Contents - Previous - Next

Chapter 3. Some social and economic aspects of erosion

Erosion crisis diversity
Who is concerned by erosion control?
The importance of exceptional rainstorms
Erosion effects in different regions
Effects of erosion on the eroded site: loss of productivity
Negative off-site effects of erosion
The economic rationale for land husbandry
Criteria for the success of soil conservation projects
Morroco case study: socio-economic study of erosion control in the Loukkos Basin

Erosion is not simply a technical problem, and if erosion control has enjoyed only a qualified success so far, the reason lies not only in the failure to solve certain technical aspects of the problem to full satisfaction, but also in the need to pay more attention to the social and economic roots of erosion crises.

In the name of the public good, civil engineers have tried to impose their own solutions without bothering too much about the specific interests of each of the "beneficiaries" of engineering schemes. Here an attempt is made to enter into dialogue with the people who actually work the land, and to define their reaction in terms of their immediate concerns.

An analysis is then made of the extent of erosion problems throughout the world and the special importance of exceptional rainstorms.

After this, known facts about the cost of erosion are schematized; on the one hand, the immediate on-site effects of erosion and runoff on production, nutrient losses and the long-term productivity of degraded soil at the plot level; on the other hand, the off-site problems and damage caused by runoff when it swells peak floods, reactivates river gouging and riverbank degradation, pollutes water through nutrients and suspended matter, or silts up dams and reduces the quality of the water indispensable for the development of towns and intensive farming.

Lastly, an attempt is made to orient the choice of an erosion control strategy on the basis of the economic objectives of such engineering projects, and define the conditions for their success. A brief case study of an erosion control scheme in Morocco is also given.

Erosion crisis diversity

(comm. from C. Lilin)

Moving from one age or one country to another, major categories of erosion crises can clearly be defined - even though closer analysis will reveal the unique aspects of any given situation. If this diversity is under-estimated, a chosen approach may turn out to be unsuitable.


Population growth is reflected in a growing pressure on the natural resources of a given area, which in turn leads to their over-exploitation and degradation. The history of many countries is marked by such erosion crises, which can be classified either according to the causes of degradation or according to the response to the problem.

Factors involved in degradation. In many cases, the effects of the population/resource imbalance are aggravated by other processes. Thus in many developing countries, the presence of wide social differences in rural societies can be noted. Already disadvantaged social groups are sidelined, with their access to land resources reduced, and they are pushed onto marginal land (often also the most fragile). These disadvantaged groups also often have poor land tenure status. Now, land insecurity and conditions such as sharecropping or undivided ownership (land held in common) are disincentives to investment, inasmuch as the risks are too great or the possible benefits for the actual workers of the land too small.

Other processes also help create difficult conditions for the investment inherent in erosion control measures: scarce financial resources, for example. Marginalization also leads to the adoption of survival approaches, which favour the very short term. And these may persist even when conditions come to favour the disadvantaged sectors of the farming population.

In many countries the mechanical effects of population growth are aggravated by social processes that create what might be termed spiralling degradation. In such cases, erosion constitutes one indissociable aspect of under-development.

Working out a response. An erosion crisis can be compared to a disease that attacks a body, in this case a rural society. Historians studying past crises have shown how local societies have reacted in an effort to control erosion. For example, Blanchemanche has analysed the response to the 17th- and 18th-century erosion crises in the pre-alpine French Mediterranean region. Rural societies met the challenge by stepping up farm production through techniques such as terracing and irrigation, taking ideas from more technically advanced regions such as Tuscany. Local Úlites played a key role in the search for techniques (technology transfer), their adaptation to local conditions and their dissemination.

In many developing countries, however, local Úlites are not capable of taking responsibility for the erosion problem. Rural societies are often in crisis and traditional structures have lost their authority, while the modern structures in place are incapable of playing their role effectively. Such societies do not have the necessary local structures to meet the various challenges of under-development, and there is a great temptation for a social actor such as the State to take the place of weak local structures, setting up projects to provide the technical elements of a response to the erosion problem.

