METHODOLOGIES FOR THE ECONOMIC ANALYSIS OF SOIL EROSION AND CONSERVATION
by
Rebecca Clark
Overseas Development Group University of East Anglia Norwich
Acknowledgements
The Centre for Social and Economic Research on the Global Environment (CSERGE) is a designated research centre of the U.K. Economic and Social Research Council (ESRC).
This paper was originally written for the Forest Land Use Mapping Project (FORLUMP) Kandy, Sri Lanka, funded by the Overseas Development Administration (ODA).
The author is grateful to the following for their assistance and support in carrying out this study: Roger White (FORLUMP), Mark Lewis (ODA), Michael Stocking (UEA) and Tom Crowards (CSERGE).
ISSN 0967-8875
Abstract
Land degradation caused by soil erosion is a major threat to the sustainability of agriculture. The costs of erosion can be used to prioritise implementation of soil conservation, and economic analysis of alternative conservation technologies can be used to identify courses of action that efficiently employ available resources. A number of different approaches are currently used for the valuation and economic analysis of erosion and conservation. This paper reviews the approaches used for the on-site and off-site effects of erosion and conservation, primarily in developing countries.
The on-site effects of erosion can be valued in terms of the impacts on soil properties, using the replacement cost, or the impacts on crop production. Valuation of conservation technologies requires additional considerations: the effectiveness of the technology in reducing erosion; and valuation of additional benefits of the technology such as animal fodder production. The off-site effects of soil erosion are numerous but are excluded from many studies. Valuation is generally limited to the effects of reservoir sedimentation on hydro-electricity generation and water supplies for irrigation.
Economic appraisals of soil erosion and conservation generally use financial and/or cost benefit analysis. These appraisals are plagued with problems concerning the availability and accuracy of data, making sensitivity analysis a necessary accompaniment. The studies reviewed vary widely in the discount rates and time periods used, and the extent of sensitivity analysis. It is therefore essential that the findings of studies are accompanied by explicit details of the assumptions and any necessary cautions about the quality and reliability of data.
1 1. Introduction
Soil erosion is a process that occurs on land under agricultural production or natural vegetation. It involves the movement of soil particles on the soil surface as a result of the actions of rain, wind or disturbance of the soil surface. The soil particles, or sediment, may be deposited elsewhere on the land or enter the water system. The rates at which soil erosion occurs vary greatly between different locations and over different periods of time.
There is great concern, particularly in developing countries, about land degrad- ation resulting from soil erosion and its impacts on agricultural production. In some areas of the world it may not be possible to sustain existing levels of crop production in the long term, even with increased applications of fertiliser. This will affect the sustainability of rural communities and may even reduce national food supplies. Erosion also has effects downstream. The eroded sediment can be deposited in reservoirs and reduce hydro-electricity generation and water supplies for irrigation. The sediment can also reduce the operational efficiency of irrigation systems and impair the quality of drinking water.
Concerns about the effects of soil erosion have led to increased promotion of soil conservation technologies in developing countries. Donor and government funds have been invested in watershed management and soil conservation projects and farmers have been encouraged to invest their own funds and labour in soil conservation technologies. In order to increase the efficiency with which these resources are allocated to soil conservation activities, the effects of soil erosion need to be assessed. The costs of erosion can be used to determine the priority for action and the benefits of potential courses of action can be used to assess their suitability and efficiency in resource use.
Considerable difficulties arise in the valuation of the effects of soil erosion and conservation. Neither are traded directly in the market place so their economic values have to be inferred and there is no single recognised method for their valuation. This study aims to provide a guide to the different approaches adopted in the literature for the economic analysis of soil erosion and conservation. The possible on and off-site effects of soil erosion and conservation are briefly reviewed and different approaches that have been adopted for their valuation are examined in detail. Finally, the methods that may be used for the economic analysis of soil erosion and conservation (financial analysis, cost benefit analysis and cost effectiveness analysis) are briefly discussed.
2 The study examines only the methodological aspects of the economic analysis of erosion and conservation and makes no attempt to include or compare the results of different studies. The results of economic analyses are highly dependent on the data and assumptions and cannot therefore be examined independently. Similarly, the study does not examine the effects of different valuation approaches on the figures calculated for soil erosion and conservation. It simply presents the approaches that may be adopted and discusses some of their merits and weaknesses. It is up to the individual analyst to select the appropriate approach for a specific analysis.
The valuation of erosion and the valuation of soil conservation are examined separately in this study. To some extent they represent two sides of the same coin; the benefits of conservation are the reductions in the costs of erosion. But the valuation of conservation has further dimensions. Soil conservation cannot be assumed to completely reduce the effects of erosion; the extent of the decrease has to be specified. Soil conservation may also have benefits additional to reduced erosion, such as production of animal fodder, fruit crops or decreased production costs. The approaches adopted for valuing conservation generally involve the valuation of reduced soil loss. For this reason the reader is referred to the sections both on the valuation of erosion (Section 2) and the valuation of soil conservation (Section 3) for details on approaches for the valuation of erosion.
Valuation of the impacts of erosion are divided into those experienced on the land from which the soil is lost ("on-site") and those experienced elsewhere ("off-site"). Methodologies for valuation of the on-site effects of soil erosion are examined in Section 2. The approaches used for quantifying and valuing the on-site effects of conservation are given in Section 3. Valuation of the off-site effects of both erosion and conservation is examined in Section 4. Section 5 reviews the methods that can be used for the economic analysis of soil erosion and conservation. Finally the overall findings of the report are discussed in Section 6. This review examines analyses that have been carried out in develop-ing countries, but also includes relevant studies based in the US and Australia. The review of the literature is in no way comprehensive, but examines a sample of the studies.
3 2. Valuation of the On-Site Effects of Soil Erosion
There are two main approaches for the valuation of the on-site effects of erosion: the impact on the properties of the soil and the impact on agricultural production. The effects of erosion on soil properties can be examined from the perspective of soil as a resource, or in terms of certain soil characteristics as indicators of soil productivity (e.g. soil nutrient content, soil moisture capacity). The effects of erosion on agricultural production can be valued in terms of reductions in crop yields. This may be the reduction in market value of crop production or, where the farming system involves the production of fodder crops, erosion can be valued in terms of the decrease in livestock production. Crop yields are not dependent on soil productivity alone; yields are determined by the interaction of a number of factors such as fertiliser application rates, climate, pests, and the use of irrigation. The impact of erosion on soil properties can, however, cause declining crop yields, particularly in the absence of increased inputs.
The following sections examine the different approaches adopted by studies for valuation of the on-site impacts of soil erosion. Many of the economic analyses employ data on soil erosion rates; sources of these data are reviewed in Section 2.1. The effects of erosion on soil properties are summarised in Section 2.2 and are followed by a review of methods adopted for this approach to valuation. Section 2.3 summarises the impacts of erosion on crop production and reviews the valuation methods that have been used. Finally, alternative approaches to valuation such as hedonic pricing and contingent valuation are briefly reviewed in Section 2.4.
2.1 Soil erosion rates
Many of the studies on the economics of soil erosion and conservation employ data on erosion rates in their analyses. Several sources of these data are summ- arised below, followed by a review of their respective merits and weaknesses.
Data on soil erosion rates can be obtained from:
experimental plot studies, rapid assessment of field conditions, sediment measurements for experimental studies of sub-catchments and catchments, empirical equations, such as the Universal Soil Loss Equation (USLE) and the Soil Loss Estimation Model for Southern Africa (SLEMSA).
4 There are a number of problems that arise with these different sources of data. Gully erosion can be a major contributor to soil erosion on a catchment scale but is not measured by any of the above data sources apart from catchment studies. The USLE examines only sheet and rill erosion and plot studies generally measure only sheet erosion. Furthermore, plot studies measure only the movement of soil over a relatively small area whereas catchment-level studies measure the actual loss of soil from the land and take into account the deposition of sediment on the land surface. However, there are also problems associated with catchment-level data; the calculation of erosion rates from recordings of river sediment loads requires the specification of a sediment delivery ratio, the value of which is unknown for most catchments. Furthermore, river sediment loads may not accurately reflect erosion rates because of the deposition of sediment in the river system and the inclusion of bed load and river bank erosion in sediment measurements.
Discrepancies arise in the soil erosion rates used; some studies use gross erosion rates whereas others take the rate of soil formation into account and use net soil erosion rates in their analyses. The calculation of net rates is plagued by the absence of accurate information on soil formation rates and the lack of consensus as to whether topsoil or subsoil formation rates should be used, although the former can be ten times the latter (Stocking, 1995). Soil formation rates are, however, very low and as a result the use of gross or net soil erosion rates has a negligible effect on the results of economic analyses.
There are particular concerns about soil nutrient loss data that originate from plots using artificial de-surfacing to simulate erosion. Soil erosion is a selective process; it removes predominantly the smaller particles of the soil, which consist primarily of organic matter and clay particles. A large proportion of soil nutrients are found in these soil fractions, so erosion selectively removes nutrients from the soil. Eroded sediment has been found to be 2.5 times richer in nutrients than topsoil for several soil types in Zimbabwe (described as the sediment enrichment ratio) (Stocking, 1995). However, artificial de-surfacing experiments are non-selective and involve the progressive removal of all particles in the top soil. The results measure soil nutrient loss effectively in terms of the soil nutrient content of top soil and therefore under-estimate erosion-induced losses of soil nutrients.
The data on erosion rates and soil nutrient loss are site-specific and tend not to be transferable between geographical areas. Erosion rates and nutrient losses are affected by characteristics of the soil, climate, land, vegetation, and crop and land management systems. The transfer of data between temperate and tropical soils is especially limited. The tropics generally experience higher erosion rates than temperate zones; there tends to be a greater concentration of nutrients and organic
5 matter in the upper horizons of tropical soils; and rooting depths tend to be more restricted (El-Swaify, 1993).
Finally, problems that can be encountered in using empirical models to calculate erosion rates are illustrated by examining the USLE and SLEMSA.
