Cover Image
close this bookSoil Degradation - A Threat to Developing-Country Food Security by 2020? 2020 Vision for Food, Agriculture, and the Environment Discussion Paper 27 (IFPRI, 1999, 70 p.)
View the document(introduction...)
View the documentForeword
View the documentAcknowledgments
View the documentAcronyms and Abbreviations
View the document1. Introduction
View the document2. Evaluating the Impact of Soil Degradation on Food Security
View the document3. Past and Present Effects of Soil Degradation
View the document4. Future Effects of Soil Degradation and Threats to Developing-Country Food Security
View the document5. Policy and Research Priorities
View the documentAppendix - Types of Soil Degradation: A Glossary19
View the documentBibliography
View the documentRecent Food, Agriculture, and the Environment Discussion Papers

2. Evaluating the Impact of Soil Degradation on Food Security

The key soil characteristics that affect yield are nutrient content, waterholding capacity, organic matter content, soil reaction (acidity), topsoil depth, salinity, and soil biomass. Change over time in these characteristics constitutes “degradation” or “improvement.” Degradation processes include erosion, compaction and hard setting, acidification, declining soil organic matter, soil fertility depletion, biological degradation, and soil pollution (Lal and Stewart 1990).4 Soil quality (see box) may be improved through leveling the land, depositing sediment deposition, increasing organic matter, improving soil nutrient status, terracing, controlling erosion, improving irrigation drainage systems, or rehabilitating compacted soils and erosion gullies or other seriously degraded areas.

4See the appendix for definitions of the various types of degradation.

Change in soil quality over time can be a complex phenomenon. Quality can vary across sites, soil types, and production systems. Furthermore, soil quality is only one of many variables influencing agricultural yield, which is, in turn, only one of many factors influencing food consumption, food availability, and farm income. This complicates the evaluation and interpretation of the effects of soil degradation and the design of appropriate policies in response.

Box - Agricultural Productivity and Soil Quality

Soil quality is the inherent capability of the soil to perform a range of productive, environmental, and habitat functions. This study is concerned mainly with the soil’s productive function, hence it is important that the definitions of productivity used below in relation to soil quality are clear.

Diverse definitions of “productivity” have created some confusion. In this paper, the term “potential soil productivity” is used to refer to the potential of the soil system to accumulate energy in the form of vegetation (following Tengberg and Stocking 1997, 4), controlling for the use of other inputs. “Soil productivity” is used to refer to the actual yield of usable vegetation, also controlling for input use. “Agricultural productivity” refers to the relationship between the average or real output of economically usable products divided by an index of all fixed and variable inputs. Because economists conventionally have analyzed “land productivity” simply as total output divided by land area (assumed to be a fixed factor), soil quality has not been considered. Yet measures of change in “total factor productivity” over time that do not include soil quality are likely to overestimate the contribution of other factors. On the other hand, the effect of soil quality change on agricultural productivity is limited by its importance as a productive factor relative to other factors, and the degree of complementarity and substitutability between soil quality and other factors and inputs. Soil quality contributes relatively more to agricultural productivity in low-input production systems.

Vulnerability of Soils to Degradation

The widespread tendency to minimize the importance of soil quality for agriculture stems in part from the experience of temperate agriculture. The most productive temperate soils are geologically “new.” A result of glaciation in the last Ice Age, these soils are both fertile and relatively resistant to degradation. By contrast, though some tropical highland soils are also “new,” formed through the deposition of volcanic materials from old eruptions, most are of infertile parent material or have been highly weathered over the millenia, resulting in the leaching of soluble nutrients from soils and acidification. The higher temperatures, greater high and low extremes of rainfall, and greater rainfall intensity typical of the tropics subject soils in most developing countries to significant risk of climate-induced degradation.

Indeed, only a third of all rainfed, cultivable area in developing countries (excluding China, for which data were not available) is free of major soil-related constraints that limit production (Table 1). The 10 percent of land in steep slopes is especially prone to erosion, as are shallow soils; the extensive areas with low natural fertility require active nutrient replenishment and supplementation to sustain even moderate yields over time; and sandy soils require careful management to retain water. Chemical soil constraints are also widespread: 36 percent of tropical soils have low nutrient status; one-third have sufficiently acid conditions for soluble aluminum to be toxic for most crops (acidity is exacerbated by inorganic fertilizer application); 22 percent are tropical clays that fix phosphorus; 5 percent have critically low cation exchange capacity; and some are saline or alkaline (Sanchez and Logan 1992, cited in Tengberg and Stocking 1997, 9-10).