Such a strategy might have been effective in the very specific context of erosion in mountain areas in 19th-century France, just as a strategy of State intervention favouring the technical aspects of the problem may also be valid in treating the present erosion crisis in certain large-scale farming and wine-growing hillside areas in France, inasmuch as the rural societies in question have efficient local structures (e.g. at the level of the commune, the county council, the modern agricultural sector and the department-level agricultural authorities). The main role of the State is then that of encouraging the production and spread of suitable technology so as to help speed up implementation of measures ensuring erosion control.

On the other hand, where an erosion crisis affects rural societies that are themselves in crisis, erosion control to a large extent means reinforcement of the authority of local structures. It is important not to neglect the institutional aspect of the problem in such cases.


Erosion can be seen as a result of over-hasty modernization in certain large-scale farming or wine-growing areas of Western Europe. The last few decades have seen a series of changes, which are reflected in substantial increases both in productivity per hectare and, even more strikingly, in the productivity of the individual farm worker. These myriad and rapid changes (greater specialization of production systems, mechanization and motorization, increase in plot size, and elimination of structures instrumental in rural hydrology, etc.), have had major effects on the soil, and have led to an erosion crisis in some regions.

In tropical countries, a similar situation is observed where traditional farming has been replaced by modern, mechanized monocropping.

Similarly, in older times, changes in cropping practices or the introduction of some new crop - for example, the move to obligatory three-year rotation, or the introduction of potato-farming - sometimes led to the development of erosion.

Where erosion can be analysed as a consequence of agricultural modernization, priority must be given to the technical aspects of the question. Once the necessary technology has been worked out and tested, implementation will be all the easier inasmuch as the most forward-looking farmers are in general those most concerned by this problem, which is bound up with the introduction of new technology.


If the diversity of erosion crises is taken into greater account, it will be easier to adapt strategies to specific situations, with major variations in the emphasis on the design and dissemination of technology, treatment of the problem of under-development as a whole, and the institutional aspects.

A major difficulty springs from the fact that in developing countries the simultaneous treatment of these different aspects is vital, often constituting the key to the effectiveness of any action. Now, it is also in such difficult contexts that the production of appropriate technology can leave much to be desired, coordination of the activities of different administrative offices is problematic, and the sheer weight of projects enjoying foreign aid acts against their continuity.

Who is concerned by erosion control?

In general, big landowners ( > 500 ha) are usually little concerned by erosion, for they can easily abandon degraded land as wasteland.

In France, erosion problems are relatively rare on smallholdings (livestock production or multi-cropping mixed with livestock), for the small farm units are often well manured by animals raised under the zero grazing system. The most enterprising farmers are in fact the ones with serious erosion problems, for they contract debts to purchase large tractors and other equipment to prepare seed beds in the most advanced manner, as well as heavy-duty trailers to transport harvests. They have accepted land consolidation in order to make their farms as cost-effective as possible by eliminating all obstacles (ditches, hedges, copses) that can impede the advance of machinery. It is the people downstream who actually raise the question of erosion control, when they suffer the ill-effects of peak runoff flows, pollution of groundwater and rivers, gullying, and mud flows in inhabited areas. The big landowners should take an interest in no-till techniques, especially since the new EC agricultural policy calls for downsizing production, set-aside for particularly fragile land, and extensive livestock production on grasslands (Seguy et al. 1989; De Ployey 1990).

In developing countries, large numbers of poor small farmers are hard pressed to assure the survival of their large families (5 to 10 members in Rwanda) on tiny farms (0.2 to 1.5 ha). Despite falling yields, they cannot allow the depleted soil to rest, so that it is often barely covered (especially in semi-arid areas), fragile, located on steep slopes, and ill-protected from runoff from neighbouring plots and roads. Some families put off investing in land management until erosion damage is so serious that they have no choice. Other families in similar conditions simply up and leave, abandoning everything to try their luck in towns, or else they send some adult members to neighbouring countries to bring in a little extra income. Farmers' interest in land management depends greatly on the land tenure system. If they are the actual owners, they will invest their time (often the only input available) to mark the boundaries of their plots (hedges, low walls, lines of stones) and improve the land (organic dressing, liming, progressive or radical terracing, trenching to break up calcareous crusting, clearing the land of stones, agroforestry). It is relatively easy to introduce agroforestry or intensive cropping under orchard trees, but in cases of sharecropping or tenancy, farmers cannot improve the land they work for fear of being accused of trying to appropriate the land and therefore having their permission to work the land withdrawn, or of suffering rent rises on the basis of improvements.

