. When more and larger herds compete for the same rangelands, they may exceed the natural productivity of the area and destroy the vegetal cover, accelerating erosion. In the Sudan for instance, the growth of the pastoralist population and increased livestock density have led to the extension of grazing activity into forests and semi-arid marginal lands, causing degradation in both zones (Bilsborrow and DeLargy, 1991). For the most part, however, the relationship between human population dynamics and overgrazing is not a matter of the pastoralist population growth driving the changes in livestock density. Rapid changes in the livestock population usually are driven by external forces such as decisions by outsiders (e.g. when sedentary populations entrust cattle to nomad groups) or animal health factors (e.g. eradication of the Tsetse fly). Economic circumstances, or special goals such as the need for security, causing families to increase stocking densities, also play a part. As for proper crises, they frequently are brought about by a rapid decline in pasture resources available. In this context, droughts often are the trigger: "In dryland grazing areas, large numbers of cattle and sheep tend to build up during years of normal rainfall, too many to be supported during years of drought. When the inevitable drought arrives, graziers are naturally reluctant to cut back on herds after a single dry year. When it becomes apparent that the drought will be prolonged and serious, many ranges are already overgrazed" (FAO, 1983). An important population factor, though, is the growth of neighbouring populations, inasmuch as it leads sedentary farmers to expand the cultivated area. With the reduction of fallow and the advance of agricultural "frontiers", nomad pastoralists are increasingly restricted in their movements, and available ranges decline in quantity and quality (e.g. Little and Horowitz, 1987; Bilsborrow, 1992a). This increases density even if livestock size remains the same.12/ Degradation here is a side effect of agricultural extensification or intensification. (It is to reconcile the different logics of agriculture and pastoralism and forestall such conflicts that FAO promotes integrated agro-pastoral development models.) It is thus reported (Talbot, 1989) that in Kenya, population growth among both the Maasai pastoralists and the sedentary agricultural population led to competition for land between the two groups, overgrazing and "desertification" in certain areas. The above-cited Sudan case has a germane aspect in that the increase in fuelwood demand, which contributed to deforestation and therefore degradation, was enhanced by the growth in the population of subsistence farmers. The sedentarization of nomads, leading to the concentration of populations and herds on restricted ranges, has similar effects (Fratkin, 1981; Little, 1987). So do political conflicts that "contribute to population and livestock concentration, which in turn perpetuate ecological degradation and food shortages" (Hjort af Ornäs and Salih, 1991). 3.2.3 Improper agricultural management A set of improper practices has to do with land extension, the main problem being the gradual extension of cultivation to sloping areas and, in general, to so-called "marginal" areas (previously left aside because of the fragility of soils or of other limiting factors). This is a common phenomenon in situations of "land hunger", i.e. of high population density vis-a-vis arable land. Population growth "requires the extension of interference into new areas, and the subjection of these areas to the high levels of damage that follow initial interference. It requires the occupation of sites of lower resilience and higher sensitivity, for which existing management practices may be inadequate" (Blaikie and Brookfield, 1987). Degradation then sets in, unless particular measures are taken to protect soil structure and maintain fertility. But such measures usually are absent, since this kind of practices takes place in situations where low-cost solutions are sought because resources are lacking to invest in land protection. Examples of populations driven upland by the saturation of lowland resources, with ensuing degradation and at times ecological collapse, are numerous: Ethiopia, Haiti, Nepal and the Philippines being perhaps the best known. Population pressures play an obvious role in most of these situations, but it must also be noted that unequal land distribution can worsen those pressures notably. A different set of improper practices has to do with faulty intensification: shortening fallow, insufficient fertilization, excessive fertilization, or the various forms of inadequate management of irrigated areas. The elementary way to extract (in the short run) more produce from land which is not cultivated permanently, is to shorten the fallow period to which it is subjected. Now, shifting cultivation "works well where the ratio of land to people is high" so that the land can "be left fallow long enough [...] The chief problem with shifting cultivation today is that increasing populations and the need for higher production to feed them are pressuring many farmers to shorten or even eliminate the fallow [...] As a consequence [...] yields are lower and soil damage greater" (FAO, 1983). The above process is verified unless increasing fertilization compensates the increased rate of use of the land. Of course, insufficient fertilization as such ■whether under fallow or permanent cropping■has the same effect. Numerous examples of this process have been documented. In Africa, rapid growth in population densities has usually not led to deep changes in traditional production systems. Some intensification has taken place in the form of reduced fallow but, in the absence of other technological changes, this has led "growing poor rural populations to increasingly degrade and mine the natural resources [...] to ensure their day-to-day survival" (Cleaver and Schreiber, 1992). In the West African Sahelian and Sudanian zones, traditional production systems included techniques and enforceable rules for assuring sustainable use of the modest and fragile (low soil fertility, variable rainfall) resource base; those systems "have increasingly been disrupted, above all by rapid population growth" (Gorse and Steeds, 1987). In Egypt, "the land degradation situation in the Nile Valley has worsened markedly due to: (a) the pressure of an increasing population combined with the scarcity of cultivable land, leading farmers to ask more of the land than it can yield" (Kishk, 1986). Under permanent cropping, the need for increased produce leads to irrigation and fertilization, as well as to higher cropping intensities (multiple crops during the course of one year). As seen earlier, the typical land