PROCESSES AND CAUSES OF ACCELERATED SOIL EROSION ON CULTIVATED FIELDS OF SOUTH WELO, ETHIOPIA

Belay Tegene^*

Abstract: The processes and causes of accelerated erosion on cultivated fields in the South Welo zone of Ethiopia were assessed on the basis of information collected from field surveys, group discussions and secondary sources. The findings suggest that soil erosion by water on cultivated slopes in the zone is currently proceeding at an average rate of 35 t/ha/year. Comparable if not more intensive soil loss is also taking place due to tillage erosion on the cultivated slopes although no data is available to support this. Three very broad and inter-linked groups of factors were identified as causes for both the water and tillage induced erosion in the zone. The first group comprises the biophysical factors, i.e., the rainfall erosivity, the soil erodibility and the gradient and length of slopes, which raise the vulnerability of the arable land to accelerated erosion. The second group and the real causes of erosion were identified as the cropping and land management factors. In most of the highlands, crop cultivation is carried out without any type of terracing, while about 74 per cent of this land requires application of contour plowing, broad-based terracing, or bench terracing. The third group of factors include the social and institutional factors which not only compel farmers to cultivate fragile environments but also exert strong influence on the type of land management practices the farmers apply on vulnerable soils. The major ones among this group were rapid population growth (and shortage of land), widespread poverty, and insecure land tenure system.

1. INTRODUCTION

The Welo highlands are currently undergoing considerable sheet, rill and gully erosion. Because of accelerated erosion the soils on most of the slopes have already been reduced to depths of less than 30 cm (Hurni 1993). It must be noted, however, that although overgrazing and excessive trampling along a few pathways encourage erosion by water in the non-arable land of the South Welo highlands, the most intense runoff induced erosion is currently limited to arable land. Erosion resulting from tillage operations also contributes to pronounced soil depth loss in cultivated fields. The combined effects of these two major processes on the farmland are causing accelerated erosion in which the soil loss rate very much exceeds that of soil formation. It is to be noted that for a centimetre of soil to form it may take 200 to 1000 years on a rock under natural condition (Ehrlich et al. 1977) and at least 200 years for a topsoil under cultivation (Pimental 1993). Accelerated soil erosion constitutes one of the greatest threats to agriculture in the South Welo highlands for it is causing an irreversible loss of the soil resource on cultivated slopes.

2. METHODS AND PROCEDURES OF STUDY

The processes and causes of accelerated erosion on cultivated fields in the South Welo highlands were assessed on the basis of information collected from primary and secondary sources. The primary data was generated through field surveys carried out in four sub-catchments (Derekolli, Gido, Nibo and Wurgo catchments) having sizes ranging from about 1100 to 1700 ha and located in different agro-climatic zones. Firstly, a systematic semi-detailed (1:12500 scale) grid survey of soils, and description of site and land use characteristics was carried out in each one of these sub-catchments. At each point, selected at grid intervals of 250 meters, auger hole observations were made and site characteristics, such as landforms, slope gradients; vegetation types, land use, etc., were recorded, and soil attributes such as depth, colour, texture, structure, erosion features, etc., described. Secondly, about 60 farmers from each one of the sub-catchments were interviewed using structured questionnaires consisting of both closed and open-ended questions. The sample farmers were randomly selected from among the owners of cultivated fields picked on the basis of grid points laid for the purpose of the soil survey. Thirdly, discussions were held with six groups of farmers selected from different villages in each sub-catchment. A total of about 40 farmers from each sub-catchment were involved in the focus group discussion. In the course of the verbal interactions questions were raised by the author only to guide and direct the responses to the main themes of the study, or else the discussions were led by the farmers themselves. Thus, farmers were allowed to comment freely on their knowledge of soils, crops, soil conservation practices, etc.

3. DESCRIPTION OF THE STUDY AREA

According to the Department of Planning and Economic Development of South Welo Zone, the highlands, constituting the land above 1500 meters above sea level (m.a.s.l.), cover about 16,675 km² or 80 per cent of the total area of the zone (DPED 1993). This mountainous landscape is marked by highly variable relief, dissected terrain and rugged topography (figure 1). The altitude rises from 1500 m to above 3600 m.a.s.l. with subsequent decrease in temperature and increase in rainfall. In fact, the co-variability of temperature, rainfall and altitude has brought about the vertical differentiation of the zone into distinct agro-climatic belts traditionally recognised as kolla, woina dega, dega and wurch (table 1 and 2).

Figure 1. Location map of the study area

Table 1: Climatic Belts of South Welo Zone

Agro-climatic zones	Latitudinal range (m)	Mean annual temperature (°C)	Comparable global climatic belts
Wurch	> 3500	< 10	Alpine
Dega	2400-3500	10-15	Temperate
Woina dega	1800-2400	15-20	Subtropical
Kolla	1500-1800	> 20	Hot subtropical
Source: Adapted from Amare Getahun, 1984

Table 2. Area and Attitudinal Zones of Sub-catchments Selected for the Study

Characteristics		Sub-catchments
	Derekolli	Gido	Nibo	Wurgo
Absolute alt. range (m)	1575-2060	1920-2625	2410-2940	2760-3600
Altitude of major part (m)	1600-1800	1920-2280	2480-2720	2800-3200
Approximate area (ha)	1500	1300	1100	1700
Climatic Zone	Kolla	Woina dega	Lower Dega	Upper
				Dega

The climatic conditions of the sub-catchments were inferred from secondary sources, namely, meteorological records made in the towns of Bati, Hayk and Boru Meda. Based on this data, the annual rainfall in Derekolli, Gido, and Nibo was estimated at about 850, 1150 and 1235 mm, respectively. The precipitation at Wurgo should be slightly higher than that at Nibo because of the former's location at a higher altitude. Rainfall is in all cases bimodal with the little rains, locally known as belg, from March to May; and the big rains, meher, July to October. The mean annual temperature varies from about 20°C in Derekolli, to about 12°C in Nibo, and probably less than 8°C in the Wurgo catchment.