In Haiti, there are three kinds of land. An "A plot" will hold the owner's house, and is a multi-storey garden, encompassing fruit trees, forage for tethered small livestock, a vegetable garden, and pigs, all protected from pilfering and very well kept. "B plots" are further away, less well guarded, less intensively farmed, and less well protected against erosion. Lastly, most farmers rent some more distant land - "C plots" - which are unfenced, very little developed, and often with trees and soil in an advanced state of degradation. A recent survey showed that all farmers give priority to managing their A plots, which are the best protected - even if this means putting off management of their most degraded land which is where SWC specialists have been unsuccessfully focusing for half a century, trying out every known method of ditching and terracing (Naegel 1991).

The importance of exceptional rainstorms

When the press mentions erosion, it is usually talking about natural disasters which have led to exceptional damage and the loss of human life in the space of a few days - or even hours. People are very often not directly responsible for such disasters, which are caused by natural forces beyond our control, for example, volcanic eruptions, earthquakes, or torrential rain falling on frozen soil. However, human beings can aggravate such damage through ill-advised development. Forgetting the wisdom of their forebears, they have built structural works or homes in the path of avalanches or mud flows, or close to geological faults (San Francisco is an example here), in main river-channels or any other area subject to occasional flooding, thus increasing the catastrophic effects of such exceptional events.

The recent floods at Nţmes in southern France are a good example (Davy 1989). On 3 October 1988, a violent storm unleashed 420 mm of rain in 6 hours on two small Mediterranean catchment areas that dominate the town. The torrents and springs flowing from the limestone hills swelled inordinately and swept violently through the old town, carrying everything with them: vehicles, the contents of shops, etc. There are channels capable of evacuating such huge quantities of water, and these were respected by the Romans in ancient times. In recent years, however, they have been blocked by buildings, Highway 113 (which is designed for flooding and is not a problem), the embankment carrying railway lines, 20 metres of which were swept away (the drains being blocked by the wrecked vehicles) and lastly the motorway, which is slightly elevated. A vast area was therefore flooded, with 4 thousand million FF of damage and 11 fatalities.

The question is whether most erosion damage is a result of such very widely reported but rare disasters, which are very difficult to prevent, or is rather caused by the aggregate energy of rain falling on cultivated soil which could be better protected. A detailed study of erosion damage in the wine-growing areas of Alsace (Schwing 1979) showed that the annual cost of retrieving eroded soil from downslope and loss of inputs following normal storms was about 2000 FF/ha/yr, whereas the additional damage caused by exceptional events amounted to 15000 FF/ha/every 25 years, plus local community expenses.

While it is well known that exceptional rains generally produce major damage, the extent of such damage varies in different environments. In temperate areas, according to Wischmeier, the sum of the rain erosivity of all significant showers (over 12.5 mm) is what decides the annual erosion level at the catchment level. In subequatorial areas (e.g. C˘te d'Ivoire) the situation seems similar (Roose 1973), whereas in areas subject to frequent cyclones (e.g. New Caledonia, the West Indies, Reunion), cloudbursts are so heavy (500 mm in a few hours) that they deeply mark the landscape (regressive gullies, broad river-beds and large numbers of alluvial terraces). Similarly, in semi-arid, Sahel-Saharan or Mediterranean areas nothing may happen for years, and then in the space of a few hours, an exceptional storm, or series of storms, savagely reshapes the landscape for years or even centuries to come, with deep gullies, landslides, the undermining of wadi banks, and large-scale sedimentation in flooded plains (e.g. the events in Tunisia in 1969; Claude, Francillon and Loyer 1970). This means that it is not always easy to distinguish active gullies from forms that are a legacy of the past, and there is sometimes no direct link between forms of erosion and surrounding land use.