degradation problems arising from such practices are salinization and waterlogging of irrigated areas, and pollution by pesticides or fertilizers. The first two problems are pervasive■they actually affect more than a third of irrigated areas worldwide. Fertilizing is frequently pointed at: "In some regions of the developing world, notably areas in Asia with highly intensified rice and wheat production, excessive fertilizer use poses serious environmental risks" (Pinstrup-Andersen and Pandya-Lorch, 1994). A variant is that "the principal cause of environmental effects is unscientific fertilizer practices and not excessively high rates of application" (Rustagi and Desai, 1993). Pesticides--intrinsically, poisons--are another classical culprit: "improper pesticide use is common across much of the world [...] for most insect pests, only small amounts of insecticides are required, and [...] a large share of the insecticides applied is essentially wasted" (Pinstrup-Andersen and Pandya-Lorch, 1994). In Egypt, "irrigation practices and intensive agriculture that led to various forms of serious degradation [...] Soils are polluted primarily by pesticides, which are very intensively applied to the fields (in particular the half-million hectares under cotton plantations)" (Kishk, 1986). As seen earlier, the damages caused to land by intensified agriculture are largely avoidable through sound management. It would thus seem that population pressure has no responsibility in that degradation, other than the indirect one of triggering the move towards potentially harmful production techniques. On the other hand, population pressure, by reducing per caput access not only to land but also to other resources, can lead to "cheap" intensification: hence insufficient drainage (and waterlogging), insufficient fertilization (and loss of soil fertility), or inadequate monitoring of irrigation schemes, for instance. Safeguarding sustainability during adjustments in production systems is all the more problematic as population growth is more rapid and adaptations must be devised and executed in haste. This is the case in much of sub-Saharan Africa. On the basis of cases from Cameroon, Kenya, Malawi, Nigeria, Senegal and Tanzania, Lele and Stone (1989) have shown that when population growth is rapid, the adaptations involved in the "autonomous intensification" described by Boserup are outweighed by the environmental damage caused by deforestation and decline in soil fertility: higher yields and larger incomes do not necessarily follow from higher population densities or more frequent cropping. Pingali and Binswanger (1988), also using evidence from Africa, concluded that farmer-generated technical change is capable of sustaining slow-growing populations, but not rapid growth in both rural population and urban demand for food. In part, this is because high population pressure can "create stresses within existing systems with well-tried management practices. As the margin of subsistence grows narrower, so the pressure to maximize short-term production will grow stronger. The need to innovate will grow, but the means with which to innovate will be lacking"; as for the wealthier landowners "whose own resources are not gravely threatened by the ■downstream■ effects of degradation on the land of their poorer neighbours, [they] may welcome the growing abundance of cheap labour and see no need to embark on larger innovations which might be of benefit to all" (Blaikie and Brookfield, 1987). Clearly, these matters of technological change are crucial. Technical stagnation renders resources critically vulnerable to population pressure. On the other hand, adaptations in land management may provide wide margins for the accommodation of growing populations. The next section examines how this dilemma relates with the population variable. 3.3 Population and technological factors When one reviews the experience of agricultural systems under population pressure, the factors which happen to have been stagnating (or changing slowly) in a given case tend to appear as the weak elements of the system. In many rural settings■ particularly but not only in Africa■population has grown rapidly during the past 20 years or so, while technology and consumption levels have stagnated (or worse) and land degradation has accelerated. This could suggest that when population growth is rapid it becomes the decisive factor for the final outcome. Yet, when "technology" in turn undergoes a rapid adaptation process, its changes can offset the effects of total consumption growth (whatever the respective roles of population and per caput consumption growth in the latter). Cases of such occurrences are not too hard to find. Java for instance, "despite the serious erosion that takes place in headwater areas and on land of high environmental sensitivity that is unsuited to irrigated terracing, exemplifies the high productivity obtainable under intensive management with extremely high densities of population" (Blaikie and Brookfield, 1987). Mortimore (1993) pointed out that the Close-Settled Zone of Kano (Nigeria) exhibited a stable agricultural system in a dryland area, despite the high population density. In a study of Kenya, Nigeria, Rwanda, Tanzania and Uganda, Hyden et al. (1993) also found that in certain places "farmers have managed their lands, even under severe pressures, in a manner that has permitted sustained use to date", and concluded that high population densities could be accommodated in many parts of East Africa. Tiffen et al. (1994) described a remarkable "success story" observed in the Machakos District of Kenya (see Annex 5). They saw the case as confirming "the autonomous effects of an increased population, deriving from the availability of more mouths (more demand), more hands (more labour), and more brains (more people interacting more), accompanied by a reduction in the per capita costs of physical and social infrastructure". What conclusions can we draw from the variety of contradictory experiences? The "success stories" first need to be qualified. Out-migration has been a component of family strategies in most of the places surveyed. For one reason or another (possibly but not necessarily diversification) there has been, so to speak, an escape from local agriculture. In fact, families, not the agricultural system, have adapted to increasing pressures. The same observation would apply to Machakos.13/ In the context reviewed by Hyden et al. (1993), taking into account the intensified female labour contribution in situ to compensate male migration, the authors saw "evidence that the rural populace is working longer hours to feed itself". In other words, labour productivity had declined, and net wellbeing with it. As to environmental impact, inasmuch as families resort more