Subsistence agriculture, involving mixed farming of crops and livestock, forms the basis of the economy of the population inhabiting the South Welo highlands. An estimate made for this study based on a highly generalised map suggests that about two-thirds of the highland is now under cultivation and that nearly all the arable land is already converted to farmland. A variety of crops are produced on the cultivated land (table 3) but crop yield per hectare is very low (4 qt. for lentils to 10 qt. for sorghum). The arable land holding of most households is also less than one hectare -- the average being 0.63 ha. On the other hand, the household size is large (5.63 persons) and the minimum farm-size required to sustain such a unit is estimated at 1.75 ha (Yeraswork Admassie 1995: DPED 1993; Mesfin WoldeMariam 1991).

Livestock is kept not only to provide milk, hides and skins, and drought and pack animals, but also to serve as a means of security against clop failure. However, livestock raising is becoming more and more difficult because of the expansion of cultivation on to grazing land and subsequent contraction of pasture. Currently, livestock depends on grazing and browsing land for 70 per cent of its forage. It gets its remaining feed from cultivated fields. In fact about 60 per cent of the feed during the dry seasons is estimated to come from cultivated fields, mainly in the form of cereal straws (DPED 1993).

Table 3. Major Cultivated Crops of Southern Welo Zone in Order of Importance
Type of crop	Scientific name	Local name
Sorghum	Sorghum bicolor	Mashila
Barley	Hordeum vulgare	Gebs
Teff	Eragrostis teff	Teff'
Wheat	Triticum vulgare	Sinde
Maize	Zea mays	Bekolo
Horse beans	Vica faba	Bakela
Field peas	Pisum sativum	Ater
Chick peas	Cicer arietinum	Shimbra
Lentils	Lens culinaris	Misir
Emmer wheat	Avena sativa	Aja
Haricot beans	Phaseolus vulgaris	Fosolia
Vetch	Vicia sativa	Guaya
Finger millet	Eleusine coracana	Dagusa
Sesame	Sesamum indicum	Selit
Niger seed	Guizotia abyssinica	Nug
Flax (linseed)	Linum usitatissimum	Telba
Fenugreek	Trigonella foenum grecum	Abish

Source: Department of Planning and Economic Development of South Welo (DPED 1993)

The rapid population growth (3 per cent/year) and the severe shortage of land continue to exert high demographic pressure and severe stress on land resources and the environment. As a result, the natural forest and woodlands are now completely removed, and currently the highlands are devoid of their natural vegetation. Estimates indicate that the area currently covered by natural forests is only 11,368 ha or 0.5 per cent of the zone (DPED 1993). The area put under plantation forests in the last two decades was not more than 10,000 ha and, as Mesfin Tadesse (1990) rightly argues, this forest has not yet significantly contributed to the vegetation cover of the highlands. Moreover, because the plantation forests were imposed on the community, a considerable portion of them were destroyed by the peasant farmers following the fall of the `Derg' regime (DPED 1993).

4. PROCESSES OF SOIL EROSION IN THE SOUTH WELO HIGHLANDS

Estimates made on the basis of an eight-year (1982-1989) period of data collection by the Soil Conservation Research Project (SCRP 1996) suggest that currently erosion by water on cultivated slopes of the South Welo zone is taking place at an average rate of about 35 t/ha/year (table 4). However, it must be noted that this average figure masks the wide diversity of erosion in space and time. The annual erosion rates, in the eight year period, varied from 8 to 97 tons/ha. Erosion rates also vary with space depending on the slope gradient, type of crop, tillage intensity, etc.

It must also be noted that normally only a portion of the soil removed from cultivated fields leaves the catchments. In the case of Maybar, only 13 t/ha (about 40 per cent) of the soil removed from the cultivated plots is estimated to leave the experiment catchment (116 ha) in the form of suspended sediment. This suggests that most of the soil removed from cultivated slopes is deposited within the catchment sometimes causing serious damage to crops. That there is considerable deposition of sediments in the South Welo highlands is also evident from the rapid siltation of artificial ponds constructed in the last 25 years (DPED 1993).

In the wetter highlands, i.e., the 'dega' and 'woina dega' zones, erosion by water is also to some extent aided by mass wasting (soil creep, landslides, etc.). The latter carries considerable amounts of soils and other materials under the influence of gravity, with water playing only a subsidiary role. The incidence of mass wasting has significantly increased in the wetter parts of the South Welo zone since the introduction of level bonds.

Table 4. Soil Loss (t/ha) Registered on Cultivated Soils at Maybar Soil Conservation Research Station, South Welo Zone (1982-1989)

year	Rainfall (mm)	Erosivity (Rm; J/mh)	Plot I (16 % slope)	Plot II (37% slope)	Average
1982	1432	479	57	2	30
1983	1122	459	13	14	14
1984	721	592	75	119	97
1985	1093	302	90	33	62
1986	1465	423	14	1	8
1987	915	250	3	12	8
1988	1347	466	36	54	45
1989	1407	414	15	19	17
Average	1188	423	38	32	35

Source: Adapted from SCRP (1996)

Table 5. Soil Loss Rate Estimates Made for Different Land Cover/land

Use in Ethiopia

Land cover/land use Area (%) Soil loss (t/ha/yr)	Area (%)	Soil loss (t/ha/yr)
Cropland (annuals)	13:1	42
Perennial crops	1.7	8
Currently uncultivable land	18.7	5
Totally degraded landscapes	3.8	70
Grazing/browsing land	51.0	5
Wood and bushland	8.1	5
Forests	3.6	1
Average	100.0	12
Source: liurni (1993)

In addition to the volume eroded by runoff and mass wasting, tillage operations cause downslope transfer of a large mass of soil although there is no quantitative data to support the statement in the case of Ethiopia. In Thailand, soil loss by manual tillage on slopes of 30 to 50% was estimated at 8 to 18 t/ha per pass (Turkelboom et al. 1997). In the South Welo highlands, soils are repeatedly tilled (3 to 6 tillage passes being common) to depths of 10 to 20 cm using the ox plough, the maresha, to break up the surface soil, kill and remove weeds and prepare seedbeds (use of herbicide is unknown in the South Welo highlands). Soils are also intensively tilled so that they are pulverized and the small seeds of crops such as wheat, barley, sorghum and teff (Eragrostis teff) germinate and grow well. For example, fields are ploughed five to six times in the preparation of seedbed for teff, one of the most widely cultivated crops in the zone. Such intensive ploughing, however, causes a downslope movement of huge amounts of soil, causing tillage erosion. Farmers explain, in detail, how the traditional tillage operations push the soil, and clods downhill during the preparation of seedbed.