Poor land husbandry can have quite different long-term effects on different types of soils, and costs of and returns to soil improvement can vary substantially, depending upon soil resilience (the resistance to degradation) and soil sensitivity (the degree to which soils degrade when subjected to degradation processes). For example, ferralsols, which have low available nutrient supplies, strong acidity, low available phosphorus, no reserves of weatherable minerals, and easily lost topsoil organic matter, demonstrate low resilience and moderate sensitivity to water erosion. Even with good soil cover, yields decline rapidly without a combination of structures and biological measures to control erosion. By contrast, luvisols, with moderate nutrient levels, low-to-moderate organic matter content, and weak topsoil structure prone to crusting, have moderate resilience and low-to-moderate sensitivity. Maintaining their productivity requires both tillage practices that maximize surface water infiltration and biological measures that maintain soil cover (Tengberg and Stocking 1997; see Figure 1). While some soils, like alfisols, can be maintained for a long time with only inorganic fertilizer application (if farmers make sure that they do not crust), Luvisols require complementary use of organic inputs because they are low in organic matter to begin with (Swift 1997).

Table 1 - Share of land with terrain and soil constraints in total rainfed land with crop production potential


Sub-Saharan Africa

Latin America and the Caribbean

Near East/North Africa

East Asia (excluding China)

South Asia

Developing countries (excluding China)


Steep slopes (16-45 percent)







Shallow soils (<50 centimeters)







Low natural fertility







Poor soil drainage







Sandy or stony soils







Salinity, sodicity, or excess of gypsum







Total land with crop production potential affected by one or more constraintsa







No major constraints







Source: N. Alexandratos 1995, Table 4.2, p. 155.
aIndividual constraints are nonadditive, that is, they may overlap.

Figure 1 - Effects of soil erosion on maize yields for different types of soil - Luvisol soil type

Figure 1 - Effects of soil erosion on maize yields for different types of soil - Ferralsol foil type

Source: Tengberg and Stocking 1997, figures 4 and 6.

Assessment of the Effects of Soil Degradation

An assessment of the productivity-related economic effects of soil degradation that is relevant to policy-making first requires estimates of the changes over time of the type, scale, and rate of physical soil quality at a subregional or higher scale. These changes must then be linked to consequent changes in agricultural yield or production costs, and these, in turn, to resulting changes in consumption, market supply, farm income or economic growth, and the long-term value of the resource base.

Assessing Soil Quality Change Over Time

Methods for soil quality assessment were developed mainly for use at the plot level, and are problematic to scale up, even when substantial plot-level data are available (Halverson, Smith, and Papendick 1997). No developing country has in place a national monitoring system for soil quality. Researchers trying to assess soil quality change above the plot level, have used approximate measures, including

· Consultation with experts, long familiar with particular regions, who provide a ranking or qualitative assessment of the scale and processes of degradation within the region, according to agreed-upon criteria (see, for example, Oldeman, Hakkeling, and Sombroek 1991);

· Review and comparative evaluation of published studies on degradation from many different sites within a region (see, for example, Lal 1995; Dregne and Chou 1992);

· Extrapolation of the results of case studies, field experiments, and other micro- or watershed-level data to the national level (see, for example, cases in B996); and

· Estimates constructed from examination of secondary data on land use change, representative ecological conditions, and so on (see, for example, Rozanov, Targulian, and Orlov 1990).

Assessing the Effects on Agricultural Productivity

The effects of soil degradation on agricultural productivity (see box) vary with the type of soil, crop, degradation, and initial soil conditions, and may not be linear. Lower potential production due to degradation may not show up in intensive, high-input systems until yields are approaching their ceiling. Reduced efficiency of inputs (fertilizer, water, biocides, labor) could show up in higher production costs rather than lower yields.

Effects on productivity are most commonly estimated using coefficients based on plot-level experimental trials or cross-sectional farm surveys. Many researchers estimate production effects using the Universal Soil Loss Equation.5 Since trial and survey data are unavailable for a number of soils and degradation processes, studies often base assumptions about aggregate physical yield effects on degradation-yield relationships taken from the literature or estimated by soil experts. Few studies use historical time-series data on yield and production cost; even fewer attribute yield or cost change to soil quality change, controlling for other variables.

5 The Universal Soil Loss Equation (USLE) was developed in the 1970s to estimate erosion risks and levels in temperate agriculture, but it has been adapted for tropical conditions. The USLE equation is A = R*K*L*S*C*P, where A = long-standing average annual soil erosion in metric tons/hectare; R = rainfall erosive factor (which depends on the frequency, quantity, seasonal distribution, and kinetic energy of heavy rainfall); K = soil credibility factor (dependent on soil type); L = slope-length factor; S = slope steepness factor; C = farming practice and crop-type factor (dependent on the stage of cultivation and the cover by crops, other vegetation, or residues); and P =- soil conservation measures (which depend on farm management practices). The USLE was developed and further refined for use at the farm-plot level, but it has been widely applied (and some would say, misapplied) at the landscape and even national levels to estimate erosion (Wischmeier and Smith 1978).