to purchases of agricultural products, possible damages are simply shifted to the areas where production takes place. Otherwise, these stories do not show that rapid population growth has a positive effect on environmental and economic outcomes, but simply that it does not necessarily lead to a catastrophe■surely a well-accepted statement anyway. There is no evidence that the same (or greater) improvements could not have been made under slower population growth; there is no evidence either that the improvements are due to population growth rather than to any other factor. In fact, the benefits listed by Tiffen et al. arise not from population growth, but from a sufficient population density. They may be convincingly associated with a density threshold, but it is unlikely (as admitted by the authors) that they will continue to improve indefinitely if population density keeps growing at the same pace, as should be the case if population growth had an intrinsic positive effect. And there is of course no guarantee that comparable improvements will occur for other populations growing at the same rate (although that possibility is not excluded either, provided that favourable economic conditions exist and policies are adequate). It would be easy to point to situations (in Ethiopia and other parts of Africa, the Philippines, Haiti, etc.) where population growth conditions of the same kind as those observed in Machakos (or indeed more benign) were associated with stagnation or even ecological collapse. Naturally, such cases do not prove, either, that "rapid population growth leads inexorably to environmental degradation", the proposal which Tiffen et al. purport to disprove, but that few, if any, actually propound. In fact, the only correct question is: Would the outcome have been better (or worse, or identical) if population growth had been slower? Clearly, neither observation nor experimental methods can provide a definitive answer to such a question, so any answer is debatable. That technological change may be driven by population growth (Boserup, 1965, 1981; Simon, 1986) is a matter of common wisdom: "Necessity is the mother of invention". But, why ascribe to population growth alone the effect of technological changes, including those which bring levels of wellbeing above what they were before the population reacted to reduced per caput resources?14/ There seems to be no other answer than the desire to improve levels of living, in which case this can just as well occur in the sheer absence of population growth. The hardship caused by diminishing resources per caput may be a stronger incentive to innovate than aspirations to more wellbeing, but "population pressures are a clumsy and cruel stimulant to development" (Hirschman, 1958). Favouring such pressures (with their uncertain efficiency and the attendant problems of maternal and child health in high fertility settings) in the hope of production gains is an odd proposition. Policy-wise, it is critically important to address the ambiguity in Boserup's hypothesis regarding the innovation process itself. Blaikie and Brookfield (1987) ask: what is it that makes population pressure result in degradation rather than innovation? They point to a variety of explanations, including "the lack of access to productive resources on the part of the cultivator", and revolving around the various kinds of pressures which lead farmers to extract more from the land than it can sustainably give. The next section provides an overview of those social and institutional factors which mediate between pressures for increasing output on the one hand, and the actual changes in land use on the other. 3.4 Social and institutional factors 3.4.1 Agrarian structures and poverty Land degradation on a holding depends in part on how intensively the land is exploited, and in part on the holder's ability and willingness to undertake conservation measures. These two factors in turn are influenced by farm size■although not in an entirely linear manner. Consider the contrast between large and small farms in an agroecological zone. Small holdings may be "mined" in order to extract enough for family subsistence; their holders cannot afford leaving a large portion of the farm under fallow; the output does not enable long-term investment in soil conservation or amelioration, or productivity-raising implements. At the same time large (at times absentee) landholders, who maintain or increase their well-being simply by concentrating resources, need a lesser, more easily sustainable rate of use: hence extensive exploitation schemes. They also can easily take land out of production for anti-erosion works. Certainly, faulty practices or an excessive rate of use can also be found on larger farms; what is meant here is that they are not a necessity in that case, because of their high resources-per-person ratio. On the other hand, it has been argued that labour for conservation works may be more readily available on smaller farms in densely populated areas. Large estates would typically have to hire labour for this task as they do for other purposes. An extreme aspect of this question is the situation of areas where out-migration has left too little manpower to carry out conservation works, e.g. maintain terraces (Collins, 1987). A "natural" factor in the fragmentation of land into small holdings is population growth. "All across the developing world, farm size is shrinking as farmers continue to subdivide holdings among their children. In countries such as Malawi, Rwanda, Haiti, and Bangladesh, population growth rates are high, and the non-farm sector is still in its early stages of development. Farms now average less than 0.5 hectares in some areas. Ever-increasing numbers [...] have become nearly or entirely landless" (Clay et al., 1994). Another factor, of course, is social inequality within the population leading to skewed structures of land ownership. Pressures towards land degradation are stronger in this case, because land quality and vulnerability are usually not equally shared either: "Inequities in land ownership may also encourage soil erosion. In Andean Latin America, for example, wealthy ranchers often use the relatively level valley floors to graze cattle, forcing the small, poor landowners onto the steep slopes to produce subsistence crops" (FAO, 1983). If the smaller holdings occupy marginal, more vulnerable areas such as slopes or poorer soils in need of longer fallow or fertilizing, not only such areas will be needlessly settled, but also they will likely be overexploited as their occupants cannot afford restraint in resource use. The respective weights of demographic pressure and social inequality in causing land fragmentation vary from place to place; certainly, both aspects must be tackled.15/ Population