That tillage erosion is a major process behind soil loss on slopes in the South Welo highlands is evident from a number of features of the landscape. The first evidence relates to the more widespread occurrence of shallower soils and frequent rock outcrops on cultivated hill summits and ridge divides when compared to those on the backslopes (Weigel 1986; Paris 1985; Belay Tegene 1997). Had runoff been the major cause of erosion, soil loss would have been more severe on the backslopes where the steeper gradients and the larger volumes of water would increase its eroding capacity. The very shallow soils on ridges and summits, and rock outcrops on divides, are primarily results of tillage erosion.

Further evidence comes from the more rapid build-up of terraces following construction of bonds on cultivated fields. Terraces generally form very rapidly on cultivated fields throughout the South Welo highlands and these rates cannot be explained on the basis of water erosion alone. A fast rate of terrace development due to tillage operations was also observed on experimental plots in Southern Ethiopia (Belay Tegene 1992). The rapid development of traditional terraces above strips of land left in cultivated fields in the Derekolli catchment and some other parts of South Welo highlands is also primarily the result of tillage erosion. The large volume of surface stones mantling the footslopes at the base of the cultivated hillslopes provides further evidence for the downslope movement of soil due to tillage operations (Belay Tegene 1997). Most of these stones originate on the rock outcrops on the hill summits and backslopes and gradually move downhill together with the soil as a result of tillage operations causing severe erosion and degradation of the slopes. Thus, soil depth loss on cultivated fields must be much higher than what is induced by runoff and hence estimates ignoring tillage erosion lead only to gross underestimation of the problem.

5. CAUSES OF ACCELERATED EROSION IN THE SOUTH WELO ZONE

5.1. Major Categories of Factors of Erosion

Three closely inter-linked groups of factors interact in a dynamic system to accelerate soil erosion resulting from both runoff and tillage operations in the South Welo highlands. The first group comprises the biophysical factors that make the soils vulnerable to erosion. These are primarily the steep slopes, the high rainfall erosivity, and the soil erodibility in the case of erosion by water, and the steep slopes in the case of tillage erosion. The second group of factors which Blaikie (1983) rightly referred to as the "real causes" are the non-conservation based farming, which involve intensive tillage of soils, reduction in the frequency of fallows, lack of effective soil conservation practices, etc. The third group of factors, which encourage erosion in the South Welo highlands, comprises the social and institutional factors. These are primarily related to the demographic, economic and political conditions, that not only compel farmers to cultivate fragile environments, but also exert a strong influence on the type of land management practices applied on vulnerable soils.

Almost all models developed to predict erosion by water, in one way or the other, incorporate biophysical, cropping and land management factors. For example, the `Universal Soil Loss Equation' (USLE), expressed as A (soil loss) = RKSLCP, is designed to predict sheet and rill erosion on cultivated fields. The independent variables constituting the equation are three groups of biophysical factors -- rainfall erosivity (R), soil erodibility (K), length and steepness of the slope (SL), and two soil and crop management factors, i.e., C and P (Wischmeier and Smith 1978). However, these kinds of models generally ignore the contribution of tillage operations to accelerated soil depth loss, and as a result soil conservation projects usually tend to be insensitive to the factor. Most of the models that are used in the prediction of erosion rates also fail to integrate the social and institutional factors as predictors of soil loss. In fact, Lewis (1992) argues that one of the reasons for the persistence of erosion throughout the world is lack of success in incorporating the social factors in soil loss models. Almost all the models also ignore the contribution of tillage erosion to soil depth loss.

5.2. The Biophysical Factors

In mountainous areas, such as those of the South Welo highlands, one of the most critical biophysical factors determining vulnerability of soils to erosion is the slope gradient. For example, in the four study sub-catchments, 54 to 74 per cent of the cultivated land has gradients of more than 12 per cent, while 14 to 23 is marked with gradients of more than 27 per cent (table 6). In some of the sub-catchments, more than a tenth of the cropland is marked by slope gradients exceeding 36 per cent. These steep slopes encourage erosion by increasing the volume and velocity of runoff and by encouraging the down slope flux of soil due to tillage.

The high erodibility of the soils of the South Welo highlands is also another factor encouraging erosion by water. The soils with depths of less than 30 cm, i.e., the Leptosols, cover most of the slopes and account for more than 20 per cent of the total cultivated land (table 7). Such shallow soils cannot absorb a large proportion of the rainwater during the wet seasons and as a result cause excessive runoff and erosion. These shallow soils also exacerbate the impacts of the drought years, as they cannot store sufficient water to provide the moisture needed by the crops in the period between the consecutive rains, leading to poor biomass yield and crop cover. The deeper soils on the gentler slopes, which are mainly Vertisols and vertic intergrades (i.e., Vertic Luvisols, Vertic Cambisols, etc.) are also highly erodible because once their cracks close up they develop poor infiltration capacity, generate runoff and encourage rill erosion. The vulnerability of these soils is also increased because of continuous cultivation and subsequent degradation of the soil organic matter. Where the organic matter is reduced to very low levels, the clay easily slakes, disperses and develops surface seals of low permeability, subsequently encouraging runoff and erosion.

