Assessment of soil erosion in cultivated fields using a survey methodology for rills in the Chemoga watershed, Ethiopia

Abstract

Soil erosion by water is recognized to be a critical economic problem in highland Ethiopia. However, nearly all the available information about its severity and economic costs are extrapolated from plot and micro-watershed level studies which are too few in number to represent the diverse environments of the country. Moreover, plot and watershed level studies do not show actual soil losses from cultivated fields, while understanding the magnitude of soil loss at the field scale is important for practical conservation planning. This paper reports results of field-scale soil erosion assessment that employed a survey methodology for rills and was conducted over two wet seasons (the years 2000 and 2001) at two sites, Kechemo and Erene, located in the upstream and downstream reaches of the Chemoga watershed, northwestern highland Ethiopia. The two wet seasons average rill erosion magnitudes were 13.5 Mg ha⁻¹ in the Kechemo and 61 Mg ha⁻¹ in the Erene. Assuming that interrill erosion contributes 30%, actual soil losses were around 18 Mg ha⁻¹ in the Kechemo and 79 Mg ha⁻¹ in the Erene. These estimates, which are well in agreement with results obtained by measurements in a nearby experimental micro-watershed, reveal that soil erosion is a threat to agricultural production in the study area and conservation measures are needed. Soil erosion showed significant spatial (between and within the two sites) and temporal variations. Hence, soil and water conservation (SWC) measures that fit well into local-scale circumstances will be realistic and acceptable to the farmers. Additionally, the problem of soil erosion should be tackled in the watershed context, because there is a strong physical interdependence between upstream and downstream areas. Finally, the study confirms that the rill survey approach gives good semi-quantitative information on soil erosion in real life situations of diverse farming and land use practices in a fast and inexpensive way; and it is commendable for practical conservation-oriented soil erosion assessment purposes.

Introduction

Soil erosion is presently a global environmental crisis that is severely affecting global food security (Pimentel et al., 1993, Pimentel et al., 1995, Lal, 2001). Although many countries of the world suffer from the problem of accelerated soil erosion, the developing countries suffer more because of the inability of their farming populations to replace lost soils and nutrients (Erenstein, 1999). In Ethiopia, one of the least developed countries of the world, accelerated soil erosion by water constitutes a severe threat to the national economy (Sutcliffe, 1993, Hurni, 1993). As estimated by Hurni (1993), soil loss due to water erosion is about 1493 million Mg per annum. Of this, nearly half is estimated to come from cultivated fields, which account for only about 13% of the country’s total area. The estimated rate of soil loss in the cultivated fields is 42 Mg ha⁻¹ per year, or 4 mm of soil depth per annum, which by all measures exceeds the rate of soil formation. With this rate of loss, much of the slopes of the highlands, where there is only a thin soil layer, will be totally stripped of the soil mantle in less than two centuries.

A study by FAO (1986) estimated an even higher rate of soil loss: 2 billion Mg per year in the country as a whole and around 100 Mg ha⁻¹ per year from cultivated fields. According to this study, some 50% of the highlands was already ‘significantly eroded’, and erosion was causing declines in land productivity at the rate of 2.2% per annum. The study also predicted that erosion would have reduced per capita incomes of the highlands population by 30% by the year 2010. On the other hand, around 85% of the 64 million Ethiopians derive livelihood directly from the land resource; and this population is currently growing at the rate of about 2.7% per annum, requiring food production increases of at least the same rate (Sonneveld, 2002).

The reported estimates of soil loss in the country are certainly worrisome. However, nearly all the available information about the magnitudes, severity and economic costs of soil erosion are extrapolated from plot and micro-watershed level studies which are too few in number to represent the diverse environments of the country. Moreover, plot and watershed level studies have their own limitations. Strictly speaking, plot level data indicate only the magnitude of soil loss at that particular area which is confined and excluded from interaction with its surroundings. Deposition, which is the other important process in the erosion-sedimentation continuum at larger spatial scales, is almost completely excluded by the plot level measurements. Generally, extrapolating results from the researcher-manipulated plots to larger spatial scales can be misleading because erosion shows extreme spatial variation with variations in rainfall energy, gradient and length of slopes, inherent soil characteristics affecting its erodibility and land use and land management practices. The variability in soil erosion is very large even between replicated plots, which is a result of natural variability and measurement variability (Nearing et al., 1999).

On the other hand, watershed level data are spatially aggregated, in the sense that only sediment yields at a point outlet are measured and how much of it comes from which part of the watershed remains unknown. In other words, watershed level data do not indicate actual soil losses from cultivated fields, which are the units of land use and management by the farmers, to be readily used as inputs for planning of soil and water conservation (SWC) intervention. Thus, as Herweg (1996) has succinctly put it; “… there is a measurement gap between plot and catchment level”, for SWC planning purposes. For practical conservation planning purposes, understanding the magnitude of soil loss at the field scale, or the spatial distribution of the problem, is very important. It has immediate significance to development agents and field technicians for a targeted and cost-effective conservation intervention by identifying most vulnerable landscapes and setting of priorities.

Rill erosion constitutes one of the mechanisms of soil loss by water on agricultural land. It is probably the most important form because in addition to being an erosion feature in itself, rills serve the purpose of transporting materials supplied by the interrill (splash) erosion. Hence, assessment of soil loss by surveying rill erosion gives a good understanding of the process of land degradation due to erosion by water. Rill erosion survey is a semi-quantitative method for assessing the extent of erosion damage under field conditions, without involving expensive instrumentation, long lead times and/or sophisticated modelling (Herweg, 1996). It is a more conservation-oriented method of soil erosion assessment than the plot and watershed level studies (Herweg, 1996). Some researchers also argued that good field surveys of erosion produce results fairly comparable with test plot derived data (Govers, 1991, Evans, 1993). According to Herweg (1996), results from erosion survey are within 15% accuracy for careful measurements.

Obviously, however, being a semi-quantitative and qualitative assessment, survey results cannot be taken as accurate estimations of soil loss. But, the low cost and the ease with which it can be applied under natural conditions at times compensate for the precision and the high cost test plot and watershed level measurements command. Generally, erosion survey is currently accepted as a good alternative approach to soil erosion research for it has multiple advantages (fastest, cheapest and under actual on-farm situations) (Turkelboom and Trébuil, 1998). Vandaele and Poesen (1995) also underline the fact that measuring volumes of erosion features can be done quickly and with a satisfactory precision (they measured rills and ephemeral gullies in their study). Turkelboom and Trébuil (1998) recommended more studies to be carried out using the survey approach for improving the methodology and to use it for practical conservation planning purposes. The objectives of this study were to: (i) assess the magnitude of soil erosion at the scale of cultivated fields by using a survey methodology for rills in a typical watershed (the Chemoga watershed) in the northwestern highland Ethiopia; and (ii) evaluate the significance of the rill survey methodology as an alternative approach to soil erosion research.