The Universal Soil Loss Equation (USLE) was developed in the 1960s using an extensive set of data for the US Midwest. It takes the form of:
E = R . K . L . S . C . P where: E = mean annual soil loss, R = rainfall erosivity index, K = soil erosivity index, L = slope length, S = slope steepness, C = crop factor, P = conservation practice factor.
Values are available for the above factors for the US Midwest, but can not be assumed to be representative of other geographical regions. A number of studies have applied the USLE in other areas, in tropical as well as temperate zones. Some of these adapt several, if not all, of the factor values for local conditions. However, the adjustments are not always explicit and it is often unclear how representative the USLE is of the area studied. Morgan (1986) insists that "The equation [USLE] should not be applied to conditions for which the factor values have not been determined".
The USLE was originally derived from experimental data collected over a period of 20 years and so calculates average soil loss over the long term. Since the USLE examines only sheet and rill erosion and excludes gully erosion it calculates soil movement, not soil loss, as it does not account for the deposition of sediment. In summary, the USLE is not suitable for studies in areas for which appropriate factor values are not available, for studies on a catchment scale and for the estimation of soil erosion rates in the short term.
A simpler empirical model that requires less data is the Soil Loss Estimation Model for Southern Africa (SLEMSA). This was developed in the 1980s in Zimbabwe and is similar in structure to the USLE. Lu and Stocking (1994) state that "it has the advantages of reasonable accuracy, simplicity, flexibility and suitability for less developed countries with a limited amount of data". SLEMSA takes the form:
6
E = K . C . X where: E = predicted mean annual soil loss, K = mean annual soil loss from a standard plot, C = crop cover factor, X = a factor of slope length and angle.
2.2 Soil properties
The following section starts with a summary of the impacts that erosion can have on the properties of the soil. It then reviews the different methods that have been adopted for valuation of the impacts of erosion on soil properties.
Erosion can affect all aspects of the soil: physical, chemical and biological. The following list provides an indication of the range and number of possible impacts; for further details the reader is referred to a standard text such as Hudson (1981). The effects of erosion on soil may include:
reduction in the depth of top soil; reduction in soil nutrient content. Soluble nutrients (e.g. nitrates and some phosphates) are leached from the soil in runoff; relatively insoluble soil components (e.g. insoluble phosphates) are lost in sediment, including reserves of soil nutrients in particles of organic matter (e.g. nitrogen) (Lal, 1987). Studies generally only examine the impacts of erosion on nitrogen, phosphorous and potassium, but other soil nutrients are also affected; reduction in soil organic matter content. This affects soil nutrient content (organic matter is the major source of nitrogen), soil fauna and hence soil structure; increase in the risk of soil acidification or toxicity; loss of applied chemicals in runoff and sediment; reduction in rainfall infiltration ie: increase in runoff; reduction in soil moisture capacity as a result of the selective removal of organic matter and clay particles; undesirable soil surface characteristics e.g. crusting and sealing, which further reduce infiltration; removal of soil to the extent that it exposes subsoil or bedrock at the soil surface; impairment of the soils physical structure: texture, structure and bulk density;
7 reduction in the transmission of water in the soil due to deterioration of soil structure; reduction in gaseous exchange resulting from deterioration of soil structure; reduction in soil fauna, involved in nutrient recycling and the maintenance and improvement of soil structure; loss of land area due to gullying and landslides.
Valuation of the impacts of erosion on soil properties focuses on the reduction in soil nutrient content and the physical loss of soil. It takes into account few, if any, of the other impacts listed above. The most common approach for valuing the loss of soil and soil nutrients is the replacement cost. This is based on the cost of replacing lost soil nutrients with artificial fertilisers; it may also include the cost of physically returning eroded sediment to the land. The valuation may be carried out with a view to the replacement of nutrients as a potential course of action, and so include the costs of the labour required to apply artificial fertilisers (Kim and Dixon, 1984). Alternatively a more conceptual approach may be adopted which includes only the cost of artificial fertiliser in the valuation (Stocking, 1988).
The replacement cost has major inherent problems; it represents a potential course of action that cannot be adopted for both practical and environmental reasons. The approach values erosion in terms of its effects on the soil content of several nutrients (generally nitrogen, phosphorous and potassium) and may also incorporate the effect of erosion on soil depth. It does not examine any other aspects of erosion, such as its effects on the soils physical structure, moisture capacity, organic matter content, soil fauna and the levels of the many other soil nutrients. El-Swaify (1993) states that "The replenishment of soil nutrients (as added fertilisers) is by itself insufficient to restore original soil productivity. Physical and biological rehabilitations are often also required.". This is illustrated by an example in Thailand (Attaviroj, 1990) where the loss of rice yields as a result of erosion can not be recovered using artificial fertilisers alone, because of the reliance of rice on soil organic matter and soil structure.
The replacement cost also fails to recognise the different forms in which nutrients exist in the soil. Erosion removes nutrients from the soil not only in forms available to plants but also from soil reserves of nutrients in fixed forms that are unavailable to plants. However, artificial fertilisers supply nutrients only in available form. The replacement of soil nutrients with fertilisers therefore over- supplies nutrients in available form and fails to replenish soil reserves of fixed nutrients. Furthermore, artificial fertilisers are subject to volatilisation and leaching, which makes them highly inefficient at replacing soil nutrients; these
8 losses should be taken into account in the calculation of the replacement cost, though in practise they are ignored. The actual replacement of large quantities of soil nutrients using artificial fertilisers would result in negative off-site effects; the leaching of nutrients into water supplies could reduce water quality and increase the risk of algal blooms which would impose additional costs.
In reality, farmers are unlikely to apply the quantities of fertiliser required to replace all lost soil nutrients, but will apply the quantities that are economically viable in terms of the response of crop yields. Thus the replacement cost is more acceptable when it is perceived as a theoretical approach to the valuation of soil erosion rather than as a potential course of action. However, the replacement cost does focus on just one of the impacts of erosion on the soils properties and does not necessarily provide an indication of the economic value of soil as a resource.
A number of approaches adopted in the literature for valuation of the effects of erosion on soil properties are reviewed below. It is purely by chance that they all involve use of the replacement cost.
The cost of actually replacing both lost soil and soil nutrients is used by Kim and Dixon (1984) to value erosion in Korea. In the area studied, sediment is eroded from hill land and deposited on paddy land below where it causes damage to rice crops. The valuation of erosion includes the cost of removing sediment from the paddy fields, transporting it by lorry to the hill land and spreading it back onto the land. It also incorporates the cost of the artificial fertilisers required to replace the nutrients lost from the sediment and the labour required to apply the fertilisers. The valuation includes the cost of irrigation required to replace one-third of the runoff and the hill farmers are expected to compensate paddy farmers for the loss in yields that results from the smothering of rice crops with sediment. It is assumed that all of these costs are incurred to hill farmers on an annual basis. The figures used for erosion rates, soil nutrient losses and runoff are based on field data.
A similar approach is adopted by Samarakoon and Abeygunawardena (1992) for a study based in Sri Lanka. The study uses soil erosion rates calculated using the USLE with factors adapted for the conditions in the study area. Two different approaches are used for valuation of the replacement cost of soil erosion. The first approach values erosion in terms of the cost of replacing both the soil and the nutrients lost from crop land. Eroded soil is replaced with soil taken from fallow land (assumed to be four to five miles away), transported by hired truck and spread on the land used for crop production. The difference in nutrient content (nitrogen, phosphorous, potassium and organic matter) between the original top soil on the crop land and the soil from fallow land is valued in terms of the market price of
9 equivalent quantities of artificial fertiliser and organic matter. The valuation includes the cost of the labour required to load, unload and spread the soil but does not include the labour inputs required to apply the fertilisers. The second valuation approach is based on the cost of addressing the effects of soil erosion. Erosion is valued in terms of the cost of ploughing the soil that remains on the crop land and replacing the lost nutrients with artificial fertilisers. It is assumed that erosion has removed all top soil from the crop land, so the nutrient deficit is calculated as the difference in nutrient content between top soil and subsoil.
In contrast to the above static perspective of soil nutrients, van der Pol (1992) examines the impacts of erosion in the context of the dynamic soil system. The study investigates soil erosion in Mali in terms of pools of soil nutrients, their associated inflows and outflows and the resultant effects on soil nutrient balance. Only pools of soil nutrients that are in forms available to plants or that are present in soil organic matter are included. The study examines the following inflows into the soil's nutrient pools:
fertiliser applications, organic manuring, incorporation of crop residues, biological nitrogen fixation by relevant crops, asymbiotic nitrogen fixation, recycling of leached nutrients and biological fixation by trees growing in fields, atmospheric deposition of nutrients in rain and dust, enrichment due to alteration and dissolution of soil minerals, nutrient content of seeds,
and the following outflows of nutrients from the soil's nutrient pools:
nutrient uptake by crops, leaching losses, erosion losses, volatilisation/denitrification losses, irreversible fixation of phosphorous and potassium.
The net soil nutrient balance is given by the outflows from the soil's nutrient pools less the inflows and is calculated for each of the major nutrients: nitrogen, phosphorous and potassium. The results are valued in terms of the market prices of equivalent quantities of artificial fertilisers; the sum of the values for nitrogen, phosphorous and potassium provides the total value of the soil nutrient deficit. The valuation includes the cost of lime that would need to be applied to prevent
10 nitrogenous fertilisers from causing soil acidification. However, the cost of the labour required to apply the fertilisers is not taken into account. The value calculated for soil erosion is used to determine the proportion of farm income that is derived from "soil mining" ie: the ratio between the economic value of nutrient depletion and farm income. It is also used to calculate a sustainability quotient in terms of the proportion of farm income that remains after deduction of the value of soil erosion. The results are used to assess possible approaches for reducing soil erosion in the study area. The data used in the study are obtained from literature based in West Africa and Mali.