The other biophysical factor that encourages erosion in these highlands is the rainfall erosivity. The rainfall in South Welo is concentrated in five to six months, and hence is very erosive. In fact, about 40 per cent of the erosive rains occur in the two summer months of July and August (table 8). These intensive rains encourage accelerated erosion on cultivated fields not only by detaching soil particles but also by generating the runoff needed to entrain and transport them. Rainfall intensity is much higher in the kolla zones, where rainfall is slightly lower than in the woina dega and dega zones. This condition is very well observed in the rainfall records of the Awash basin, which showed intensities of up to 140 mm per hour for altitudes of less than 1800 m and below 100 mm per hour for those above 1800 m (FAO/UN 1965; Constable 1984).

Tile 6. Distribution of Cultivated Land by Slope in the South

Welo Highlands

Slope			Sub-catchments
%	Derekolli	Gido	Nibo	Wurgo	Average
<12	46	26	47	37	39
12-27	40	57	35	41	43
27-36	5	15	6	15	10
36-47	5	1	8	4	5
47-57	2	0	2	3	2
> 57	2	0	3	1	2

Table 7. Distribution of Cultivated Land by Soil Depth in the

Selected Catchments

Soil depth		Sub-catchments
(cm)	Derekolli	Gido	Nibo	Wurgo	Average
<30	22	27	15	25	22
30-90	41	26	36	33	34
>90	37	47	49	42	44

Table 8. Monthly Rainfall (P) and Erosivity (Rm) in Maybar, South Welo (1982-93)

Source: SCRP (1996)

The frequent occurrence of drought in Welo is another factor encouraging erosion. The erect of drought years is evident from the very high soil loss (119 t/ha) registered in 1984, at the Maybar station, despite the very low total rainfall recorded in the year (SCRP 1993) (see also table 4 in the previous section). Drought years encourage erosion primarily for two reasons. Firstly, the soil has very little cover in the drought years and hence is very vulnerable to erosive rains. Secondly, in drought years, rainfall tends to be concentrated within a few weeks. For example, in Maybar, of the annual erosivity of 592 J/mh registered in the drought year of 1984, 80 per cent occurred in the month of March, causing 99 per cent of the annual soil loss in one of the cultivated plots (SCRP 1996).

5.3. The Cropping and Land Management Factors

The factors that directly contribute to accelerated erosion on cultivated fields in the South Welo Zone are the cropping and land management practices. Most of the crop cultivation in these highlands is practiced on steep slopes and shallow soils without appropriate conservation measures. A land capability classification carried out on the bases of slope and soil depth of the croplands shows that 26 per cent of the cultivated land is not at all suitable for cultivation and hence should be put out of production and brought under pasture and/or forest. About 40 per cent of the cultivated land requires application of contour plowing and broad-based terracing while 34 per cent cannot be cultivated without bench terracing (see tables 9 and 10).

However, in spite of the management requirement, most of the cultivated land in the South Welo highlands is not treated with structural conservation measures other than contour furrows left by the plow (contour plowing). Even in the drier highlands, i.e., the kolla zone, where not only erosion but also moisture deficiency is a problem, the traditional soil and water conservation measures applied were not effective enough to arrest accelerated erosion completely. The level 'normal' bunds (and to some extent the level 'fanya juu' bunds) constructed in some areas through food-for-work programs in the 1970s and 80s, and the graded broad-based terraces that had subsequently formed since then have been destroyed in many places following the change in government in 1991. The justification given by the farmers for their action is that these newly introduced structures put large areas of land out of production, allowed very little space for turning the plow oxen, and encouraged water logging, particularly in the wetter highlands. They also pointed out that where they are not properly maintained, bunds may concentrate runoff and cause severe rill and gully erosion in the down-slope fields.

Table 9. Capability of Cropland of the Wurgo, Nibo and Gido Catchment Based on the

Treatment Oriented Classification

Capability Per cent of Recommended conservation treatments

Class cultivated land

C 1 40 Contour cultivation; strip cropping; broad-base terrace

C2 28 Bench terracing

C3 6 Bench terracing

C4 0 Bench terracing and farming operations by hand labor

P 24 Soil depth too shallow for cultivation; use for improved

pasture or rotational grazing system

FT 0 Tree crops with bench terracing; inter-terraced areas in

grass; use contour planting; diversion ditches; mulching  

F 2 Maintain as forest land

Moreover, even in the kolla zone, where indigenous conservation structures are practiced, the traditional conservation measures were not able to completely arrest accelerated erosion (Belay Tegene 1998). Firstly, the inter-structure spacing in traditional terraces are in most cases too wide for the respective slope gradients and soil depths to effectively arrest soil erosion. This encourages insidious sheet wash and rill erosion on the upper and middle parts of the very wide inter-structure spaces. Secondly, most of the traditional structures are not allowed to develop into low gradient or level terraces because of the combined effects of wide interstructure spacing and the practice of frequent bund destruction. Such widely spaced and graded inter-structure slopes encourage both tillage and water erosion.

Table 10. Characteristics and Recommended Treatments of the Land Capability Classes According the Treatment - Oriented Capability Classification Scheme (Sheng 1972)

Group Class Characteristics and Recommended Treatments

Suitable for tillage C 1 Up to 12 % slope; soil depth normally over 10 cm;

contour cultivation; strip cropping; broad base terraces

C2 Slopes 12-27 %; soil depth over 20 cm; bench terracing

(construction by bulldozers); use of four-wheel tractors

C3 Slopes 27-36 %; soil depth over 20 cm; bench terracing on

deep soil (construction by small machines), silt-pits on

shallower soils; use of small tractors or walking tractors.

C4 Slopes 36-47 %; soil depth over 50 cm; bench terracing

and farming operations by hand labor.

P Slopes 0-47%; soil depth too shallow for cultivation; use

for improved pasture or rotational grazing system; zero

grazing where land is wet.

FT Slopes 47-58%; soil depth over 50 cm; use for tree crops

with bench terracing; inter-terraced areas in permanent

grass; use contour planting; diversion ditches; mulching.

F Slopes over 58% or over 47 %; where soil is too shallow

for tree crops; maintain as forest land.