Most research methods provide only a rough estimate of the nature and relative importance of degradation across large areas, though a few valuable studies disaggregate by type of soil, topography, location, crop, or farm household.

Indicators of Economic Impact

Many different indicators have been used in research on the economic effects of soil degradation. Welfare effects have been measured by changes in the number of food-insecure households or malnourished children; the amount of food consumed from farm production; the level of rural household income or consumption; the degree of community- level food self-sufficiency; and the rates of migration. Effects on agricultural supply have been measured by changes in average crop yields or aggregate crop production, aggregate market supply, export or import levels, and level and variability of crop prices. Economic losses have been assessed by comparing the value of lost production, the value of inputs needed to compensate for lost nutrients, or current or discounted future income streams to farm income, national income, or economic growth rates, or by measuring changes in input efficiency. Effects on national wealth have been measured only by changes in the aggregate amount or quality of agricultural land (Scherr 1997a).

Evolution of Methods for Impact Assessment

Studies of the productivity-related economic effects of soil degradation can be divided into three periods. Those published in the late 1970s and 1980s were intended mainly to draw public attention to the issue. They used rather simplistic approaches, calculating gross aggregate effects of soil erosion on agricultural lands (assuming little use of conservation practices) and resulting gross economic losses.

Global and regional analyses published in the early 1990s were more systematically designed and reflective of broad field experience. They relied mainly on secondary data, literature reviews, and surveys of regional soil experts, and used fairly simple economic models, if any. National and sub-national studies used similar methods, but with more disaggregated data, to construct models that measured impact. Typically in the early 1990s, the economic impact of degradation was measured in terms of the value of lost yields, the value of plant nutrients lost through erosion, or the costs of soil rehabilitation. These changes were valued at market prices. The approaches of this period have been criticized for their degree of aggregation, simplistic assumptions about degradation-production relationships, failure to examine least-cost alternatives to rehabilitation, and failure to consider likely farmer or market responses to supply or cost shifts. Since the mid-1990s a third generation of studies has used more sophisticated models and methods for collecting and analyzing data to disentangle causal relationships and explore variation in soil conditions and management (see, for example, Enters 1998). Many projects have begun to collect primary data from representative soil, farm, or village units in order to develop more reliable biophysical yield models for different types of environments, degradation, and soil management. Research increasingly focuses on effects at the national and subnational levels, and this allows for more policy-relevant analysis (Scherr 1997a).

Predicting Future Effects: Conceptual Challenges

Even with the best information on past and current trends, three other central issues must be considered before predictions about future trends regarding soil degradation can be made with any confidence:

(1) To what extent is soil degradation reversible at an economically reasonable cost?;

(2) To what extent will farmers respond on their own to protect or rehabilitate their soils?; and

(3) To what extent will structural change in agricultural economies affect our reliance on currently degrading soil resources?

Reversibility of Soil Degradation

Where soil degradation is reversible at low-to-moderate economic cost (relative to agricultural product prices and land values), even significant degradation may result in little long-term economic loss. Prevention is not always cheaper than a cure. For example, farmers who cease to undertake soil-protecting investments during prolonged periods of low food prices may resume those practices when prices rise. Farmers also may mine soil nutrients (soil capital) over a period of time in order to accumulate alternative forms of more economically valuable capital, but subsequently use that capital to rebuild soil resources. Land abandonment after prolonged soil degradation could serve to keep the land fallow long enough for it to recover key long-term productive attributes.

If, on the other hand, degradation through lack of proper soil husbandry in the short term leads to permanent reductions in the soil’s productive potential, strategies leading to degradation are less likely to be economically justifiable. What constitutes “irreversibility” is a matter of some debate among soil scientists due to inadequate research. Only nutrient depletion and imbalance and surface sealing and crusting can be rapidly and relatively cheaply reversed (Table 2). Many water, nutrient, and biological problems in soils can be reversed over 5-10 years through soil-building processes and field- or farm-scale investments and management changes. Some types of physical and chemical degradation, such as terrain deformation and salinization, are extremely difficult or costly to reverse. The feasibility and cost of soil rehabilitation depend in part on soil type, production system, and severity of degradation. For many soil types, little is known about the effects of degradation or the thresholds for soil quality below which future investment in restoration is uneconomic.