A more conceptual approach to the replacement cost of soil erosion is adopted by Stocking (1988). This study, based in Zimbabwe, values erosion solely in terms of the loss of soil nitrogen and phosphorous content, costed at market prices of equivalent quantities of artificial fertilisers. Regression analysis is applied to an extensive set of experimental data from Zimbabwe in order to quantify the relationships between erosion rates and losses in soil nitrogen and phosphorous content. The relationships are assumed to be representative for the whole of Zimbabwe on the grounds that there is little variation in characteristics of soil types in the country. Erosion rates are calculated using SLEMSA for the four main agricultural land uses in Zimbabwe, and the relationships derived in the study are used to estimate erosion-induced losses of nitrogen and phosphorous. The author notes that only a small fraction (4%) of the nitrogen lost would become available to plants in any one year. However, the study examines loss of total nitrogen because decline in unavailable nitrogen is felt to have a significant impact in the long term. In the case of phosphorous, erosion-induced losses are calculated only for the available form. The study calculates the total quantities of nitrogen and phosphorous lost due to soil erosion for the whole of Zimbabwe and estimates the national cost of erosion in terms of the cost of equivalent quantities of locally available artificial fertilisers. Stocking (1988) discusses the problems that can arise with the use of this valuation technique and states that "The financial costing through the equivalent amount of fertilizer...must be seen purely as a financial 'handle' to quantify that impact."
Bishop and Allen (1989) adopt the valuation approach developed by Stocking (1988) for their study in Mali, with the adaptation that loss of only available nitrogen is valued. The study uses linear regression analysis of experimental data from Ibadan (Nigeria) to estimate relationships between erosion and losses of nitrogen, phosphorous and potassium from the soil. Nitrogen is valued based on 4% of estimated total nitrogen loss; this is the proportion of soil nitrogen content assumed to be in available form in any one year (based on Stocking, 1988). Such adjustment is not necessary for phosphorous and potassium as it is assumed that these are lost only in available form. Total losses of nitrogen, phosphorous and
11 potassium are valued in terms of the cost of equivalent quantities of artificial fertilisers. The study also values erosion in terms of reduction in crop yields (Section 2.3.2); comparison of the results indicates that the cost of erosion in terms of soil nutrient loss is significantly lower than the cost of yield loss. The authors attribute this to the exclusion of other impacts of erosion from the nutrient loss calculations, such as reduction in soil moisture capacity and deterioration of soil structure.
In addition to soil and soil nutrient loss, some studies also examine the effects of erosion on other soil properties. Rosenberry et al (1980) value erosion in terms of the impacts on soil properties that are actually experienced by the farmer. For the study area in Iowa, the valuation of erosion includes the cost of fertiliser required to maintain fertility at recommended levels (as indicated by soil analysis) and the value of yield decline that occurs even with increased fertiliser application rates. It also incorporates the cost of additional fuel required to carry out cultivations as a result of the deterioration of soil physical structure.
The study by Pimentel et al (1995) includes the impacts of erosion on runoff, soil moisture capacity and soil fauna. The analysis is based in the US and is carried out for a specific set of environmental conditions (specified in the footnote to Table 1). The long term impacts of erosion (approximated by a time period of 20 years) are broken down and quantified in terms of the effects on runoff, soil nitrogen, phosphorous, potassium and organic matter content, soil moisture capacity and soil fauna. The costs of replacement, the effects on energy requirements and the impacts on crop yields are calculated for each of these factors (see Table 1).
Pimentel et al (1995) value the increase in runoff that results from erosion in terms of the cost of replacing the lost water with ground water irrigation. The replacement of losses in soil nutrient content is valued in terms of equivalent quantities of artificial fertilisers and the cost of replacement of soil depth is simply an estimate. The increases in energy requirements that result from erosion are calculated as the fuel required for irrigation to replace water lost in runoff and the energy required to replace soil nutrients with artificial fertilisers. Based on the figures in Table 1 the study estimates that approximately 10% of the energy used in US agriculture is off-setting the effects of erosion.
12 Table 1: The Long Term Effects of Soil Erosion on Different Soil Characteristics and the Resultant Replacement Costs, Increases in Fuel Requirements and Reductions in Crop Yields a
Factor Quantity lost per annum Replacement cost ($/ha/year) Increase in energy requirements (000s cal/ha/year) Crop yield loss after erosion over a 20 year period (%) Water runoff 75 mm 30 700 7% Nitrogen 50 kg/ha 500 Phosphorous 2 kg/ha 100 3 8% Potassium 410 kg/ha 260 Organic matter 2 t/ha - - 4% Soil depth 1.4 mm 16 - 7% Water holding capacity
0.1 mm
-
-
2% Soil fauna - - - 1% Total on-site effect - 146 1460 20% b
Source: Pimentel et al (1995)
a for a maize crop under conventional tillage on a 5% slope, with 700 mm rainfall, on a loamy soil, 15 cm deep with 4% organic matter and a soil erosion rate (water and wind) of 17 t/ha/yr.
b total reduction in crop yield is not equal to the sum of the reduction in yield for each factor as some of the factors are independent and/or overlap.
2.3 Crop production
Farmers generally experience the effects of erosion in terms of the impact on crop yields. This section examines the different mechanisms by which erosion affects yields and the variability experienced in the extents of its impacts. The effects of erosion on systems of crop production are also discussed. The different approaches adopted in the literature for valuation of the impacts of erosion on crop yields are reviewed. The approaches are grouped according to the source of data employed in the study: field data, data originating from empirical models and data from other literature.
Erosion affects crop yields through the changes it causes in properties of the soil (Section 2.2). The mechanisms by which erosion can affect crop growth include:
13 reduction in soil nutrient content, reduction in soil moisture holding capacity due to loss of soil depth, soil organic matter and clay content, reduction in available soil water due to increase in run off, surface crusting and sealing and greater bulk density of the soil. Reduction in available soil water increases the frequency, duration and intensity of crop drought stress (Lal, 1987), reduction in rooting depth due to decreased topsoil depth and exhumation of impermeable layers in the soil profile e.g. stone lines and ferricrete, reduction in soil water transmission, reduction in soil fauna populations involved in nutrient recycling and soil aeration, reduction in seed germination and the proliferation and development of roots due to degradation of the soils physical structure, increase in the washing away of seeds and fertilisers from the soil surface, reduction in the timeliness of farming operations. Seed bed preparation can be delayed by surface crusting and soil compaction; planting can be delayed by reduction in soil moisture capacity; and variability in soil fertility can cause within-field variations in crop maturity resulting in increased harvest losses (Lal, 1987), increase in the risk of reductions in crop yields due to soil acidification and toxicity.
The above mechanisms are interactive and cumulative in their effects on crop yields. However, the extent to which erosion actually reduces yields is also determined by a number of other factors. Shallow rooting crops such as cereals tend to be more sensitive to the effects of erosion than deep rooted crops such as tubers (El-Swaify, 1993). The crop management system can mask the effects of erosion on yields through increased fertiliser application rates, the adoption of different cultivation practices and the use of irrigation. In general, cropping systems that involve low levels of inputs tend to be more vulnerable to the impacts of erosion on yields (El-Swaify, 1993). The distribution of nutrients in the soil profile (nutrients may be evenly distributed through the profile or concentrated in specific horizons) affects the impact of erosion on yields, as does physical structure of the subsoil and climate, through the influence of rainfall and temperature on crop drought stress.
Measurement of the effects of erosion on crop yields is problematic. Yields are determined by the interaction of a number of factors such as soil type, climate,
14 crop and land management system, which makes it difficult to isolate the effects of erosion. The impacts of erosion can also be masked by the use of fertilisers, irrigation and improved technology. As a result there is a general lack of data on the impacts of erosion on yields and the data that are available are highly site and crop specific. This causes major problems for the valuation of soil erosion; as Boj (1991a) states, "there is no simple relationship between soil loss and production".
Assessment of the impact of erosion on crop production is further complicated by the dynamic nature of agriculture. If yields are falling, farmers are unlikely to continue growing a crop until yields drop to negligible levels. Cultivation of the crop will cease when yields reach levels at which production becomes un- economic. Furthermore, farmers will respond to long term decline in yields by adjusting the levels of inputs used and/or by adapting the cropping system. As erosion proceeds farmers are likely to switch to growing crops that are less sen- sitive to the effects of erosion, that require lower levels of inputs, and/or have less variable yields. There is evidence that such changes in cropping systems have occurred in Java in response to soil erosion (Magrath & Arens, 1989).
Most studies examine the effects of erosion only in terms of the existing crop-ping system and current levels of inputs. Few studies include adaptations made by farmers in response to erosion-induced yield loss. One such study is the economic valuation of soil erosion in Java carried out by Magrath and Arens (1989). The authors account for adaptations in the farmers' cropping systems through inclusion of several different cropping systems for each region in Java (see Table 2). It is assumed that the cropping systems with lowest net returns operate on land that has experienced the greatest loss of soil and is therefore the most degraded.
Despite a lack of data on the effects of erosion on crop yields, many studies value the impacts of erosion in terms of the loss in crop production. A review of the different valuation approaches employed in the literature follows. The studies are grouped according to the source of data: field data, empirical models, and data from the literature.
15 Table 2: The Cropping Systems Examined for Different Regions of Java in the Study by Magrath and Arens (1989)
Region & cropping system Estimated net farm income (Rp/ha) West Java: Cassava, maize, upland rice & legumes. 139,496 Cassava, maize & upland rice 49,531 Cassava 1,279 Central Java: Intercropped maize, cassava & legumes 10,183 Intercropped maize & cassava 6,698 Jogyakarta: Intercropped maize, cassava & legumes 11,279 Intercropped maize & cassava 8,220 East Java: Level land: intercropped maize & cassava 298,327 Level land: cassava 145,005 Terraced hill land: intercropped maize & cassava 58,130 Terraced hill land: cassava 27,806
Source: Magrath and Arens (1989)
2.3.1 Field data Field data used to quantify the effects of erosion on crop yields is generally obtained either from interviews with farmers or from experimental plots. Problems can be encountered with the quality of data from farmer interviews; it is unlikely that farmers can isolate the impacts of erosion on yields from the effects of other factors, such as climate and crop management. Farmers may not be able to accurately recall historical yield data and the data provided by farmers may be biased. The latter can result from farmers perceptions of the interviewer and the manner in which questions are asked.