Wetland, liable to P . Slopes 0-47 %; use as pasture

flood; also stony

land F Slopes over 47 %; use as forest

Gullied land F Maintain as forest land

The other major factor contributing to accelerated erosion on cultivated fields is the poor plant cover. The markedly reduced soil cover in such vulnerable soils leads to a sharp increase in accelerated erosion. It must be noted that erosion reaches substantial levels in the tropics where the plant cover is reduced to less than 30 or 40 per cent (Seuffert 1991). As commonly observed in almost all cultivated fields the crop canopy is much lower than the minimum required level for a number of reasons. Firstly, the cultivation of annuals leaves the land exposed to the direct impacts of raindrops for most of the year. In fact in some of the rainiest months the fields are at the seedbed stage of the cropping cycle and therefore very much exposed to erosive rains. Secondly, the poor soil organic matter and fertility management system induces low fertility and poor water holding capacity of the soils and significantly reduces the cover and protection the crop provides even during the maturity stage. The lowest soil cover of course occurs in the kolla zone where low total rainfall coincides with higher rainfall intensity. It must be emphasised, however, that some of the cultivated soils, particularly those on the footsteps and backslopes are very well covered with stone (table 11) and therefore effectively protected from erosion. In fact, erosion by water would have been much higher than the estimated 35 tons/ha/year had a considerable proportion of the cultivated fields not been covered with stones. It should also be noted that where they are very large and their cover rate high, stones are likely to interfere with tillage operation and make crop cultivation difficult.

Accelerated erosion on the cultivated fields is also encouraged by the plough, which increases the erodibility of the surface soils. Apart from directly causing erosion, intensive tillage of the surface soils also breaks and pulverises soils, and makes them vulnerable to detachment by raindrops and to entrainment by runoff. The pulverisation of the surface soils also increases the volume of runoff generated on the cultivated land and result in dense network of rill and accelerated erosion on the soils. The seasonal pattern of land use also contributes to accelerated erosion in the dega zone. For example, the Leptosols, i.e. the shallow soils on the slopes, which are extremely susceptible to drought and crop failure, are more intensively cultivated in meher, i.e., during the big rains. The overlap or coincidence of the very erosive rains and the seedbed stage of the meher crops further accelerates erosion on these wetter highlands. Vertisols in these highland zones are also vulnerable to solifluction and sliding when they occur on slopes.

Table 11. Proportions of Cultivated Land with Various Percentages of Stone

Cover in the Four Sub-catchments of South Welo Zone

 

Stone cover %			Sub-catchments
	Derekolli	Gido	Nibo	Wurgo	Average
<10	63	39	42	-	48
1.1-20	18	21	24	-	21
20-30	7	20	5	-	11
31-50	8	15	6	-	10
>54	4	5	3	-	4

5.4. The Social and Institutional Factors

Social and institutional factors exert influence on erosion because the rate and the spatial and temporal distribution of the process are determined not only by the interactions of the biophysical factors and management factors but also by human circumstances. In fact, it is the social and institutional factors that determine the way land is used and managed and it is only where the land use is inappropriate that accelerated land degradation occurs. As Sanders (1992: 21) rightly argues: "... farmers and other land users rarely deliberately degrade the land from which they have to make a living and feed their families. Incorrect land use and bad management must therefore be due to either ignorance or, more likely, to economic, social, and political pressures that force farmers to use the land in the way that they do." In the case of the South Welo highlands most of the farmers are very well aware of the ongoing accelerated erosion. In fact, as noted in the previous sections, in some areas farmers apply indigenous technologies to control 'erosion; but these conservation measures are not effective enough to completely arrest the problem (Belay Tegene 1998; Dessalegn Rahmato 1991; Yeraswork Admassie and Solomon Gebre 1985). It is primarily the social and institutional factors that compel farmers to practice land use and land management that accelerate erosion and the major ones are population pressure, widespread poverty and the insecure land tenure system.

The rapid population growth and the ever increasing need for more land to maintain required food production is forcing farmers to encroach onto marginal areas such as steep slopes and drier regions. Ethiopia has witnessed a more than three-fold increase in population since 1950 (CSA 1985 and 1988). This population growth, in the absence of technologies, for agricultural intensification and opportunities for off-farm employment, has resulted in a severe shortage of arable land and continuous expansion of cultivation on to the more fragile environments (steep slopes, drought prone areas, etc.) (McCann 1994; Messerli and Aerni 1978). Because of this problem, the total area of land each farmer cultivates has fallen steadily and is now very low. As pointed out earlier the current average arable land holding is 0.63 ha/family while the minimum required to support a family is 1.75 ha. Some farmers have reported that one of the ways by which they managed to provide more food for their families was by extending cultivation on to very steep and fragile slopes. Field observations have revealed that slopes as high as 60 per cent are commonly cultivated without any conservation measures.

Demographic pressure and shortage of land have also compelled farmers in many places to practice continuous cropping and to abandon completely even seasonal fallowing, raising the cropping intensity in most of the double cropping areas to 200 per cent. Farmers pointed out that abandoning seasonal fallowing and continuous cultivation of land was the strategy they adopted to provide food for the growing population. In some cases, the shortage of cultivated land has also led to complete exclusion of legumes from the rotation and its replacement by continuous cultivation of cereals (the main staple food).

Population growth has also caused contraction of grazing land and this, in turn, has forced farmers to rely heavily on cropland for both food and fodder. Currently, the livestock depends heavily on aftermath (stubble) grazing on the farmland. According to some estimates about 60 per cent of the feed during the dry seasons comes from cultivated fields mainly in the form of cereal straws and stover (DPED 1993). Grazing animals normally consume everything that is green and leave behind bare and very much trampled farmland by the end of the dry season. This allows incorporation of very little organic matter into the soil and subsequently leads not only to increased erodibility but also to reduced fertility and sharp decline of biomass yield and crop cover.

