Contents - Previous - Next

Hazards

Rainfall was 100 to 250 mm less than the average. Only one set of major rainstorms was recorded, totalling 130 mm in three days, and falling in the summer on dry soil.

TABLE 48
Runoff (% of rainfall), erosion (t/ha/yr), yields (t/ha) and net revenue (in diners [28 DA = $US 1]) for 15 erosion plots at the INRF station at Ouzera, Algeria (cf. Arabi and Roose 1992)

Med. = median; Max. = maximum in 1990
* The apricot harvest was very poor because of an attack by insects.

Rainfall is much less forceful than in West Africa. The annual average rainfall ratio is about 0. l at the Médéa station, 0.5 in Côte d'Ivoire and 0.25 in the mountains of Cameroon, Rwanda and Burundi.

Average annual runoff (Kaar %) (Table 48) is modest (0.5 to 4% of rainfall), and the maximum during a rainstorm is 8 to 36% 140% in exceptional cases). On the bare control fallow, annual runoff was still modest (10 to 18%) compared with levels in a tropical environment (25 to 40% in Côte d'Ivoire). However, if the bare soil is packed down or waterlogged during the winter, runoff can exceed 50 to 80% on marl and red fersiallitic soil.

With similar soil conditions, slopes and tillage techniques, it was noted that plant cover especially with improved techniques - effectively reduced erosion risks.

As many other writers have noted, tillage improved infiltration. For example, under the grapevines on the plot where tillage was replaced by herbicides in order to control weeds, runoff increased significantly because the surface horizon was compacted, but at the same time erosion decreased because the soil was more cohesive. In the case of an exceptionally heavy rainstorm, the water storage capacity of the soil would soon be exhausted, and runoff could increase to such an extent on the tilled plot that it would carry away the tilled - and therefore less cohesive - horizon, at least on the steepest slopes. This is frequent in the fields.

Under natural vegetation (scrubland developing into grassland or forest), very considerable litter cover (more than 80% of the surface covered) meant that although runoff was frequent on soil compacted by overgrazing it was never dangerous (max KR £ 7%). However, it has often been observed in Algeria that drainage lines and gullies arise on overgrazed rangeland (especially on tracks worn by livestock) or even in some forest plantations degraded by grazing.

Runoff generally begins after 20 mm of rain on dry ground, and after 3 mm on compacted or moist ground. The precise point at which rain gives rise to runoff clearly depends on the specific features of each rainstorm (intensity, but also capacity to saturate the soil), but first and foremost on the condition of the soil surface (humidity of the top 10 cm, presence of fissures, worm holes, slaking crusts, litter, pebbles and large clods). The most copious runoff occurs only when all the conditions are optimal - usually between November and March - or during an exceptional summer storm (every five years).

Sheet erosion was a very modest 0.1 to 2 t/ha/yr on cultivated fields and 1.5 to 18 t/ha/yr on bare fallow despite steep slopes (12 to 40%), for rainfall is fairly gentle (R_USA < 50) and the soils are very resistant (K = 0.01 to 0.01), rich in calcium-saturated clay, and often stony.

Even if erosion were to reach 19 t/ha/yr (1.27 mm of soil), it would take two and a half centuries to remove 20 cm of topsoil. It has been proved in trials that sheet erosion selectively removes organic matter, clayey and loamy colloids and nutrients, while rill erosion removes soil unselectively. Thus, when rill erosion sets in, it usually removes the topsoil, especially on steep slopes. If sheet erosion appears not to be the most powerful process on slopes, rill erosion and especially creep caused by farm implements appear to be the forms most active in transforming mountain landscapes.

For example, at Ouzera on an orchard planted 30 years ago, 30 cm of soil is now missing between the tree trunks. Even if recorded sheet erosion is as much as 1.5 t/ha/yr (0.1 mm), in 30 years 3 cm of soil would have been lost, while 27 cm must have been moved by dry mechanical erosion during the twice-yearly deep cries-cross tillage with a crawler tractor. Tillage therefore hastens the transformation of mountain slopes, contributing to the formation of banks on the edges of fields.