Anecdotal information collected from farmers on yields can be combined with erosion rates and used to estimate the relationship between erosion and yield. Valds P. (1994) collected data from farmers on crop yields over the preceding six years for a study of two areas in Honduras. The data from each area for scenarios with and without conservation were assessed using linear regression analysis. The resultant equations formalise the decline in yield that occurs due to the effects of erosion and the increase in yield that results from implementation of soil conservation measures (diversion ditches with live barriers). For one of the study areas, linear regression analysis produces relationships with correlation
16 coefficients, R, of 0.73 and 0.77 respectively, indicating a high level of explan- ation. For the other study area the relationships have R of 0.10 and 0.51, which indicates that the equations provide low levels of explanation of the data collected. The author states that this low level of explanation "could be due either to faulty data (since the data depend on farmer recall) or, more likely, to the failure to account for the many other factors that are also influencing yields (for example, weather variations, differences in agroecological conditions within the sample, and changes in practices)."
Valds P. (1994) uses the equations to project crop yields over time for each of the two study areas, for scenarios with and without conservation, and uses the results to assess the economic viability of investment in soil conservation measures. It is assumed that for scenarios without conservation, production ceases when yields fall to a specific level below which production is assumed to be no longer profitable. In scenarios with conservation, the analysis limits yields such that they cannot increase above the maximum sustainable yield that can be achieved in the region.
Another approach that uses farmer recall data is the collection of data on the rates of yield decline perceived by farmers to be occurring as a result of soil erosion. This is the approach used by Cuesta (1994) in the Turrubares area of Costa Rica. The farmers questioned in the study estimated that in the absence of soil conser- vation, cocoa yam yields would fall by as much as one-third per annum. The authors felt that this estimate concurred with characteristics of the soil in the area and their analysis indicated that erosion would result in farmers abandoning cocoa yam production after cultivation of a plot of land for only four years. Despite the severity of this result the study does not examine possible responses of the farmers to declining crop yields, such as adaptations of the cropping system.
Boj (1991a) uses field data on soil depth as indicators of the potential for erosion to have an impact on crop yields. This technique is based on the relationship between soil moisture capacity and soil depth and on the potential for soil moisture capacity to act as a constraint on crop yields. Tables presented by Stocking and Pain (1983) are used to determine the minimum soil depth for different crops and soil types below which soil moisture capacity becomes a constraint on crop yields. Boj determines that the minimum soil depth required for sorghum in the study area in Lesotho is 27 cm. The average depth of soil in the area is estimated as 25 cm, from which it is deduced that erosion affects yields due to its impact on soil moisture capacity. A study that uses data from experimental plots is that by Hernndez (1994), based in the Dominican Republic. Experimental data are available on erosion rates and crop yields for the study area over a period of five years, but with only one
17 observation for each crop and land management system per year. Hernndez discusses the weakness of the data set and states that "Nevertheless, these data are used for lack of any other information on the relation between yields and degradation". Linear regression analysis is applied to the data to estimate the relationships between soil loss and crop yield decline. These relationships are used in conjunction with predicted soil erosion rates, calculated using the USLE with factors modified for Dominica, to project crop yields over the period of the analysis.
2.3.2 Empirical models A number of empirical models are used to estimate the effects of erosion on crop yields. The models range from relatively simple relationships derived from regression analysis to highly complex computer simulations of crop growth, such as the Erosion-Productivity Impact Calculator (EPIC). They are all based on specific assumptions and are generally representative only of the geographical area in which they were developed. Some of the models require small quantities of data that are relatively easy to obtain whereas others require large data sets such as are rarely available in developing countries. The absence of data is addressed in some studies by the use of data from outside the study area. It should, however, be remembered that the results of model-simulations are only as representative as the input data used.
Models to estimate yields should be accompanied by yield thresholds to ensure that simulations remain representative of the cropping system studied (Valds P., 1994). Lower thresholds can be specified such that crop production ceases when yields fall to uneconomic levels. Upper thresholds can be established such that simulated yields cannot exceed those actually achievable by farmers in the study area. Even with the use of thresholds, caution is required when models are used to extrapolate outside the range of the original data from which the models were derived (Hernndez, 1994).
An example of a complex computer simulation model is the Erosion-Productivity Impact Calculator (EPIC) which was developed in the US in the 1980s and has since been widely used. The model simulates interactions between soil, climate and plant growth and examines the impact of erosion on both the physical and chemical properties of the soil to estimate the effects of erosion on crop yields. It comprises eight sub models: hydrology, weather, erosion, nutrients, plant growth, soil temperature, tillage, and economics. EPIC has its disadvantages, summarised by Boj (1991a): What limits the use of EPIC is its great complexity and a need for data that cannot be satisfied in the average developing country context. The model also requires a team of advanced personnel to run them and to sensibly
18 interpret the results. Most studies that employ EPIC are based in the United States; reviews of two such studies follow.
Colacicco et al (1989) use EPIC to calculate the economic value of erosion on a regional basis for the whole of the US. EPIC is used to quantify the effects of erosion on crop land, which are divided into:
permanent loss: the effect of erosion on crop yields through reduction in rooting depth and soil moisture capacity; temporary loss: the increase in fertiliser application rates that results from loss of soil nutrients.
The study uses data on erosion rates from the 1982 US Natural Resources In- ventory. It predicts that, under given assumptions, erosion will have little effect on crop yields in the US over the next 100 years if farmers increase fertiliser applications at the rates simulated by EPIC. Over the 100 year period maize yields are predicted to decline by only 4.6% and wheat yields by 1.6%. Crop and fertiliser prices are applied to the EPIC data to calculate the economic value of erosion. The results are used to identify whether crop yield loss or soil nutrient loss is the major effect of soil erosion in different regions of the US and to identify the areas in which economic loss due to erosion is most concentrated.
EPIC is employed by Benson et al (1989) to study the effects of erosion on soil productivity over a 100 year period and to examine its effect on yields over the soil's productive life. Comparison of the effects of erosion on different soil types in the US is facilitated by the calculation of a soil productivity index: the cumulative yields simulated by EPIC for the different study areas are divided by average national yields for the different crops. The impact of erosion on soil productivity over a 100 year period was found to be insignificant for the soils examined in the study. However, analysis of productivity over entire soil life indicated that the four soil types vary significantly in the rate of productivity decline and in the length of soil life (300 to 4,700 years).
Grohs (1992) uses both the EPIC and the "CERES-maize" computer models to estimate the impact of soil erosion on crop yields in Zimbabwe. CERES-maize is a simulation model that was developed in the 1980s in the US. It assumes that the nutrients required for crop production are available in the soil and do not limit crop growth; the impacts of erosion are determined solely in terms of the reduction in soil moisture capacity (Grohs, 1992). The study uses soil erosion rates calculated using SLEMSA. CERES-maize estimates reductions in maize yields of 0.3 to 1.4% per cm soil loss whereas EPIC calculates yield reductions of 0.7 to 3.3% per cm. The difference in results is attributed to the inclusion in EPIC of the effects of
19 erosion on soil nutrient content as well as on soil moisture capacity. Based on the results of CERES-maize and EPIC and findings in the literature, Grohs decides to use rates of 1, 2 and 3% crop yield decline per cm soil loss for the analysis. These figures are employed to calculate the impact of erosion on farm income on a district level for the whole of Zimbabwe. The analysis is based on the following assumptions:
all crops experience the same rate of yield decline as that estimated for maize, there is no change in the cropping system, the cost of labour for harvest and the cost of transporting produce both decrease in proportion to crop yield, there is a linear relationship between yield decline and loss of income (ie: crop prices are assumed to remain constant).
Less complex models of the relationship between soil loss and crop yield have also been developed. Day et al (1992) use empirical relationships between soil moisture capacity and crop yield to model the effects of erosion on yields in Western Mali. The model is calibrated using data on seasonal rainfall, soil infiltration rates and maximum potential crop yields in the study area. The data used on erosion rates and the effects of erosion on yields are based on Lals data for Nigeria. The cost of the erosion that occurs over a period of one year is calculated in terms of the present value of the resultant losses in yields that occur over the following ten years.
Bishop and Allen (1989) use a simple model developed by Lal in Nigeria to calculate the effects of erosion on crop yields in Mali. The model was developed through multiple regression analysis of experimental data on cowpea and maize crops. It takes the form:
Y = C - bx
where: Y = yield (tons/ha), C = yield on uneroded (newly cleared) land, b = coefficient for crop and slope, x = cumulative soil loss (tons/ha).
The model is used in the study of Mali to estimate the effects of erosion on millet, millet/cowpea, sorghum and maize crops in different areas across the country. The authors note that "Use of the model under Mali conditions is questionable, but certainly more appropriate than anything available from temperate zones, given our data constraints."
20
The region studied in Mali is broken down into areas with different physical and land use characteristics. Erosion rates are calculated using the USLE with factors adapted for conditions in Mali and it is assumed that no soil loss takes place in areas where sediment deposition is known to occur. Total soil loss is calculated for each area over a ten year period and is transformed into crop yield loss using Lals model. It is assumed that in the absence of erosion, initial crop yields would be higher but would also decline over time due to the depletion of soil fertility by crops. The economic value of soil erosion is calculated as the difference in value of crop production between scenarios with and without erosion over a period of ten years. The value of crop production is calculated based on farm budgets developed in Burkina Faso. The analysis accounts for decreases in labour inputs for weeding and harvesting that result from crop yield decline. The study assumes that land is cultivated for a period of ten years and then fallowed for an unspecified period after which soil fertility and crop yields are fully restored.
Cuesta (1994) calculates the effect of soil erosion on coffee yields using a model of crop yield response to nitrogen loss. A representative erosion rate (0.8 mm/year) is calculated for the study area of Heredia (Costa Rica) using the USLE with factors adapted for local conditions. The author estimates that soils in the study area loose 4.32 kg of available nitrogen per mm of soil loss. Research indicates that, for the coffee yields prevailing in the area (11,348 kg/ha), crop response to nitrogen fertiliser is 35.9 kg coffee per kg nitrogen fertiliser. Based on these data it is calculated that coffee yields decline at a rate of 155 kg per mm soil loss; soil erosion in the study area is therefore estimated to reduce coffee yields by 124 kg per year.