The population of South Welo Zone is not only large but also very poor and this has further contributed to accelerated soil erosion and land degradation. Reports indicate that more than 85 per cent of the rural population in Ethiopia lives well below the poverty line suggested by the World Bank, which is 30 US dollars per person per month (Dercon and Krishnan 1996). Moreover, the standard of living of the rural population of Welo is among the lowest even by Ethiopian standards (Dessalegn Rahmato 1991). For example, the data provided by Mesfin WoldeMariam (1991) suggests that the average per capita income of farmers in South Welo is not more than 124 Birr per year (i.e., at the exchange rate of the time this was equivalent to less than 60 dollars per annum).

Obviously, these poor farmers not only tend to raise their income with little concern for the long-term impacts on the environment, but also try to satisfy their short-term needs at the expense of their long-term benefits. As clearly pointed out by the World Commission on Environment and Development (1987: 28): "Those who are poor and hungry will often destroy their immediate environment in order to survive... " They cultivate very steep slopes, indiscriminately cut trees to sell wood and charcoal, and their livestock overgraze grasslands, each of which intensifies the erosion process on slopes. Moreover, these poor farmers cannot invest on modern inputs such as chemical fertilisers, improved seeds, and herbicides. As a result, the average grain and seed yield is very low (less than 10 quintals per ha) (DPED 1993) and hence more land is required to feed the population. This in turn leads to further expansion of crop cultivation on to the steep slopes and fragile environments aggravating the problem of accelerated erosion, subsequent loss of the productivity of the land and further impoverishment of the community inducing an "environment-poverty trap" (Pearce et al. 1990).

Many farmers have also identified land tenure as one of the major factors behind accelerated erosion. In South Welo Zone, as elsewhere in Ethiopia, land is owned by the state and redistributed periodically to accommodate the landless. Under this arrangement a farmer has only a short term use right, i.e., the right to use a piece of cultivated land until redistribution takes place. According to the reports of farmers at least three major and a number of minor redistributions of land have been carried out in South Welo after the initial distribution following the 1975 proclamation. For instance, a study conducted by Yeraswork Admassie (1995), shows that one-third of the farmers in three administrative units of Welo had to completely or partially change their cultivated plots three or more times because of the redistribution (table 12). This study also suggests that at the time more than 50 per cent of the farmers had been using the land they cultivate for not more than 18 years while about 27 per cent had tenure duration of less than 5 years. Farmers argue that this redistribution induces tenure insecurity and reduces the farmer's incentive to invest on conservation and other management practices on the land.

Even in the periods between land distributions, the transfer of use rights is possible through inheritance, only with the approval of the peasant associations. It must be noted that though the government now tries to assure farmers that there will be no other redistribution, this has failed to convince them as the last distribution was done in spite of repeated similar statements in the past. Moreover, farmers know that redistribution is carried out to give land to those who have none, and that there will always be a landless population that has to be accommodated through further land redistribution schemes.

Table 12. Number of Times Plots Were Changed Due to Redistribution and Percentages of Farmers Affected in Three Administrative Units of Welo after the Agrarian Reform of 1975

Administrative unit	Number of times plots were re-adjusted or changed
	Once	Twice	Three or more
Yeju	23	48	65
Ambassel	40	45 .	16
Dessie Zuria	48	34	18
Avergage	37	42	33

Source: Adapted from Yeraswork Admassie (1995:166)

Although land redistribution is intended to address the pressing demand for land by the landless population, the practice is detrimental to soil protection and conservation. The tenure arrangement ensures neither long-term use-right nor transferability of use-right and thereby creates serious tenure insecurity among the farmers. As Mesfin Wolde-Mariam (1991) rightly points out farmers are not sure if they or their children will still be cultivating the same land in four to five years' time. This encourages land degradation since the peasants have little incentive to invest in the land they will not be farming in years to come (Campbell 1994). They cannot invest on improved management and resource conservation measures under this condition, because the benefits from such investments are realised only in the long term. In fact, an `IUCN Sahel Programme Study' (Hutchinson 1991: 135) notes, this land tenure system has "restricted the peasantry's potential to increase agricultural output or maintain appropriate conservation-based farming systems".

Moreover, even the short-term use right is ensured only as long as there is a crop in the field. The cultivated land is held as private as long as the crop is there. After the annual crops have been removed all the members of the community have the right to graze their domestic animals on the aftermath. This alternation between use as private domain and as common property resource discourages farmers from leaving crop residue, i.e., stalks, leaves, etc. on farms thereby leading to depletion of the soil organic matter and acceleration of erosion. The state ownership of land has also created conditions for the expansion of crop cultivation into traditional grazing and forestland and marginal areas such as the steep slopes. A farmer can cultivate the non-arable land (no man's land) provided he obtains permission from the peasant association and this can be easily attained by corrupting the leaders.

6. SUMMARY AND CONCLUSION

Crop cultivation in the South Welo highlands without appropriate conservation measures had in the past exposed, and continues at present to expose the soils to stress, thereby accelerating the process of soil erosion. Accelerated erosion resulting from the combined effects of both water and tillage operations constitutes the greatest threat to agriculture in these highlands. This paper has identified three major and inter-linked groups of factors that induce accelerated erosion in the zone. The first group comprises the biophysical factors,, such as rainfall erosivity, soil erodibility, and the length and gradient of the slope which mainly determine the vulnerability of soils to erosion. However, it should be emphasised that these factors only create the condition for accelerated erosion but cannot be the direct causes.

The second group, which constitutes the real causes of erosion, are the nonconservation-based land management and cropping system. Observations made in different parts of the South Welo highlands clearly indicate that a considerable proportion of crop cultivation is practiced on steep slopes and shallow soils. In fact, a capability classification shows that 40 per cent of cultivated land requires application of contour plowing and broad-based terracing, 34 per cent bench terracing, while 26 per cent is not at all suitable for cultivation and hence should be put out of production and brought under pasture and/or forest. In most of the cultivated fields, no terracing is practiced and, as a result, erosion is proceeding at accelerated rates. Even where soil conservation is traditionally practiced erosion is not completely arrested because in most cases the structures are not allowed to develop into low gradient or level terraces because of the combined effects of wide inter-structure spacing and the practice of frequent bund destruction (Belay Tegene 1998).