pressure in turn contributes to unequal practices, because a deteriorating population/resources situation leading to decreasing average well-being contributes to trigger or accelerate land concentration: "With more people, the increased demand for food results in increased competition for arable land, tending to change land prices. In the common situation in which farmers with small plots have much less access to credit and new technology than those with large plots, this may result over time in a smaller proportion of the rural population owning land, smaller average size of plots for the majority of small farmers who continue to own land, and an increase in the average size of large farms. This process of increasing socioeconomic differentiation has been well documented in Latin America and may be occurring in Africa and Asia as well" (Bilsborrow and DeLargy, 1991). Overall, poverty is usually seen as adding considerably to resource overuse in developing countries. "Poor households are often virtually forced to overuse natural resources for daily subsistence. Thus, landless farmers colonize tropical forests, or grow cassava and maize on highly erodible hillsides. Rural households in fuelwood- deficit countries strip foliage and burn crop and animal residues for fuel rather than using them for fertilizer and this contributes to desertification. Underemployed men in coastal villages overexploit already depleted inshore fisheries. A cycle of poverty and natural resource degradation is established" (Repetto, 1987). For Blaikie and Brookfield (1987), "since the expansion is carried out largely by those displaced from older areas by poverty, or by other pressures of social or political origin, the new land has to be managed by those with the fewest resources to devote or divert to its management"; Nepal provides an illustration of those situations where "poverty is the basic cause of poor management, and the consequence of poor management is deepening poverty". But this view has also been said to be superficial. While recognizing that the poor "are more likely to gather free fuels [...] gather every last dried piece of dung for fuel, instead of leaving it to fertilize the soil [or] migrate for work, leaving the wife at home too burdened to take on any extra work to conserve the soil", Harrison (1992) notes that "bigger farmers [...] are more likely to use tractors [or] own more livestock [who, if] not properly managed, can do more environmental damage than humans". He points at situations such as that of Lesotho where the poorest "possess neither fields nor livestock. Since they have no access to land, they cannot degrade it. [...] those who do most damage [are those] who own cattle and, among these, the wealthiest 23 percent who own 74 percent of the cattle. Livestock degrade the highlands in the summer months. In winter they eat stubble and trample down terrace edges". Harrison also sees the tendency to move into forest or marginal land more as a matter of generation (the young being the prime candidates) rather than of socioeconomic category. As for getting hold of large concessions of forest land, clearing and farming them with hired labour and tractors, the better-off are clearly the most likely to do that. He notes that "even before it is degraded, a marginal area by nature does not usually produce enough surplus to lift its inhabitants out of poverty. Poor areas and poor people destroy each other". There is much value in this analysis, as well as in the observation that "consumption and waste per person is also lowest among the poorest" and the conclusion that "all in all, the poor probably tread lightest of all upon the earth, and do less damage to the environment than any other group. They are victims, not perpetrators" (Harrison, 1992).16/ It is also worth remembering that average access to natural resources is unquestionably affected■among others, but often in a big way■by population density: population pressure "is an important and reinforcing link in reducing that access to sectors of an agrarian population" so that, while not causing inevitably land degradation, it "may almost inevitably lead to extreme poverty when it occurs in underdeveloped, mainly rural, countries" (Blaikie and Brookfield, 1987). 3.4.2 Land tenure With regard to factors which foster or discourage land conservation efforts, it has been argued that only private ownership makes it worthwhile for peasants to care about the sustainability of their farming methods: "systems of land ownership as well as tenure and business arrangements which do not provide security to the farmer" are held to be "major obstacles to conservation" (FAO, 1983). A number of authors ■notably E. Boserup■have thus contended that the move from collective to individual land ownership, which usually emerges as land becomes scarce, promotes investments in the productivity of landholdings. Much has been made, in this respect, of the overuse of common property resources (CPR), a feature which is present in many societies: "Population growth is most likely to result in land degradation when land is held in common without rules governing its access" (Jolly and Torrey, 1993). The fate of CPR has been much studied, e.g. in India by Jodha (1991), who noted that increased population, along with changes in market relations and land privatization, have led to declines in CPR size, increased pressure, dwindling communal management; all this precipitated land degradation. Cleaver and Schreiber (1992) also observed that rapid population growth often is responsible for breakdowns in communal land management, failures in resource control and local "tragedies of the commons". But communal tenure does not have to lead to such outcomes. Where strong social and cultural sanctions hold, use can remain sustainable; the problem actually lies with open access resources.17/ Also, traditional tenure systems have often been misunderstood and underrated. For instance, "it is often suggested that communal tenure is the norm in West Africa and that individuals therefore have little incentive to make any long-term on-farm investments. This argument is questionable on three counts. First, the term "communal tenure" is used very loosely to cover different forms of ownership (by a chief, by a lineage...) and, more important, different forms of management (by entire groups, by representatives on behalf of group members, by individuals). [Second,] not all forms of communal tenure entail disincentives for long-term investments". On balance, lack of individual tenure does not appear to be an unsurmountable obstacle (Gorse and Steeds, 1987), as long as population pressure or other factors do not erode the rules which are essential to the system. The "Boserup argument" also "fails to address whether the induced