2.3.3 Data from the literature Some studies rely on other literature for data on relationships between crop yields and soil loss. Such data may be drawn from the findings of experimental research or from other economic analyses using secondary data.
For a study based in India, Magrath (1990a) uses experimental data on the impact of artificial soil removal on maize production to estimate erosion-induced losses in yields. The author discusses the inadequacies of the data; the experimental removal of soil tends to underestimate the effects of erosion on crop yields. The data on yield decline are adjusted to represent the erosion rate estimated for the study area, giving rates of erosion-induced yield decline of 1.25% per year for the first five years of cultivation and 0.95% per year in subsequent years. Magrath comments that yield decline would be expected to increase with time "as sheet erosion gives way to rill erosion leading to larger soil losses". However, such
21 processes were not examined in the experimental research used as the basis for the analysis.
Magrath and Arens (1989) also use experimental data from other studies for their economic valuation of erosion for the whole of Java. The crops grown in Java are divided into those that are relatively sensitive to the effects of erosion (maize, soya beans, groundnuts, green beans and upland rice) and those that are relatively insensitive (cassava). A geographical information system (GIS) is used to identify 25 different categories of land in Java, based on soil type and slope. The relationship between yield and erosion rate for each of the land categories and for both types of crop are estimated using experimental data from other studies. Erosion rates are estimated for each of the four main cropping systems on each land category, based on previous studies and judgement of local conditions. Total loss of crop production due to soil erosion is calculated for the whole country, assuming either the production of only sensitive crops or of insensitive crops. The results indicate that for the whole of Java, erosion causes an average decline in the production of sensitive crops of 6.7% per annum and 4.2% per annum for insensitive crops.
An alternative approach is to use relationships between yields and erosion rates already derived from experimental data in other studies. Mitchell et al (1980) use crop yield index data published by the Illinois Cooperative Extension Service for their study based in Illinois, US. The crop yield index provides data on per-centage yield for land with different categories of slope, favourable or unfavour-able subsoil, different categories of erosion rate, and different levels of crop management. Mitchell et al calculate prevailing erosion rates with the USLE and use crop index data to assess the impact of erosion on yield for land with different combinations of characteristics.
Boj (1991a) draws together data on relationships between erosion rates and yield decline from seven different studies based in Lesotho, Ethiopia, India, El-Salvador and the US, the majority of which examine the effects of erosion on maize and sorghum crops. To facilitate comparison Boj converts all the data into one common unit, percentage relative loss in yield per millimetre soil loss. The results from the different studies range from 0.001 to 0.8 % and are given in full in Table 3. Boj summarises these figures by rounding them up to 1% "In order not to exaggerate the degree of exactness". This is then used as the rate of yield decline (as a proportion of the previous year's crop) for both maize and sorghum in the area studied in Lesotho by Boj. Table 3: The Relationships Between Crop Yield Decline and Soil Erosion Rates in the Literature Examined by Boj for a Study based in Lesotho
22
Literature source Location of study % Relative yield loss per mm soil loss Hurni (1985) Ethiopia 0.2 Lal (1981) Nigeria 3.9-20.8 Lockeretz (1981) US 0.1 Lyles (1975) US 0.2 Nobe & Seckler (1979) Lesotho 0.001-0.5 Swanson & Harshbarger (1964) US 0.1-0.2 Vittal et al (1990) India 0.8 Wiggins (1981) El Salvador 0.2
Source: Boj, 1991a.
N.B. the data for Nigeria (Lal, 1981) is not used by Boj (1991a) as the conditions are not considered to be relevant to Lesotho.
2.4 Other valuation methods
Contingent valuation and hedonic pricing are alternative methods for valuing the on-site effects of soil erosion. Contingent valuation involves questioning individuals (such as farmers) about their willingness to accept compensation for the effects of erosion or their willingness to pay for the benefits of reduced erosion e.g. as a result of soil conservation. This technique can also be used to assess the value of the off-site effects of erosion and conservation.
Hedonic pricing uses land prices to estimate the economic value of soil erosion. Sale prices or rental charges for plots of land that experience different rates of soil erosion are assessed using regression analysis. The results quantify the relationship between land prices and soil loss and so can be used to infer the cost of erosion. Several such studies have been carried out in the US. For example, Miranowski and Hammes (1984) examine the effects of both depth of topsoil and potential soil erodibility on land prices in Iowa and are able to estimate costs of each of these factors.
There are, however, a number of problems that arise in the valuation of soil erosion with hedonic pricing. The use of land prices to estimate the cost of erosion requires data on land prices and a well developed market for agricultural land. In most developing countries neither of these conditions are met. Hedonic pricing assumes that individuals take into account erosion-induced degradation in making decisions whether to purchase or rent a piece of land. In many cases individuals are unaware of the extent of degradation due to insufficient information or the
23 masking effects of technology. Individuals may not even be concerned with the quality of the land, especially when making purchases on the basis of investment speculation or for reasons of personal security or personal status.
24 3. Valuation of the On-Site Effects of Soil Conservation
There are many different soil conservation technologies employed around the world. Some are well recognised and have formed the basis of much of the research on soil conservation whilst others are less well known and are adaptations by farmers to their local environmental conditions. Each conser-vation technology is suitable for certain characteristics of land (slope, soil type, availability of stone), climate and farming system.
Soil conservation technologies comprise of conservation measures and conservation practices. On-site soil conservation measures can be divided into physical measures such as earth and stone bunds, contour drains and terraces, and biological measures such as contour grass strips, alley cropping and contour hedgerows. Conservation practices are cultivation practices that promote soil conservation, such as intercropping, under sowing with cover crops, minimum tillage, contour planting and surface mulching. Conservation measures can also be constructed off-site to reduce the loss of soil on a catchment scale. Such measures include check dams, silt traps, diversion drains and gully control structures. Further details on soil conservation technologies are given in Morgan (1986).
On-site soil conservation technologies reduce the effects of erosion through a number of different mechanisms. These include:
reduction in field length, which reduces the rate of overland flow (e.g. contour bunds), reduction in slope (e.g. terraces), reduction in the rate of overland flow, which increases rainfall infiltration (e.g. contour planting), removal of excess surface runoff (e.g. contour drains, diversion ditches), increase in rainfall infiltration, resulting from reduction in overland flow, improvement in soil structure and increased soil organic matter (e.g. terraces, organic manuring), reduction in the impact of precipitation on the soil surface (e.g. cover crops), binding of the soil surface to reduce soil particle displacement (e.g. conservation tillage), reduction in the temperature of the soil surface (e.g. mulching), trapping of eroded sediment (e.g. contour hedges).
The impacts experienced by farmers from the implementation of conservation technologies vary with different technologies, environmental conditions and farming systems. The general assumption is that farmers experience an increase in
25 crop yields. This may occur as a direct result reduced soil erosion, or indirectly as a result of increased soil moisture capacity, improved soil nutrient content, and/or trapped fertile sediment transported by overland flow. However, yields can also decrease as a result of implementing soil conservation. Excessive increases in soil moisture capacity can cause water logging and reduce yields; the construction of terraces may bring subsoil to the surface and decrease yields. The levels of inputs employed in crop production may increase or decrease according to the conservation technology adopted. The farmer may experience a loss in crop area due to the occupation of land by conservation measures (e.g. contour bunds) and the adoption of some conservation technologies may increase damage to the crops e.g. vegetative strips can harbour pests such as rats.
Some conservation technologies increase the efficiency of agricultural production through, for example, increased returns per unit of labour, seed or fertiliser. Additional benefits of conservation can also include increased supply of fuelwood (from agroforestry), animal fodder (from grass strips) and increased nutritional value of food production (from fruit trees). In summary, the impacts of conservation on agricultural production are complex and highly situation specific.
Studies of the economic benefits of soil conservation run into problems of the dynamism of cropping systems discussed in Section 2.3. Increases in yield resulting from soil conservation may prompt farmers to alter the levels of inputs used for crop production. They may also prompt farmers to shift to growing crops that are more valuable, require greater expenditure on inputs, and/or have greater variability in yield. An example of such a response is given by Barbier (1990) in Java where the construction of terraces resulted in the cultivation of more valuable crops. For the sake of simplicity though, most studies assume that the adoption of soil conservation does not cause any change in the cropping system or the levels of inputs used.
A general problem encountered in the economic analysis of soil conservation technologies is quantification of the benefits. Boj (1986b) states that Quanti- fication of physical effects (such as difference in crop yields) of conservation projects is often a weak spot in available studies. Scant empirical evidence, sometimes collected with unclear or questionable methods play a distressingly large role. Experimental data may be available on the impacts of conservation on soil loss or crop yields, but results tend to be site specific. Alternatively, empirical models or data from other studies may be used to estimate the benefits of conservation.
The following sections review the approaches adopted in the literature for valuation of the on-site effects of soil conservation. Approaches that value the
26 benefits in terms of the impacts on soil properties are examined, followed by those that value the effects on crop production. Valuations of other impacts of soil conservation are then reviewed including the impacts of conservation on: the area under crop production; the costs of production; the nutritional value of crop production; and the effects of conservation on livestock production, land-use value and environmental quality. Use of contingent valuation is also examined.
3.1 Soil properties
Few studies value soil conservation in terms of its impacts on soil properties. However, soil conservation technologies are evaluated in terms of their ability to reduce erosion rates to target levels, such as soil loss tolerance values (t- values). In most studies the t-value is based on the maximum erosion rate that permits a high level of crop productivity to be sustained economically and indefinitely. In order to reflect this the t-value tends to be equated to the estimated rate of soil formation. There is, however, considerable debate about the basis for determining soil loss tolerance values and for further details the reader is referred to Lal (1987) and Morgan (1986).