Another major economic question is whether erosion control measures are as effective for exceptional storms as for ordinary rainfall. Hydrologists generally agree (the gradex methodology) that after a certain amount of rain has fallen - whether exceptionally heavy or exceptionally long - the runoff from a catchment area tends towards 100%. This peak is attained for highly variable recurrent storms depending on the kind of rain, the status of the soil and plant cover, and how the whole catchment area has been managed. In such exceptional events, there is hugely swollen streamflow in the outlet channels and impressive sediment loads from the bed, banks and low terraces. However, at the watershed level, the more intelligently planned the erosion control measures are (terraces protected by hedges, grass banks, well-structured soil under a mulch or thick plant cover, etc.), the less danger there is of damage during such exceptional storms. Moreover, torrent-control dams are designed to withstand the effects of such cloudbursts (comm. from Mura 1992).

Runoff in the Manakazo watershed (Madagascar)

Effect of plant cover and erosion control techniques (cf. Goujon, 1972)

Plant cover

Kaar %

Max KR %

Max. flood flows l/s/ha

freq. 1/1



Burnt steppe:

- years burnt


50 - 70




- other years


40 - 50




Steppe off-limits to livestock


40 - 50






> 20




Pinus patula forest:

- 0-5 yrs


15 - 38




- > 10 yrs


1 - 5




Kaar: Annual average runoff coefficient
Max KR: Maximum runoff coefficient

The problem was considered in connection with watershed management in the Tananarive basin in the Malagasy uplands. Since this basin drains five rivers and has only one small outlet, which is partially blocked by a rocky bar, it is regularly flooded by cyclones from the Indian Ocean. These floods are all the more damaging in that they destroy rice harvests and can sometimes drive over one hundred thousand people to flee their homes (Roose 1982).

Three solutions have been examined. The first is that of broadening and deepening the outlet by blowing up the rocky bar, but this entails the risk that regressive erosion could destroy the rice fields that provide food for the capital. Secondly, part of the catchment area could be eliminated, and flood crests could be checked by building dams to store runoff from the heaviest downpours; this is a very neat solution, but costly in foreign currency. Lastly, the hill areas could be managed and afforested, erosion control structures reinforced (terraces with grass embankments) and cropping techniques improved; this solution would take some years, but it is within the financial reach of a poor country with strong governmental presence.

The only available trial results (four 4-ha catchment areas at Manakazo on the Malagasy high plateaux; Table 2) show that peak discharges from the regional control catchment (burnt savannah with Loudetia stipo´des) are ten times greater than under young pine forests (Pinus patula) and four times greater than on farmed catchments with progressive terraces (Goujon 1972). For the rare storm, peak discharges do in fact tend to blur the picture, but only for storms occurring once every five hundred years under forest and once every hundred years on farm land. Despite its long-term effectiveness, the method has not been developed on a large scale, since it takes too long (over 10 years) for forests to be effective against runoff.

Erosion effects in different regions

The severity of erosion will vary considerably from place to place.

At the 1982 New Delhi International Congress of Soil Science, Kanwar (1982) showed that of the world's 13500 million ha of land not under water, 22% is suitable for cropping but only 10%) is currently farmed. Losses in arable land have increased over the past ten years to a current rate of 7 to 10 million ha/yr as a result of erosion, salinization or urbanization. At this rate, it would take three centuries to destroy all arable land. Erosion is hence a serious world problem, although it is particularly worrying in certain regions.

Around 1930 in the United States 20% of arable land was seriously damaged by erosion as a result of the earlier and ill-considered decision by European settlers unaccustomed to such semiarid conditions to plough and farm the Great Plains. This was the grim "dust bowl" era, when dust clouds darkened the sky at noon. Such a public outcry was raised that the American government decided to set up a full-scale soil and water conservation service, offering farmers technical and financial support in each district to volunteer to join the programme. At the same time a network of research stations was set up, which 30 years later led to the formulation of the Universal Soil Loss Equation, or USLE (Wischmeier and Smith 1958; 1978). In 1986, Lovejoy and Napier observed that after 50 years of massive investment in human resources and funding, 25% of agricultural land were still losing over 12 t/ha/yr, the recognized tolerance limit. So the issue remains topical, although water pollution and water quality attract more attention today than soil conservation.