production changes are sustainable [and] does not deal with the continuing changes in tenure that often occur after the change to individual ownership", in particular the distributional aspects: "as individual owners acquire land, the potential grows for concentration of land in the hands of a few. In effect, the development of Western-style land tenure systems in Africa has sometimes led to the concentration of rights of access in certain groups and removed indigenous means of determining usage (Jolly and Torrey, 1993). Then rental and share arrangements emerge between large landowners and those without sufficient productive land; but "renters are less likely to make long-term investments, increasing the potential for degradation" (Clay et al., 1994). Indeed a number of studies have illustrated the differences of behaviour between individual owners and renters. Rented lands usually are the most degraded. But a closer look often reveals that renters with long-term use rights can be quite as inclined to improve the land as owners are. Security of tenure, not ownership, is decisive, because it enables farmers to reap the benefits from their investments (or from their restraint). Conversely short-term land leases are "among the most pernicious" arrangements from this standpoint (FAO, 1983). 3.4.3 Markets and public policies Many of the economic changes typically associated with the idea of "modernization", including the economic role of the modern state, have been seen to negatively affect the management of natural resources, both at the local level by community authorities and by individual peasants. In many places, "increasing monetarization produced marked changes in the institutions [...] extended family structures and their careful resource management practices [broke] down, and the authority of local communities, which might have taken political measures to control resource abuse [...] was increasingly constrained [...] Increasingly centralized political authority has also challenged the capability of local decision-making bodies to manage their environment" (Gorse and Steeds, 1987). The pervasive "urban bias" in macro-economic management also played its role, for instance by promoting "cheap food and fuel for urban consumers [...] To the extent that low producer prices discouraged more intensive production, and the unpredictable behaviour of public marketing agencies increased farmers' risks, land-clearing for further extensive production and/or the shortening of fallow periods was promoted" (Gorse and Steeds, 1987).18/ More generally, cheap food and low agricultural prices have kept land value low, making its conservation unattractive. The promotion of cash crops by governments in the pursuit of export gains also has often accelerated soil exhaustion, because the main crops in this group (cocoa, coffee, cotton) happen to be very nutrient-demanding. Water for irrigation has been vastly underpriced, leading to overuse and related problems. Inadequate credit facilities have rendered access to modern implements and inputs difficult for small farmers, and are thus largely responsible for failures to intensify cultivation. Added value generated by farmers has been confiscated by state marketing institutions. In sum, the distortions of agricultural policies beg redress also for the sake of land conservation. 4. CONCLUDING REMARKS 4.1 Population in the chains of explanation of degradation It must be clear at this stage that land degradation is the result of many factors■some outside human control■and that it is "futile to search for a uni-causal model of explanation" (Blaikie and Brookfield, 1987). It is probably impossible to argue satisfactorily that one of the main categories of factors is generally decisive. For one thing, the variability of situations at the local level is too great to support any generalization. For another, population change, social factors and technological factors are interlinked, so that it is impossible to assign them autonomous effects.19/ But, while no general truth can be proposed, it is necessary in situations of ongoing or impending land degradation to look for the factors on which to intervene. One useful concept in this respect is that of "chain of explanation": the chain "starts with the land managers and their direct relations with the land (crop rotations, fuelwood use, stocking densities, capital investments and so on). [The] next link concerns their relations with each other, other land users, and groups in the wider society who affect them in any way, which in turn determines land management. The state and the world economy constitute the last links in the chain" (Blaikie and Brookfield, 1987). A priori, population pressure seems to apply on the very first links of the chain. Except in accidental collapses of production systems under exogenous forces (drought, war...) the common element in land degradation is "pressure of production on resources" (Pavelis, 1983). That pressure can arise from various factors including a large or growing population, outside market demands, or the nature of crops or livestock-raising. It can also arise from institutional, social and economic conditions which lead to the extraction of surpluses from the land managers, forcing them in turn to extract from the land more than is sustainable. Such conditions are: heavy tax and tribute; very low wages; denial of access to CPR; low commodity prices due to state pricing policies or market distortions; farmer's indebtedness; and so on. In this context, population factors appear both as part of the basic conditions within which the socio-economic system operates (population density with regard to resources) and of the forces which affect its patterns of change (population growth, urbanization, migration): density is relevant to the level of direct pressure on resources; population growth and urbanization affect the volume of market demands; urbanization absorbs land, and is conducive to biased pricing policies; a large and growing rural labour force contributes to low wages; excess demand for access to CPRs may lead to shut out part of the population. 