3.2 Crop yields
The effects of soil conservation on crop yields can be quantified using field data, empirical models, data from other studies, or data that are estimates or hypothetical. A review of the valuation approaches adopted by a selection of studies follows, with the use of each different data source examined in turn.
3.2.1 Field Data Lindgren (1988) uses survey data collected over three seasons to quantify the impact of soil conservation on crop yields for a project in Kenya. Differences in yield are calculated for paired plots of land, terraced and unterraced. The plots in each pair are located in the vicinity of each other, to minimise differences in agro- ecological conditions. Crop yields are weighted for the probability of recorded rainfall in order to take into account variations in rainfall between the three seasons. The weighting is calculated from statistics on rainfall for the wet seasons in each of the twelve years prior to the survey. The differences in weighted crop yields between terraced and unterraced land are given in Table 4.
27 Table 4: The Average Difference in Crop Yields between Terraced and Unterraced Land, Weighted for Rainfall, Machakos District, Kenya.
Crop Average yield on unterraced land (kg/ha) for short and long rains Weighted average difference in yield between terraced and unterraced land (kg/ha) 95% Confidence interval (kg/ha)
Lindgren (1988) assumes that on unterraced land crop yields decline at a rate of 1% per annum as a result of soil loss and poor management. This figure is based on the work carried out by Boj in Lesotho and expert opinion in Kenya. The benefits of terraces are calculated in terms of the increase in crop yield plus the loss of yield that is prevented.
McLaughlin (1993) examines the effects of soil conservation on crop yields through the comparison of three different watersheds in Gansu Province, China. Data are available on crop production and income levels for two villages (Table 5). One of the villages, Xujiamen, has no soil conservation and the other, Daying, implemented soil conservation measures (mostly biological) half way through the period of data collection.
McLaughlin (1993) suggests that soil conservation in the study area has its greatest impact on crop yields through increasing soil moisture capacity. Linear regression analysis of data for the two villages indicates the following relationship between soil conservation and village income:
y = 0.354x 1 + 0.985x 2 - 0.283x 3 (R = 0.83) where: y = standardised village income (10,000s yuan) x 1 = annual rainfall (mm) x 2 = % land area terraced x 3 = % land area under biological conservation measures
28 The economic value of grass is omitted from this analysis, resulting in the negative coefficient for land area under biological conservation measures.
Table 5: Agricultural Production and Income in the Villages of Daying and Xujiamen in Gansu Province, China Before and After the Implementation of Biological Conservation Measures in Daying
Agricultural production (kg/mu/yr) Total village income a
Source: based on McLaughlin (1993) a total income from agriculture, animal husbandry and forestry-related income. N.B. the author notes that it is possible that the increase in the level and stability of income may result from a restructuring of the village economy which could have accompanied the implementation of soil conservation.
In addition analysis is carried out from a social perspective to examine the effects of soil conservation on the standard of living. This takes into account the economic value, as fuel and fodder, of the grass grown in biological conservation measures. The standard of living index used is calculated as the sum of the proportions of annual per capita food, fuel and fodder requirements produced within a village. The effect of soil conservation on the standard of living index is examined through
29 linear regression of data from a watershed employing mostly physical, but also biological, conservation measures (McLaughlin, 1993):
y = 0.370x 2 + 0.537x 3 (R = 0.82) where: y = standardised standard of living index x 2 = % land area terraced x 3 = % land area under biological conservation measures
The results of this analysis indicate that mechanical and more so, biological soil conservation measures have positive effects on the standard of living in the study area. The study illustrates the importance of taking into account the value of all the benefits of soil conservation, not just those that are valued in the market place.
3.2.2 Empirical models A number of different models are used to assess the impacts of soil conservation on crop yields. Applications of several of these models are outlined below.
The USLE is used by Day et al (1992) to calculate the effects on erosion rates in Mali of different combinations of soil conservation technologies and crop management strategies. Erosion rates are calculated for the different combi-nations by multiplying the maximum potential erosion rate by factors for soil physical structure and crop cover (adjusted for West African conditions). The erosion rates are entered into a soil-water balance/crop yield response model (described in Section 2.3.2) to calculate the impact on crop yields. The yield data are entered into a linear programming model of a typical farming system which maximises net cash income subject to resource constraints and the fulfilment of household requirements for grain consumption. Production is allocated between cultivation of three different groups of crops: grains (maize, millet and sorghum), groundnuts and vegetables.
Cuesta (1994) uses the USLE with factors adapted for conditions in Costa Rica to estimate the effect of diversion ditches on the erosion rate in the Heredia area. The decrease in soil loss resulting from the conservation measures is converted into a reduction in yield loss from coffee crops, based on crop nitrogen response data described in Section 2.3.2. The analysis takes into account the crop area lost to soil conservation measures.
EPIC is used by Benson et al (1989) to calculate the effects of soil conservation on crop yields in the US. The study examines three different farming practices: conservation tillage plus conservation practices; conservation tillage alone; and conventional tillage. The impacts of the three practises on the soil productivity
30 index (described in Section 2.3.2) are calculated; the results indicate that the practices all have little effect on soil productivity over a period of 100 years. However, the practices differ significantly in their impacts over the entire productive life of soils. For three out of the four soil types studied the potential productivity achievable over the soil's life with conservation tillage and conser- vation practices was double that which could be achieved with conventional tillage. Finally erosion rates for different soil conservation practices are calculated using EPIC and the ability of the practices to achieve tolerable levels of soil loss (derived from the soil productivity index) is assessed.
The "SOILEC" computer simulation model employs the USLE to calculate erosion rates and quantifies and values resultant declines in crop yields. The model can be used to carry out both long and short-term analysis. The former takes into account the long-term effects of erosion on yields and the latter takes into account only the immediate effects of erosion and takes the form of a financial analysis. Dumsday and Seitz (1985) use SOILEC to value the effects of different crop management systems on erosion-induced yield decline in Illinois, US. The study examines 54 management systems that comprise different combinations of crop rotations, tillage systems and mechanical soil conservation measures on each of 16 different soil types in Illinois.
For each soil type Dumsday and Seitz (1985) identify the most economic management systems that will reduce soil loss to nominal target levels. The costs of a unit reduction in soil erosion are compared by plotting costs per hectare per year for the different management systems against the resultant reductions in soil loss per hectare per year. From this it is possible to identify the management systems that have the lowest costs per unit reduction in soil loss and that also meet the target reductions in erosion. The study also examines the marginal costs of using the most economic management systems to achieve different target reductions in soil erosion. It is proposed that it is these marginal costs that should be used to determine the subsidy payments made to farmers for the implementation of soil conservation technologies. Dumsday and Seitz (1985) state that "Consciously or unconsciously, it is these additional costs of moving from one system to another that influence the behaviour of farm decision-makers. The average costs of erosion reduction ... are not relevant in this respect." Lu and Stocking (1994) use a combination of several empirical models to calculate the effects of different soil conservation measures on yields in China. SLEMSA is used to calculate soil erosion rates which are used as inputs to two sub models of soil properties. The sub model of soil organic matter balance is based on SCUAF (Soil Changes Under Agroforestry) and THEPROM (Theo-retical Erosion Productivity Model). A second sub model examines the soils mineral balance. The results generated for the erosion rates and the resultant changes in soil organic
31 matter content and soil mineral levels are used to calculate the effects of erosion on soil available water storage capacity (AWSC). The study uses AWSC as an index of soil productivity; an empirical model is used to calculate the effect of erosion- induced decline in AWSC on biomass production. The calculations are repeated for different soil conservation measures to determine their impacts on soil erosion rates and hence crop biomass production.
The "EROPLAN" computer simulation model is used by Prato and Wu (1991) to calculate the effects of erosion on net returns to crop production in Idaho, US. Erosion rates are calculated using the USLE for each field in the study area. EROPLAN is used to carry out simulations of nine crop rotations grown under seven different tillage/land management systems. The models output is in the form of annualised net returns to crop production adjusted for the reductions in returns that result from soil erosion. The study assesses the extent to which different combinations of crop rotation and tillage/management system achieve target levels of soil erosion and the effects that these have on net returns.
3.2.3 Data from the literature The use of data sourced from other literature is illustrated by just one study. Magrath (1990a) estimates the effects of conservation measures on crop yields through the use of data from other studies. Vetiver grass strips combined with contour planting are estimated to increase crop yields by 50% over a 30 year period, based on Greenfield's work in India. The study also assumes that contour bunds increase yields by 30% over a 30 year period, based on Tejwani's work in India.
3.2.4 Estimates A number of studies simply estimate or make assumptions about the impacts of soil conservation technologies on crop yields. One such approach is to assume that erosion-induced reductions in yields completely cease with the implement-ation of soil conservation. The validity of this assumption is highly dependent on the conservation technology and specifically its impact on yields. Bojs study in Lesotho (1991a) assumes that yields remain constant on land with terracing and drains, whereas on land without conservation yields decline at a rate of 1% of the previous years yield (see Section 2.3.3). The study calculates the benefits of conservation in terms of the losses in crop yields that are prevented. These are valued using the market price of grain and the animal fodder value of crop residues; the latter is calculated in terms of the residues protein and energy content, costed at market prices of substitutes.
Some studies assume that the reduction in yield loss that results from soil conservation is directly proportional to the reduction in soil loss. This is the
32 approach adopted by Cuesta (1994) in a study of the Turrabares area of Costa Rica. The analysis indicates that the construction of diversion ditches increases the period over which continuous cultivation can be sustained to seven years, in contrast to a period of four years on land without conservation. The construction of bench terraces is found to extend the period of cultivation on a single plot of land to 52 years.
Attavirojs study (1990) stipulates that if the construction of soil conservation measures (e.g. conservation banks) is not accompanied by the adoption of conservation farming practices (e.g. contour ploughing, mulching) then it has no effect on crop yield decline. The land development programme examined in Northern Thailand involves only the construction and maintenance of contour banks and access tracks. Attaviroj states that the programme "seeks only to control soil erosion, not soil degradation" and that "farmers continue to exploit the land". As a result, crop yield decline is assumed to occur at the same rate on land with conservation measures as it does on land without. Attaviroj states that soil moisture capacity can be a major seasonal constraint on crop production in the area but it is unclear whether the soil conservation measures reduce its effects on yields.