The third group constitutes the social and institutional factors that compel farmers to continuously cultivate vulnerable soils and use the fragile environment without appropriate conservation measures: these are rapid population growth, increasing scarcity of suitable cropland and contracting land holding per family. These, combined with widespread poverty, have left the farmer with no option other than encroaching more and more onto the steep slopes, with little concern about the long-term environmental impacts. The land tenure arrangement further aggravates the problem by providing a disincentive for soil conservation practices and discouraging farmers from investing on improved land management practices whose benefits are realised only in the long run.

Soil conservation strategies designed to arrest accelerated erosion in the South Welo highlands on a sustainable basis cannot be effective where these causative factors are not recognised and properly addressed. Firstly, sustainable soil conservation can only be possible by developing and incorporating into the farming system appropriate technologies. The latter refer to technologies that are efficient in controlling erosion, "consistent, ... with important themes or patterns in that people's culture" (Firey 1960: 30), and effective in considerably increasing the short term or immediate benefits to the farmers. Appropriate conservation technologies may be built by refining the indigenous soil/land management practices (multiple cropping, crop rotation, application of manure and compost, terracing, agroforestry, etc.) (Belay Tegene 1998, 1999). Indigenous conservation practices are based on the farmers' knowledge and skills, perception of the problems and their needs, aspirations and survival strategies; and hence are compatible with the environment, land use and farming system.

Secondly, the adoption of soil conservation can be very much improved by encouraging yield-increasing technologies so that the value of the land and the costs of erosion would be increased. In this regard measures should be taken to improve the fertility and productivity of the soil by integrating artificial fertilisers (i.e., the NPK-fertilisers) and indigenous organic matter and nutrient restoration practices such as use of manure and rotational cropping. The incorporation of the latter two improves not only the fertility but also the structure and water holding capacity of the Leptosols landscapes. Extension workers should also advise farmers to switch to higher-value crops so that they get better income per unit area and compensated for what they put in as conservation measures. This provides the farmer with incentives to protect his soil and adopt conservation measures.

Where soil conservation is integrated with the fertility management package, the loss by erosion of not only soil and rainwater but also of organic manure and the costly artificial fertilisers and other modern inputs would be considerably reduced, thereby significantly increasing the yield per unit area. These conditions improve the perceived economic advantage and profitability of the conservation packages and may provide the incentive for their adoption by farmers. Once applied, the conservation structures are also effectively maintained, leading to sustainable stabilisation of the landscape. It should also be noted that most of the investments in conservation measures yield benefits in the long run and hence they should be supported by subsidies and policy instruments that increase immediate benefits and incentives to farmers.

Thirdly, measures should be taken to remove the social and institutional constraints to soil conservation and land rehabilitation programmes. Population pressure can be gradually relieved through policies and extension education that encourage family planning and population control. The government should also encourage soil erosion control through policies that improve the security on land. Where major changes in government policies are not possible they should at least be supplemented by legislation that encourages tenure security and conservation-based farming systems. For instance, land could remain under state ownership but at the same time the farmer could have the right to use the land as long as he lives and later on pass it to his heirs.

It is also impossible to have lasting success in the control of soil erosion where in the long run appropriate measures are not taken against poverty. The creation of possibilities for non-agricultural employment could be one of the strategies that could be adopted to prevent encroachment of crop cultivation on to steep slopes. It should also be stressed that the participation of peasant farmers in conservation planning and activities is crucial for the success of conservation activities. However, participatory soil conservation and land rehabilitation can be ensured only if decision making and the benefits from activities along with responsibilities are devolved to the local community. It must be noted that the imposition of conservation measures by government agencies and the associated restrictive measures such as exclusion of livestock from closed and reforested areas, adopted as a policy during the previous government, is very much resented by the farmers and hence could not be sustained. Farmers felt that most of the land that was closed and afforested constituted important areas of cultivation and pasture, and hence they had to destroy the forest and reclaim the land.

ACKNOWLEDGEMENTS

The author gratefully acknowledges the John D. and Catherine T. MacArthur Foundation for funding this research. Thanks are also due to Ato Berhanu Tefera and the anonymous reviewers of EASSRR for their constructive comments and suggestions on the manuscript. The author is also indebted to Professors D. Crummey and Bahru Zewde, and Ato Dessalegn Rahmato for their keen interest and encouragement. The assistance of Ato Tessema Bekele and the farmers of South Welo is also acknowledged.

REFERENCES

Amare Getahun. 1984. Stability and Instability of Mountain Ecosystems in Ethiopia. Mounatin Research and Development 4 (1): 39-44.

Belay Tegene. 1992. Erosion: Its Effects on Properties and Productivity of Eutric Nitosols in Gununo Area, Southern Ethiopia, and Some Techniques of Its Control. African Studies Series A9. Bern: Geographica Bernensia.

_______. 1997. Variabilities of Soil Catena on Degraded Hillslopes of Watiya Catchment, Welo, Ethiopia. SINET: Ethiopian Journal of Science 20 (2):151-175.

_______. 1998. Potentials and Limitations of an Indigenous Structural Soil Conservation Technology of Welo, Ethiopia. Eastern Africa Social Science Research Review 14 (1):1-18.

_______. 1999. Indigenous Soil Knowledge and Fertility Management Practices in Southern Wallo, Ethiopia. Forthcoming in the Journal of Ethiopian Studies 31 (1): 123-158.

Blaikie, P.M. 1983. The Political Economy of Soil Erosion. In Progress in Resource Management and Environmental Planning, edited by T. O. Riordan, and R. K. Turner. Vol.4: 29-55. New York: John Willey & Sons Ltd.

Campbell, J. 1994. Land or Peasants? The Dilemma Confronting Ethiopian Resource Conservation. African Affairs 90:5-21.