Farmer Response to Soil Degradation

Historical evidence suggests that a linear extrapolation of current soil degradation trends will be a poor guide to future soil quality. Farmers depend upon the land for their livelihood. It is uncommon for them to be unaware of serious soil degradation unless they are recent immigrants to a new agroecological zone, the process of degradation has not yet affected yields, or its cause is invisible (acidification, for example). We should expect, therefore, that farmers will respond to degradation with new land management or investment if they perceive a net benefit from doing so and can acquire or develop appropriate technology. Trajectory 1 in Figure 2 illustrates such a process of innovation, in which increasing pressure on soil resources over time initially leads to soil degradation, but farmers eventually respond by improving soil management practices and making investments to restore, maintain, or even ultimately improve the soil’s productive potential. Empirical examples of such a process have been widely documented (Ruthenberg 1980; Templeton and Scherr 1997; Tiffen, Mortimore, and Gichuki 1994).

Table 2 - Relative reversibility of soil-degradation processes

Type of degradation

Degradation process

Largely reversible, low cost

Reversible, significant cost

Largely irreversible/very high cost


Clay pans, compactionzz zones


Surface sealing and crusting




Topsoil loss through wind or water erosion

X (if active deposition)


Terrain deformation (gully erosion, mass movement)



Waterholding Reduced infiltration/impeded drainage


Reduced waterholding capacity



X (farm scale)

X (landscape scale)

Chemical Organic matter loss


Nutrient depletion/leaching



Nutrient imbalance


Nutrient binding



X (if liming feasible)









Reduced biological activity due to soil disturbance


Reduced biological activity due to agrochemical use





Pollution (accumulation of toxic substances)


Source: Informal consultation with tropical soil experts and various texts on degradation.

Farmers respond not only by making major conservation investments such as terrace construction on steep slopes, land-leveling in irrigated areas, land drainage, and revegetation of denuded landscapes, but also by using alternative crop mixes and cropping intensities; land-clearing and fallow practices; spatial patterns and niches of crop production; tillage and planting density and timing practices; agro forestry practices; vegetation management outside crop fields; crop-residue management; livestock population, species, and feeding practices; or farming implements. Farmers may modify the layout of farm paths, fences, windbreaks, and other linear features or barriers in order to affect soil and water movement (Scherr et al. 1996).

The conservation community has discovered that farmers’ decisions about conservation practices and investments are inextricably linked to production (Shaxson et al. 1997). If good land-husbandry practices are to be widely adopted, they must not only replenish soil resources, but also contribute to increased productivity and farm income in the short term (Sain and Barreto 1996; Partap and Watson 1994). Farmer willingness to invest in soil improvement is closely associated with the overall economic profitability of farming and an economic and policy environment that facilitates commercialization, reduces price risks, increases access to infrastructure, increases security of land access, and encourages technical innovation (see, for example, Clay, Reardon, and Kangasniemi 1998; Shiferas and Holden 1997; Hopkins, Delgado, and Gruhn 1994).

When farmers fail to take action (trajectory 2 in Figure 2) or delay taking action until significant, irreversible degradation has taken place (trajectory 3), it usually means that they lack knowledge about effective means for soil improvement; lack access to the farm resources, such as labor, capital, or inputs, needed to make the improvements (a particular concern for the poor); believe the economic contribution of the plot to their livelihood is marginal; expect low economic returns from available options for soil improvement; or are uncertain about reaping the longer-term benefits of soil improvement due to tenure insecurity or price or climate risks (Scherr and Hazell 1994). Under these conditions, targeted policy action is needed to slow or reverse soil degradation. Policy intervention may also be desirable to accelerate farmer response in situations where social benefits are greater than farmers’ private benefits (trajectory 4 in Figure 2).

Figure 2 - Innovation in soil resource management under population or market pressure

Note: t0 to t3 are time periods. Trajectory 1 indicates a flexible and innovative response to degradation by farmers. Trajectory 2 indicates a failure to take action. Trajectory 3 indicates a delay in taking action until significant degradation has occurred. Trajectory 4 indicates that policy intervention encouraged farmers to respond sooner or more effectively than would otherwise have been expected on the basis of their existing incentives.

The trajectories of soil degradation and improvement vary considerably among different pathways of development. These variations result from differences in the soil resource base, demographic patterns, market integration, local institutions, and policy actions (Clay, Reardon, and Kangasniemi 1998; Scherr et al. 1996). Judicious use can be made of limited public investment resources to address soil degradation only if we are able to better predict when and how farmers will respond to degradation and intervention.

Structural Change in Agricultural Economy

Even if existing estimates of the economic effects of soil degradation in recent decades are correct, they cannot necessarily be extrapolated to 2020. There is no certainty that all of the developing world’s soils currently under cultivation will constitute important resources for agricultural production in the decades ahead. Structural changes in global and national economies, trading patterns, and infrastructure development may make some soil resources much more important than others. Technological breakthroughs may make some “problem” soils much more productive in the future, while unforeseen events may contaminate soils that are most productive at present. Thus, evaluation of future threats of degradation requires that we assess the likely future trends in the broader economy and their implications for soil management. Some possible scenarios are presented in Chapter 4. Past and present challenges are presented first, in Chapter 3.