The benefits of a soil conservation package are calculated by Mwakalagho (1986) in terms of the resultant one-third increase in crop returns. The study of a project in Malawi examines a soil conservation package involving the construction of soil conservation measures and the improvement of yields through use of fertiliser and improved seed. It is assumed that only one-third of the increase in yields that results from the package can be attributed to soil conservation. Gross margins are calculated for the periods before and after the project for each of the three main crops grown in the study area. The benefits of soil conservation are calculated as one-third of the difference between the gross margins for the two periods, averaged for all crops.
Wiggins and Palma (1980) break down the benefits of soil conservation into three different components for the purposes of valuation:
the erosion-induced decline in yields that is prevented by conservation. Two models of crop yield decline are used: in the situation where top soil is still present, yields are assumed to decline by a constant rate per cm soil loss. In the situation where no top soil remains (ie: there is subsoil at the surface) yields are assumed to decline at a negligible rate until the soil reaches a trigger depth and then yields decline rapidly to zero; the initial increase in yields that results from the implementation of conservation measures. This is attributed to the reduction in erosion-
33 induced loss of fertiliser, the effects of mulching, increased soil moisture capacity and increased infiltration; a gradual long term increase in yields that results from top soil formation, facilitated by the reduction in soil erosion.
The study examines the implementation of a soil conservation package compris- ing hillside ditches and live barriers in an area of El-Salvador. The land in the study area is divided into six different categories according to slope and the presence of topsoil or subsoil at the surface. Erosion rates are calculated using the USLE with factors adapted for local conditions and the impacts of conser-vation are estimated using models developed for each land category. Based on experimental data from the US it is estimated that in the absence of soil conser- vation maize yields decline at a rate of 2% per centimetre loss of topsoil. The initial increase in yields arising from implementation of conservation measures is estimated as 10% based on expert opinion. The results of the study are used to identify the categories of land in the study area that should be prioritised for implementation of soil conservation.
3.2.5 Hypothetical data In the absence of other data, the effects of erosion and conservation on crop yields can be estimated using hypothetical data. Indeed, hypothetical data may be used in preference to tenuous data from other sources. Use of hypothetical data should, however, always be accompanied by sensitivity analysis to assess stability of the results with regard to the assumptions made. This is the approach adopted by Obando E. and Montlvan (1994) in the economic analysis of soil conservation in Nicaragua. The study examines seven scenarios comprising different combinations of hypothetical rates of yield decline without conservation and hypothetical rates of yield increase with conservation (see Table 6). The analysis assumes that crop yields cannot exceed the maximum yields achievable in the study area under good management. The construction of conservation measures (diversion ditches with stone walls) is assumed to reduce crop area by 10%. The net present values calculated for the different combinations of hypothetical data cover a wide range (- 2798 to 3,776 crdobas/ha) (Table 6). This illustrates the effects of the values of variables and the accuracy of data (specifically the impacts of erosion and conservation on crop yields) on the economic viability of soil conservation. The authors state that "Without such accurate information, it is impossible to say whether conservation is profitable or not from the individual farmer's perspective."
Table 6: Hypothetical Scenarios of Changes in Bean Yields with and without Soil Conservation used by Obando E. and Montlvan
34 (1994) in the Economic Analysis of a Soil Conservation Project in Nicaragua
Percentage change in yield per annum a
Net present value of soil conservation Scenario Without conservation With conservation b (crdobas/ha) c
Base 0 0 -2,798 A -1% 0 -2,108 B -5% 0 51 C 0 +1% -2,112 D 0 +5% 928 E -1% +1% -1,423 F -5% +5% 3,776
Source: Obando E. and Montlvan (1994)
a for bean crops planted on land with 20% slope in the Santa Luca area, Nicaragua. b where conservation takes the form of diversion ditches with stone barriers. c calculated over a period of 100 years using a 20% discount rate.
The following sections examine approaches adopted in the literature to assess impacts of soil conservation other than those on soil properties and crop yields. Approaches adopted for valuation of the effects of conservation on land area under crop production, the costs of production, nutritional value of crop production, livestock production, land use value and environmental value are examined in turn.
3.3 Area under cultivation
The construction of physical soil conservation measures can result in a decrease in available crop area, though if the risers are also planted with crops, the construction of terraces can actually increase crop area. Some studies take into account the reduction in crop area that results from the implementation of physical soil conservation measures whereas others do not.
The study by Attaviroj (1990) based in Northern Thailand assumes that the only benefit of the construction of soil conservation measures is an increase in crop area. In the absence of soil conservation the farmers in the study area leave half
35 their land fallow each year. With the construction and continued maintenance of conservation measures (such as contour banks) it is assumed that land can be continuously cultivated without any increase in input levels, effectively increasing the crop area.
3.4 Costs of crop production
Implementation of soil conservation measures can have a number of effects on the costs of crop production. For example, construction of conservation measures can reduce crop area, decreasing the level of inputs required; some studies take this into account whilst others do not. The following two studies both examine reduction in labour inputs as the major benefit of soil conservation.
Cuesta (1994) identifies the major benefit of soil conservation in the Cartogo area, Costa Rica, as the reduction in labour inputs required for crop production. The major costs of erosion are identified as the increases in labour inputs required to remove rocks exposed on the soil surface and to apply greater quantities of fertiliser. Erosion exposes the less fertile B horizon of the soil and washes away artificial fertilisers, making it necessary for farmers to increase fertiliser application rates. It is assumed that the labour costs incurred by erosion decrease in proportion to the soil erosion rate. The data on labour and fertiliser costs required for the analysis were collected from a sample of six farmers and the results are adjusted for crop area lost to conservation measures.
The benefits of an agroforestry project in Panama are examined by Vsquez and Santamara (1994) in terms of the reduction in land clearing costs. The agri- culture in the area is migratory; farmers clear land, cultivate it for one year and then leave it fallow for five years. Farmers that had adopted the projects conser- vation package estimated that it would permit them to continuously cultivate land for a period of twenty years. The benefits of soil conservation are valued solely in terms of the reduction in land clearing costs to the farmer; it is assumed that the conservation package has no effect on crop yields. In the absence of conservation, farmers incur the cost of clearing land every year whereas farmers with conservation clear land only once every 20 years. In the latter case the farmers have the additional costs of the construction (every 20 years) and maintenance of the conservation measures.
3.5 Nutritional value of crop production
36 Soil conservation projects can change the crops produced by farmers. As a result they may affect the nutritional value of crop production as well as the level of crop yields. This is illustrated by a project studied by Holmberg (1990) in Kenya, which is estimated to increase food crop production by 45%. Through the promotion of crops with higher nutritional value the project effectively increased food production by an additional 5%.
3.6 Livestock production
The Eppalock Catchment Soil Conservation Project in Victoria, Australia, was primarily targeted at pasture improvement; its on-site benefits are quantified by Abelson (1979) in terms of the rise in returns to livestock production. The increases in stocking rates that result from the project are calculated from survey data for different soil/rainfall zones in the area. It is estimated that in the absence of the project, stocking rates would have increased but by a smaller magnitude. The impact of the project is calculated as the difference between the increases in stocking rates with and without the project. The project also resulted in an increase in the proportion of livestock that are cattle as opposed to sheep and in greater wool production per sheep, both of which are taken into account in the analysis.
The adoption of some soil conservation technologies on land under crop production can have additional benefits for livestock production. The "fanya juu" terraces examined by Holmberg (1990) in Kenya are planted with fodder grass along their top edges. The fodder grass is used by farmers to increase livestock production and this is taken into account in the economic analysis.
3.7 Land-use value
The benefits of the soil conservation package promoted by the Phewa Tal Watershed Management Project in Nepal are calculated in terms of their effects on "land-use value" (Fleming, 1983 & Hufschmidt et al, 1983). The latter represents the total economic value of production on the land. The economic analysis values the increase in production in the project area in terms of the rise in land-use value. The project aims to conserve soil through the establishment and protection of pasture, hay fields and scrub. Grass and tree foliage are assumed to be used as animal fodder and are valued in terms of the market price of milk and the market prices of the nitrogen and phosphorous fertiliser contents of the dung. The
37 fuelwood produced is valued in terms of its substitute value. The land-use value is calculated for scenarios both with and without the project and the difference between them is the value of the benefits of the project.
3.8 Environmental quality
In addition to the benefits of increased livestock production (Section 3.6), Abelson (1979) calculates the environmental value of the soil conservation project in Australia. The environmental value is calculated as the difference in average prices of land improved under the project (Australian $ 375/ha) and unimproved land ($ 150/ha). The difference in prices ($ 225/ha) is broken down into the benefits of increased productivity ($ 175/ha) and a residual $ 50/ha attributed to the value of environmental improvements. It is assumed that for the scenario with the project, the environmental value of the area would increase linearly from zero in 1960 to $ 50/ha in 1975. For the scenario without the project, land improvements would have taken place more slowly and the environmental value would not have risen to $ 50/ha until 2010.
3.9 Contingent valuation
Contingent valuation is adopted by Purvis et al (1989) for valuation of the on-site effects of filter strips under a conservation programme in Michigan, US. Farmers were questioned about their willingness to accept compensation in return for the establishment of filter strips on their land. The aim of the filter strips was to reduce the sediment washed off the land and into streams. Filter strips have few benefits to the farmer (access, amenity, recreation); the farmer bears the cost of a reduction in crop area and, under the given programme, 50% of the costs of establishment. The data for the analysis were collected using postal question- naires. The sample of farmers was divided into twelve groups, each of which was offered a different annual compensatory payment in return for the establishment of filter strips. Data were collected on the willingness of farmers to accept the payment offered and, at the given level of payment, the area of land they would be willing to place under the programme. Tobit analysis was carried out on the data and the resultant equation was used to estimate the area of land that filter strips could be established on for different levels of compensatory payment.