4.2 Relevance for population programmes Policy-wise, tackling the land degradation issue will generally entail addressing two separate questions: (1) How is land degradation brought about? and (2) Why does land management fail to be effective? (Blaikie and Brookfield, 1987). While the latter question leads to seek interventions in the economic and social sphere, the former directs attention to population and other "pressures". But both require a correct understanding of the situation at hand. Population growth usually appears as the major cause for environmental (e.g. land) degradation in situations where other elements of the local system (consumption levels, production techniques, institutions, social structures) are stagnant. It seems natural then to causally associate the changing elements, i.e. the growing population and the worsening state of the environment. This is not to deny that appropriate changes in the socio- economic and institutional setting would have positive effects on environmental dynamics, enabling the system to better absorb the stresses brought about by population growth. But the latter proposition is of a hypothetical nature, and not a priori superior to the alternative hypothesis that if population growth were slower, environmental stress would be smaller. As far as interpreting the facts is concerned, one can only say that degradation was driven by population growth and enabled by unfavourable socio-economic conditions. In those cases where a series of adverse changes have occurred (population growth, growing agrarian inequities, deteriorating terms of trade for agriculture etc.), designating the main culprit of land degradation (hence the main target for policy intervention) seems an impossible problem to solve by objective means. In fact, the diagnoses given in literature often seem dictated by intuition (if not by prejudice) rather than by sound analysis. But an interesting aspect emerges when diagnoses are viewed in a policy perspective. Take for instance the following conclusive statement from a case study (DeWalt et al., 1993) in environmental degradation: In southern Honduras, environmental degradation and social problems often attributed to population pressure arise from glaring inequalities in the distribution of land, the lack of decent employment opportunities, and the stark poverty of many of the inhabitants. It is not the carrying capacity of the land that has failed to keep pace with population growth. Neither is population growth the primary cause of the impoverishment of the Honduran ecology and its human inhabitants. The statement may be correct, but its policy implications are very weak. It suggests that it is unequal land distribution, lack of employment and poverty that must be attacked in priority for the sake of environmental relief. But those evils ought to be attacked in their own right, regardless of the effect of that attack on the environment! As for population, policies aiming at slower growth never are justified by the sole purpose of limiting environmental impact: there usually are, by common standards, more direct, important benefits to such policies (including alleviating some of the pressures leading to inequalities in access to resources, unemployment and poverty). The fact remains that population growth entails a greater rate of exploitation of natural resources. Adjustments can be made on other factors to diminish that impact, and where those other factors have the main role, the potential for action is great. But adjustments have costs, constraints may limit them, and it is best anyway to tackle all the factors rather than a few, especially when it is known that a "vicious circle" type of dynamics is at work. An additional observation is needed. In considering the possible value of changes in population dynamics, the role of labour force dynamics is critical. All members of the population are roughly equivalent contributors to demand for land products, but members of the labour force are parts of the production system, and therefore their movements affect not only overall population density but also the functioning of that system. For instance, where high population densities have been put to use to develop labour-intensive land management systems, "such systems require abundance of labour [...] also for their maintenance. If some of that labour is withdrawn, as by an increase in off-farm employment opportunities, or by emigration [...] the consequences can be disastrous" (Blaikie and Brookfield, 1987). Barring such circumstances, however, it usually pays to alleviate excessive labour pressure on the land by seeking to diversify economic opportunities. But the earlier observation serves as a reminder that action must be based on solid knowledge of the system one deals with. In view of the linkages identified in the preceding pages, it appears that population-oriented research can contribute to situation assessment and policy formulation with regard to land degradation problems. The UNCED has outlined some relevant ideas in this respect (United Nations, 1992), emphasizing the need for continuous improvements in research, communication with decision makers and public information.20/ A first objective would be "to assess human vulnerability in ecologically sensitive areas", so as to identify priorities for action. For this purpose, the study of demographic trends should be part of all studies of changes in land use and quality (especially at the local level). In the longer run, through accumulated knowledge enabling comparative studies, the aim is "to develop a better understanding of the relationships among demographic dynamics, technology," cultural behavioral norms, and land resources. This requires strengthened interdisciplinary research, emphasizing community-level experience. The eventual aim, where land use is concerned, is the formulation of integrated policies taking into account population concerns. In this respect an immediate practical output should be an identification of vulnerable geographic areas (taking into account inter alia trends in population) as well as of vulnerable populations (which are not necessarily those living in vulnerable areas). UNCED also recommended country assessments of population-supporting capacities, which have rarely been undertaken so far. The relevance of demographic features appears also "in formulating human settlement policies", for good planning of land use entails taking account "of resource needs, waste production and ecosystem health". In effect, policies dealing with population distribution and migration appear clearly more relevant than those affecting national population growth in our case. For populations in subsistence economy, trends in density are directly relevant to assess pressure on the land. But it remains necessary in all cases to take into account the pressure exerted by market demands from urban populations. Finally, to further integration of the respective concerns, "population programmes should be implemented along with natural resource management and development programmes at the local level that will ensure sustainable use of natural resources, improve the quality of life of the people and enhance environmental quality". This will entail developing locally appropriate frameworks for action, based on a participatory process, giving special attention to social features-in particular the role of women as resource managers. REFERENCES Bilsborrow, R. (1992a): Rural