Soil erodibility was seen to be very low, even after three years of untilled fallow (K = 0.01 to 0.02). However, sheet and rill erosion are increasing each year.

Erosion was high on red fersiallitic soils (10 to 19 t/ha/yr), average on grey vertisols (2 to 3 t/ha/yr) and minimal on brown calcareous soils (1.5 to 2 t/ha/yr). It seems that gravel offers very effective protection.

TABLE 49
Effect of soil type and slope (%) on runoff 1%) and erosion (t/ha/yr) on bare fallow (cf. Arabi and Roose 1992)

TABLE 50
Effect of improved farming practices on runoff (average and maximum as % of rainfall), erosion (t/ha/yr) and net income ($US 1 = 28 diners)

It is difficult to comment on the likelihood of runoff for different soils, since the slopes vary as much as the soil types. However, it does seem clear that average annual runoff and maximum daily runoff decrease as the gradient increases, at least on bare fallow (Table 49) - an astonishing conclusion that was previously reached by Heusch in Morocco (1970) and Roose in Côte d'Ivoire (1973).

This result throws doubt on the validity of the equations of Ramser, Saccardy and others, for whom the space between terraces should decrease as the gradient increases. Heusch has already shown in Morocco that the position of a field in the toposequence can be more important than the gradient of the slope itself.

Suggested improvements: influence of the farming system

[Plate 23]

Better plant cover (high crop density, fertilizers, rotations with pulses, a winter crop between fruit trees and grapevines) steadily but undramatically reduced runoff and field erosion (Table 50).

More important, however, is the very marked improvement in crop yields and farmers' income (see Table 48). The traditional cereal crop can bring in 2500 dinars/ha/yr, while the improved, intensive cereal/legume rotation brings in 35800, and up to 42000 or 65000 diners when this rotation is introduced between rows of grapevines or fruit trees.

These results show that it is possible both to intensify mountain farming and to reduce the risk of degradation of the rural environment.

The yields recorded on the traditionally worked runoff plots (100 m²) were as poor as those on nearby farmers' fields (700 kg/ha/yr of winter wheat, 2800 kg of grapes, 800 kg of apricots, these latter being diseased). On the neighbouring erosion plots treated with improved techniques, yields reached 4800 to 6500 kg/ha/yr of winter wheat, 4000 kg of grapes, plus 3400 kg of dried beans.

Furthermore, straw, pulsecrop leaves and other crop residues also increased significantly (from 0.2 to 2 or 3 t/ha/yr) so that livestock production and the availability of manure or organic residues can in the long run improve soil fertility and resistance to erosion.

Yield improvements are unlikely to be as spectacular on large areas as on the small erosion plots (100 m²), but the first step has been taken: that of showing that farming can be intensified while steadily improving the rural environment.

That this is a viable investment is shown by the following. After subtracting the extra costs of improved seed, fertilizers, pesticides, herbicides, the work of soil preparation and harvesting, farmers are still left with a much higher net income than that from traditional crops.

1) extensive grazing in wooded areas can bring in 500 DA/ha¹
2) traditional winter wheat 2500 DA/ha
3) grapevines or traditional apricot orchards 10-17000 DA/ha
4) improved rotation: wheat × fodder pulse 28-33000 DA/ha
5) combination of this rotation with grapes and apricots 42-65000 DA/ha

¹ 5 Algerian diners (DA) = 1 FF ~ $US 0.2 at 1992 exchange rates.

This would indicate that returns can be multiplied tenfold for cereals and threefold for grapes if an intensive system is adopted. If the traditional wheat-grassland rotation is replaced by intensified orchards, returns are multiplied twentyfold. This is one of the benefits of mountains with a mild, moist climate, for farming cannot always be intensified if the soil is too shallow and rainfall is less than 400 mm.

With such evidence in hand, it is possible to interest farmers in changing and improving their practices so as to increase conservation of water and soil fertility. Indeed, after three years of trials, the hill farmers copied the project methods and achieved better yields than the project itself in 1991, a year when rainfall was well distributed.

Contents - Previous - Next