In France, a survey by Henin and Gobillot (1950) established that an estimated 4 million ha of farmland had been degraded by water or wind erosion. Since the danger was considered limited, little research funding was available in this area. Thus France still has no proper erosion control technology, which causes considerable problems in the case of impact studies.

Taking the European Community as a whole, De Ploey (1990) estimates that 25 million ha have been seriously affected by erosion. France is thought to account for 5 million of this total, and the cost of erosion-caused damage is put at 10 thousand million FF, excluding the intrinsic value of the lost soil, which is hard to quantify.

Much more dramatic figures in tropical countries were cause for alarm. In 1977 Combeau reported that 80% of the land in Madagascar was suffering accelerated erosion. Also 45% of the surface area of Algeria is affected by erosion, which translates into 100 ha of arable land lost for every day of rain.

In Tunisia, Hamza (1992) estimated the average annual sediment load transported by the different watersheds. Assuming an average soil depth of 50 cm, the equivalent of 15000 ha of land are washed into the sea by water erosion each year!

More serious than these dramatic extrapolations are the soil losses recorded on 100 m▓ plots established since 1950 by ORSTOM and the CIRAD institutes (Roose 1967, 1973, 1980a), under Professor Frederic Fournier's influence. Soil losses range from 1 to 200 t/ha/yr (and up to 700 tonnes in mountain areas on 30 to 60% slopes) under crops adapted to average forest slopes (4 to 25%), with losses of 0.5 to 40 t/ha under millet, sorghum, groundnut and cotton on the long tropical ferruginous pediments of the Sudano-Sahelian regions (Roose and Piot 1984; Boli, Bep and Roose 1991).

If an apparent surface horizon density of 1.2 to 1.5 is assumed, the amounts removed by erosion range from 0.1 to 7 mm (and even 15 mm in mountain areas) depending on topography, climate and crop. This corresponds to 1 to 70 cm (150) cm/century or 0.2 to 14 m in the past two thousand years, though the same soil has obviously not been cropped for two thousand years! Land exhausted after 2 to 15 years of relatively intense and unbalanced cropping (removal and losses not being made up for by replacements and supplements) was left fallow, which has the primary effect of reducing erosion (Roose 1992b).

The length of soil life can also be estimated on the basis of mean annual soil loss, the depth of soil to which roots can reach, the rate of soil fertility regeneration, and the soil yield curve as a function of the depth of the arable layer (Elwell and Stocking 1984). In a forest environment, with aggressive rainfall and steep slopes, soil losses can be considerable, and degradation very fast (a few years). However, soil regeneration is equally fast, for degraded soil provided it is quickly covered by vegetation.

In semi-arid areas, the life span can be several decades, despite slight slopes and aggressive rainfall, but the restoration of soil fertility is slower in that biomass production is poor in low-rainfall areas and the soil is greatly depleted.

Analysis of the sediment load of hundreds of American and European rivers shows that there is a semi-arid climatic zone (mean annual rainfall 350 - 700 mm, depending on how continental the watershed is), where specific degradation of watersheds is greatest. In lower-rainfall zones, the specific sediment decreases with rain energy (Fournier 1955). In higher-rainfall areas, the plant cover intercepts a good part of the energy of rain and runoff energy (Fournier 1955 and 1960). Although this is statistically true at the macrofocus of an entire watershed, it is not so at village level, and even less so on the plot level. The specific management system used for each plot leads to major local variations - a valid reason for developing cropping techniques encompassing erosion control.

The economic impact of erosion can be analysed from two perspectives:

• the on-site perspective of plots on which the signs of runoff and erosion have developed;
• the off-site perspective of damage further downstream.

Contents - Previous - Next