Constable, M. 1984. The Degradation of Resources and an Evaluation to Combat It. Working Paper 19. Ethiopian Highlands Reclamation Studies. Addis Ababa LUPRD/MoA and FAO/UN, Ethiopia.

CSA. 1985. Population Situation in Ethiopia - Past, Present, and Future. Addis Ababa: Central Statistical Authority.

________. 1988. Population Projection of Ethiopia - Total and Sectoral (1985-2035). Addis Ababa: Central Statistical Authority.

Dercon, S. and P. Krishnan. 1996. A Consumption-based Measure of Poverty for Rural Ethiopia in 1989 and 1994. In The Ethiopian Economy, Poverty and Poverty Alleviation, edited by B. Kebede, and M. Taddesse. Proceedings of the 5^th Annual Conference on the Ethiopian Economy, pp. 77-100. Addis Ababa: Ethiopia Department of Economics, Addis Ababa University, and the Ethiopian Economic Association.

Dessalegn Rahmato. 1991. Famine and Survival Strategies: A Case Study from Northwest Ethiopia. Uppsala: Nordiska Afrikainstitutet.

DPED. 1993. A Preliminary Survey on Utilization and Conservation of Natural Resources. Department of Planning and Economic Development for South Wollo, Bureau of Planning and Economic Development for Region Three, Dessie, Ethiopia. Unpublished

Ehrlich, P. R, A. H. Ehrlich and J. P. Holdren. 1977. Ecoscience: Population, Resources, Environment. San Francisco: Freeman.

Ehrlich, P.R., A. H. Ehrlich and G. C. Daily. 1997. The Stork and the Plow: The Equity Answer to the Human Dilemma. London and New Haven: Yale University Press.

FAO/UN. 1965. Report on Survey of the Awash River Basin. Rome: FAO/tJN.

Firey, W: 1960. Man, Mind and Land: A Theory of Resource Use. Illinois: Free Press of Glencoe.

Hudson, N. 1995. Soil Conservation, 3rd edition. Amen, Iowa: Iowa State University Press.

Hurni, H. 1993. Land Degradation, Famine, and Land Resource Scenarios in Ethiopia. In World Soil Erosion and Conservation, edited by D. Pimental. Cambridge: Cambridge University Press.

Hutchinson, R. A 1991. Fighting for Survival: Insecurity, People and Environment in the Horn of Africa, edited by R A. Hutchinson (based on study by B. C. Spooner, and N.Walsh). Gland, Switzerland: IUCN.

Lewis, L. A 1992. Terracing and Accelerated Soil Loss in Rwandian Steeplands: A Preliminary Investigation of the Implications of Human Activities Affecting Soil Movement. Land Degradation and Rehabilitation 3:241-246.

McCann, J. 1994. Ethiopia In Drought Follows the Plow: Cultivating Marginal Areas, edited by M. H. Glantz. Cambridge: Cambridge University Press.

Mesfin Tadesse. 1990. An Illustrated Guide to the Trees and Shrubs in the Red Cross/UMCC-DPP Areas in Welo, Ethiopia. Addis Ababa: ERCS.

Mesfin Wolde-Mariam. 1991. Suffering under God's Environment: A Study of the Predicament of Peasants in North-Central Ethiopia. Bern, Switzerland: The African Mountain Association and Geographica Bernensia.

Messerli, B., and K. Aerni. 1978. Cartography and Its Application for Geographical and Ecological Problems, edited by B. Messerli, and K. Aerni (With Maps). Bern, Switzerland: Geographica Bernensia.

Paris, S. 1985. Soil Survey of the Borkena Area (Welo). AG:DP/ETH/82/010, Field doc. 7., LUPRD/UNDP/FAO. Addis Ababa: Land Use Planning and Regulatory Department, Ministry of Agriculture, Ethiopia.

Pearce, D. B., and E. Markandia 1990. Sustainable Development: Economics and Environment in the Third World. Worcester: Billing.

          Pimental, D. 1993. Overview. In World Soil Erosion and Conservation, edited by D.

          Pimental. Cambridge: Cambridge University Press.

Sanders, D. W. 1992. Soil Conservation: Strategies and Policies. In Soil Conservation for

Survival, edited by Kebede Tato and Huns Hurni. (n. p.): Soil and Water Conservation Society.

SCRP. 1996. Data Base Report (1981-1993) Series I: Maybar Research Unit (First Draft). Soil Conservation Research Project, Addis Ababa, Ethiopia.

            Seuffert, O. 1991. Ecomorphodynamics and Soil Erosion. Applied Geography and 

Development 38:7-21.

Sheng, T.C. 1972. A Treatment-Oriented Land Capability Classification Scheme for Hilly Marginal Lands in the Humid Tropics. Journal of the Scientific Research Council Jamaica 3:93-112.

Turkelboom, F., J. Poeson, I. Ohler, K Van Keer, S. Ongprasert and K. Vlassk. 1997.

Assessment of Tillage Erosion Rates on Steep Slopes in Northern Thailand Catena 29:29-44.

Weigel, G. 1986. The Soils of the Maybar/Welo Area: Their Potential and Constraints forAgricultural Development. African Studies Series A4. Bern, Switzerland: Geographica Hernensia

Wischmeier, W.H., and D. D. Smith. 1978. Predicting Rainfall Erosion Losses -A Guide to Conservation Planning. Handbook No. 537. United State Department. of Agriculture.

World Commission on Environment and Development. 1987. Our Common Future. Oxford: Oxford University Press.

Yeraswork Admassie. 1995. Twenty Years to Nowhere: Property Rights, Land Management and Conservation in Ethiopia. PhD. Thesis, Department of Sociology, Uppsala University.

Yeraswork Admassie and Solomon Gebre. 1985. Food-for-Work in Ethiopia: A socio-economic Survey. Report prepared for the World Food Programme and the Ministry of Agriculture. Research Report No. 24. Addis Ababa: Institute of Development Research (IDR), Addis Ababa University.