poverty, migration and the environment in developing countries. Three case studies. Washington: World Bank. Bilsborrow, R. (1992b): Population growth, internal migration, and environmental degradation in rural areas of developing countries. European Journal of Population, Vol. 8, pp. 125-148. Bilsborrow, R. and P. DeLargy (1991): Population growth, natural resource use and migration in the third world: the cases of Guatemala and Sudan. In: Resources, environment and population (K. Davis and M.S. Bernstam, eds.), Supplement to Population and Development Review, pp. 125-147. New York: Population Council/Oxford University Press. Blaikie, P. and H. Brookfield (1987): Land degradation and society. London: Methuen. Boserup, E. (1965): The conditions of agricultural growth: the economics of agrarian change under population pressure. London: Allen & Unwin. Boserup, E. (1981): Population and technological change. Chicago: Chicago University Press. Brown, L. et al. (1990): State of the world 1990. New York: W. W. Norton. Clay, D.C., M. Guizlo and S. Wallace (1994): Population and land degradation. EPAT/MUCIA Working Paper No 14. Madison: University of Wisconsin. Cleaver, K.M. and G. Schreiber (1992): The population, agriculture and environment nexus in sub-Saharan Africa. World Bank Working Paper No 6. Washington. Collins, J. (1987): Labour scarcity and ecological change. In: Lands at risk in the third world: local level perspectives (P. Little and M. Horowitz, eds.), pp. 19-37. Boulder: Westview Press. Cruz, M.C.J. (1994): Population pressure and land degradation in developing countries. In: Population, environment and development, pp. 135-147. New York: United Nations. DeWalt, B.R., S.C. Stonich and S.L. Hamilton (1993): Honduras: population, inequality, and resource destruction. In: Population and land use in developing countries (C.L. Jolly and B.B. Torrey, eds.), pp. 106-123. Washington: National Research Council/National Academy Press. Engelman, R. and P. LeRoy (1995): Conserving land. Population and sustainable food production. Washington: Population Action International. FAO (1980): Natural resources and the human environment for food and agriculture. Environment Paper No 1. Rome. FAO (1983): Keeping the land alive. Soil erosion - its causes and cures (by H.W. Kelley), Soils Bulletin No 50. Rome. FAO (1984): Land, food and people. Rome. FAO (1986a): African agriculture: the next 25 years. Rome. FAO (1986b): Atlas of African Agriculture. Rome. FAO (1992): Protect and produce. (Revised edition.) Rome. FAO (1993a): Key aspects of strategies for the sustainable development of drylands. Rome. FAO (1993b): Sustainable development of drylands and combating desertification. Rome. FAO (1994): Water policies and agriculture. (Reprint of special chapter of The state of food and agriculture 1993.) Rome. Fratkin, E. (1991): Surviving drought and development: Ariaal pastoralists of northern Kenya. Boulder: Westview Press. Gorse, J.E. and D.R. Steeds (1987): Desertification in the Sahelian and Sudanian zones of West Africa. World Bank Technical Paper No 61. Washington. Hardin, G. (1968): The tragedy of the commons. Science, No. 1262, pp. 1243-1248. Harrison, P. (1992): The third revolution. Environment, population and a sustainable world. London: I.B.Tauris. Hirschman, A. (1958): The strategy of economic development. New Haven: Yale University Press. Hjort af Ornäs, A. and M. Salih (1991): Research and development issues for African drylands, Ambio, Vol. XX, No. 8, pp. 388-394. Hyden, G., R. Kates and B. Turner (1993): Beyond intensification. In: Population growth and agricultural change in Africa (B. Turner, G. Hyden and R. Cates, eds.). Gainsville: University of Florida Press. ISRIC/UNEP (1991): World map of the status of human-induced soil degradation (by L.R. Oldeman, R.T.A. Hakkeling and W.G.Sombroek). Global Assessment of Soil Degradation (GLASOD), 2nd revised edition. Wageningen/Nairobi. Jodha, N. (1991): Population growth and common property resources: micro-level evidence from India. In: Consequences of rapid population growth in developing countries, pp. 209-230. New York: United Nations. Jolly, C.L. and Torrey, B.B. (1993): Introduction. In: Population and land use in developing countries (C.L. Jolly and B.B. Torrey, eds.). Washington: National Research Council/ National Academy Press. Kishk, M.A. (1986): Land degradation in the Nile Valley, Ambio, Vol. 15, No 4, pp. 226-230. Lele, U. and S.W. Stone (1989): Population pressure, the environment and agricultural intensification: variations on the Boserup hypothesis. MADIA Discussion Paper No 4. Washington: The World Bank. Liamine, N. (1993): Desertification - The issues and challenges, Ecoforum, Vol 17, No 1/2 (April) pp. 11-21. Little, P. (1987): Land use conflicts in the agricultural/ pastoral borderlands: the case of Kenya. In: Lands at risk in the third world: local-level perspectives (P. Little and M. Horowitz, eds.), pp. 195-212. Boulder/London: Westview Press. Little, P. and M. Horowitz, eds. (1987): Lands at risk in the third world: local-level perspectives. Boulder/London: Westview Press. Maloney, C. (1990-1991): Environmental and project displacement of population in India. Two parts. Field Staff Reports (Asia) Nos 14 and 19. Sausalito: Universities Field Staff International/Natural Heritage Institute. Marcoux, A. (1994): Population and water resources. (Population and the environment: a review of issues and concepts for population programmes staff. Part I.) Rome: FAO. Mortimore, M. (1993): Northern Nigeria: land transformation under agricultural intensification. In: Population and land use in developing countries (C.L. Jolly and B.B. Torrey, eds.), pp. 42-69. Washington: National Research Council/National Academy Press. Nelson, R. (1990): Dryland management - The "desertification" problem. World Bank Technical Paper No 116. Washington. Norse, D., C. James, B.J. Skinner and Q. Zhao (1992): Agriculture, land use and degradation. In: An agenda of science for environment and development into the 21st century (Dooge et al., eds.). Cambridge: Cambridge University Press. Pavelis, G.A. (1983): Conservation capital in the United States, 1935-1980. Journal of Soil and Water Conservation, Vol. 38, pp. 455-458. Pingali, P. and H.P. Binswanger (1988): Population density and farming systems: the changing locus of innovations and technical change. In: Population, food and rural development (R.D. Lee et al., eds.), pp. 51-76. Oxford: Clarendon Press. Pinstrup-Andersen, P. and R. Pandya-Lorch (1994): Poverty, agricultural intensification and the environment. Paper presented at the 10th Annual General Meeting of the Pakistan Society of Development Economists, Islamabad. Repetto, R. (1987): Population, resources, environment: an uncertain future, Population Bulletin, Vol. 42, No. 2 (July). Rustagi S. and G.M. Desai, 1993: Fertilizers and environmental concerns. Mimeo. Washington: IFPRI. Sfeir-Younis, A. (1986): Soil conservation in developing countries - A background report. Washington: The World Bank. Simon, J. L. (1986): Theory of population and economic growth. Oxford: Blackwell. Talbot, L. (1989): Demographic factors in the resource depletion and environmental degradation in East African rangelands. In: Population and resources in a changing world: current readings (K. Davis et al., eds.), pp. 315-323. Stanford: Morrison Institute for Population and Resource Studies. Tiffen, M., M. Mortimore and F. Gichuki (1994): Population growth and environmental recovery: policy lessons from Kenya. Gatekeeper Series No 45. London: International Institute for Environment and Development. Umale, D. L. (1993): Irrigation-induced salinity: a growing problem for development and the environment. World Bank Technical Paper No. 215. Washington. UNEP (1992): The world environment 1972-1992. Two decades of challenge (M.K. Tolba et al., eds.). London: Chapman & Hall. United Nations (1977): Round-up, Plan of Action and Resolutions of UNCOD. New York. United Nations (1992): Agenda 21. The United Nations programme of action from Rio. New York. World Resources Institute (1993): World Resources 1992-1993. Washington. Wu, S., S. Xu and S. Wu (1987): Ecological agriculture within a densely populated area in China. Agriculture, Ecosystems and Environment, Vol. 27, pp. 597-607. NOTES 1/ See Marcoux (1994). 2/ Lord John Boyd Orr (first Director-General of FAO), in 1948. 3/ Some soils erode easily under the action of rain and runoff, while others are remarkably resistant, mostly due to their ability to absorb rainfall rapidly (FAO, 1983). That ability is also important because it affects the formation of underground water reserves which are so crucial for agriculture and the populations in general. 4/ See preceding paper in this series. Such problems will increasingly affect the world's coastal areas, currently estimated to be home to two thirds of the world's population. 5/ The idea that the Sahara keeps expanding is much questioned nowadays, on the basis of studies which show that vegetation usually reoccupies the lost ground when rain returns. "The only conclusion is that within a sparsely populated belt of some 200 km at the southern fringe of the Sahara, biological productivity changes from year to year according to rainfall fluctuations. In areas where the soil was destroyed, the decline of biological productivity would be permanent" (UNEP, 1992: pp. 140-141) 6/ The proportion of degraded area to the total area minus wasteland (which may be regarded as the "useful" area) is of course higher: 17% globally, with highs of 25% in Central America, 23% in Europe and 22% in Africa. 7/ The final loss would be less because much of the land lost to crops could be used as pasture; that loss would still amount to 25 percent in Africa and central America, however. 8/ One may distinguish two slightly different types of "enabling" factor. One creates conditions such that degradation can begin (e.g., deforestation in an area where forested land was stable despite water and wind action). The other creates conditions such that degradation is more intense than it would be otherwise (e.g., differing slope or soil texture in pieces of land exposed to the same action of water and wind will result in different rates of degradation; this could be labelled an "accelerating" effect. 9/ "*" denotes a quantity too small to be rounded up to the smallest unit (1% in this case), while "-" means nil. 10/ Many industrial activities are known to have potentially deleterious effects on soil fertility because of emissions of pollutants. Examples of serious damage can easily be found at the local level (especially in eastern Europe), but it is globally marginal in terms of areas affected. As for linkages with population variables, the expansion of such activities is driven by the levels and growth rates of aggregate incomes rather than of population. 11/ See FAO estimates cited in Harrison (1992), p. 100. 12/ This can have an additional, altogether different effect, due to the lesser soil protection under crops as compared to pasture, when dryland farmers "extend croplands into more marginal lands during good years, pushing back graziers in the process. When drought begins, the new cropland lacks defences and the soil may emerge from the drought too degraded even for livestock" (FAO, 1983). 13/ It would also apply to cases where diversification occurred within the agricultural sector. In China, Wu et al. (1987) have documented a more diversified exploitation of local ecosystems (with e.g. development of fishponds) under population pressure. This is different from the scheme of intensification fostered by population growth: on the contrary, there has been exploitation of new resources, i.e. extensification; low productivity of traditional activities has been escaped, not remedied. 14/ This point arises for instance in respect of a diagram in Tiffen et al. (1994: p. 14) titled "Positive effects of population growth". The said effects essentially consist in a larger economy, including of course a larger production (= larger total incomes), but the diagram registers this as "higher per capita incomes". The "per capita" is an unsupported insert. 15/ In a country with very unequal agrarian structures like Guatemala, redistributing land entirely would "buy" no more than twenty years at current population growth rates before a situation of saturation returns (R. Bilsborrow, communication at the Round Table on Population, Environment and Development, International Academy of the Environment, Geneva, 1993). 16/ Much as the poor are sometimes pointed at as the main actors of environmental degradation, women, being the water and fuelwood collectors, or the ones pushed up the slopes with their animals or food crops, also are at times unjustly accused. 17/ In Hardin's words (1968), it is the absence of enforced rules ("freedom in a commons") which "brings ruin to all" through the pursuit by each individual of his/her self-interest. 18/ A similar mechanism has objectively promoted deforestation through inadequate costs of wood collection. 19/ "Any attempt to find the cause of land degradation is somewhat akin to a "whodunnit", except that no criminal will confess, and Hercule Poirot is unable to assemble the suspects [...] for the final confrontation. [...] However, any general statement about the causes of land degradation is of a very different order from the usual ■whodunnit■, except perhaps in the case of the Orient Express, where all the suspects were found guilty!" (Blaikie and Brookfield, 1987). 20/ The quotations in this section are from chapter 5 of Agenda 21.