In summary, valuation of the on-site effects of erosion and conservation can be carried out using a number of different approaches. These are generally based on the impacts of erosion and conservation on soil properties or crop yields but may also include other factors, such as crop area and costs of production. The valuation
38 of the off-site effects of erosion and conservation involves additional complications and has received less attention in the literature.
39 4. The Off-Site Effects of Soil Erosion and Conservation
Erosion can have a number of effects downstream from the land on which soil loss occurs. The impacts can be positive as well as negative and are described as off- site effects. This section provides an overview of some of the off-site effects that can be experienced as a result of soil erosion and discusses problems that can arise with quantifying them. It then briefly reviews approaches adopted in the literature for valuation of a number of off-site effects of erosion and conservation.
The off-site effects of soil erosion include:
increase in risk of reduced crop yields due to soil salinisation and toxicity, caused by the ground water rise and sedimentation that result from increased run off, the smothering of crops by sediment, reduction in crop yields due to water logging of low lying land caused by increased runoff, provision of fertile sediment on flood plains which may increase yields or may smother crops,
damage to structures such as roads, bridges, terraces, and the foundations of buildings, provision of sand for use as a building material, reduction in the navigability of waterways and harbours,
reduction in reservoir storage volume which results in: - reduction in water supplies for irrigation, - reduction in hydro-electricity generation capacity, - damage to the equipment used for hydro-electricity generation, - reduction in flood control, - reduction in dry season flows for drinking water and industrial use,
reduction in the efficiency of irrigation systems, damage to the equipment used in irrigation systems, increased costs of drainage system maintenance,
reductions in the quality of water supplies for drinking water and industrial use as a result of: - increased concentrations of agrochemicals, metals and salts in the water, - increased water turbidity due to its sediment content,
40 - increased risk of algal blooms as a result of water nutrient enrich-ment; algal blooms reduce the quality of water for consumption and reduce the aesthetic and recreational value of water resources.
reduction in the diversity and size of biological populations in rivers, lakes and reservoirs, reduction in the recreation and amenity value of water resources, damage to fisheries, inland and coastal, damage to coral reefs.
The valuation of the off-site effects of soil erosion and conservation is problematic. In many cases it is difficult to quantify let alone value the impacts of sediment on, for example, downstream crops and fisheries. Furthermore, it is hard to isolate the effects of soil lost from agricultural land from those of other sources of sediment such as landslides, river bank erosion, and erosion caused by construction projects.
The benefits of erosion experienced off-site are rarely discussed in the literature, even though eroded sediment is often employed beneficially by entrepreneurs down stream. Sediment can provide farmers with a source of fertile nutrient-rich soil, for example on bunded paddy land. Traps may even be constructed by farmers to create new fields through the accumulation of sediment, such as the rock check dams constructed by farmers in Karnataka State, India (Prem Kumar, 1994). Also the sand content of eroded sediment may also be deposited in rivers, from which it can be removed for use in the building trade, such as in the Mahaweli Catchment, Sri Lanka.
In terms of the off-site costs of erosion, difficulties arise in quantifying the effects of reservoir sedimentation on, for example, hydro-electricity generation and water supplies for irrigation. The storage capacity of reservoirs is divided into live and dead storage capacity. The dead storage capacity lies at the bottom of the reservoir and the live storage capacity on top; the latter supplies the water for hydro- electricity generation and irrigation. Sediment deposition can occur in both the live and the dead storage capacity and its distribution between the two is specific to the characteristics of the individual reservoir. This makes it difficult to determine the extent to which sedimentation affects hydro-electricity generation capacity and water supplies for irrigation. To overcome this, studies make assumptions about the relationship between reservoir water yield and loss of live or total reservoir capacity due to sedimentation.
Despite the numerous downstream effects of soil erosion, few studies actually incorporate them into the valuation of erosion and conservation. This is because of
41 the difficulties encountered in quantifying and valuing off-site effects and the complications that arise when the study area does not cover an entire catchment or sub-catchment. Those studies that do examine downstream effects focus primarily on the impacts of reservoir sedimentation on hydro-electricity generation capacity and water supplies for irrigation. The following sections examine the approaches adopted for the valuation of the effects of erosion and conservation on hydro- electricity generation, irrigation, navigation, water quality, deforestation and road maintenance.
4.1 Hydro-electricity generation
The approaches used for valuation of the effects of erosion on hydro-electricity generation differ primarily in terms of the type of reservoir storage examined (live storage or total storage) and the basis for valuation of a unit of electricity.
Gregersen et al (1987) quote a USAID study based in the Panama watershed. It values the effect of sedimentation on hydro-electricity generation in terms of loss in live reservoir storage capacity. A one acre-foot of live storage capacity is estimated to generate 130 KW of power. The least cost alternative method for generating the electricity is using fossil fuels, at a cost of $ 0.5/KW. The value of loss of live reservoir storage capacity due to sedimentation is calculated as $ 6.50 per foot acre or $ 4,975 per million m.
For their study of Java, Magrath and Arens (1989) examine the impact of sedimentation on reservoir water yield using ranges of figures. The authors assume that the actual impact of sedimentation on water yield lays between the impacts of sedimentation causing loss of total reservoir storage capacity and loss in only dead storage capacity. For the nine major reservoirs in Java examined in the study, it is estimated that dead storage capacity accounts for 20% of total capacity. The results indicate that for all of the reservoirs, sedimentation results in an average 0.5% loss in total reservoir storage capacity per annum and a 2.3% loss in dead storage capacity per annum. Based on this it is estimated that reservoir sedimentation reduces hydro-electricity generation capacity by between 0.5 and 2.3% per annum. The loss in generation capacity is valued in terms of the average of the on-peak and off-peak prices for electricity in Java.
Wiggins and Palma (1980) assume that there is a linear relationship between reservoir sedimentation and hydro-electricity generation capacity for their study in El-Salvador. The relationship is based on the estimated total loss of hydro- electricity generating capacity that would occur with complete siltation of the Cerrn Grande reservoir. It is assumed that this siltation would reduce hydro-
42 electricity generation from not only the Cerrn Grande reservoir but also from two reservoirs downstream (one of which was under construction at the time of the study). Sedimentation decreases power generation capacity only in the dry season; during the rains the flow of the river is sufficient to maintain hydro-electricity generation at full capacity. Based on expert opinion it is estimated that 75% of the soil erosion from the catchment is deposited in the Cerrn Grande Reservoir; this results in a 25% reduction in hydro-electricity generation capacity of the three dams. The loss in generation capacity is valued in terms of the least cost alternative source of power; electricity generated by thermal power stations.
4.2 Irrigation
The approaches employed in four different studies for valuation of the effects of erosion and conservation on water supplies for irrigation are reviewed below.
Abelson (1979) values reduction in reservoir yield due to sedimentation in terms of the resultant decline in output from dairy farms that use the water for irrigation. Based on the advice of the Australian State Rivers and Water Supply Commission, it is estimated that a 1% decrease in storage capacity of the Eppalock Reservoir results in a 1.5% decrease in water yield. The study assumes that the catchments soil conservation project reduces reservoir siltation rates from an estimated 615 megalitres (Ml) in 1960/61 to 246 Ml in 1967/68 and that sedimentation is stable at this rate from then onwards. Without the project it is assumed that conservation works and land improvements would have been carried out more slowly and siltation would have decreased linearly from 615 Ml in 1960/61 to 339 Ml in 1990/91. The value of water supplied to dairy farms for irrigation is calculated in terms of the social value of milk production. The price paid to Australian farmers for butterfat was subsidised by the Government; the study takes this into account by calculating the returns to dairy farmers based on world market prices.
The costs of reservoir sedimentation in Morocco are calculated by Brooks et al (1982) in terms of the value of loss in yield from irrigated crop production. Reductions in water yield for irrigation and in the land area under irrigated crop production are assumed to be directly proportional to the rate of reservoir sedimentation. The authors note that it is unlikely that reduction in reservoir dead storage capacity would directly affect water yield every year, but as the reservoir gradually fills with sediment the risk of seasonal shortfalls in water supply increases. Reduction in water supply is valued in terms of the loss in value of crop production on land that has to revert from irrigated to rainfed cropping. It is assumed that the costs of operation and maintenance of irrigation systems decrease in proportion to reservoir sedimentation.
43
The study by Brooks et al (1982) values the benefits of a watershed manage-ment project in terms of reduction in reservoir sedimentation. Soil erosion rates are calculated using a version of the USLE modified for conditions in Morocco. The erosion rates are transformed into reservoir sedimentation rates using a sediment delivery ratio based on the reservoirs sedimentation levels and trap efficiency. These sedimentation rates are used to estimate the effects of different phases of the projects activities on reservoir sedimentation. The supply of irrigation water saved as a result of the project is valued in terms of the additional returns to crop production on land that is able to remain irrigated as opposed to rainfed. It is assumed that the operation and maintenance costs of the irrigation system increase in proportion to the supply of irrigation water.
A slightly different approach is adopted by Attaviroj (1990) for a study based in Northern Thailand, which covers a catchment with four major reservoirs and sixteen smaller ones. The study examines the effects of reservoir sedimentation on the area of land under irrigation. However, the loss in irrigated land area is calculated by dividing the annual reduction in reservoir storage capacity by the quantity of water required to irrigate a unit area of land. Sedimentation is valued in terms of the loss in returns from the decrease in production of irrigated rice crops. The results represent the loss in land area that could potentially be irrigated, but it is unclear how representative this is of the actual impact of reservoir sedimentation on the area.
The cost of the deposition of silt in irrigation channels is examined by Magrath and Arens (1989). This is valued in terms of the resultant increase in the operation and maintenance costs of irrigation systems. Previous studies indicate that high rates of return can be achieved through additional expenditure on operation and maintenance to remove silt. Based on this information the authors assume that the cost of removal of silt is less than the loss in returns to irrigated crop production caused by siltation; the former (the least cost of siltation) is used for the valuation. The study estimates that the cost of silt removal represents 15% of the total expenditure on operation and maintenance of irrigation systems required in Java.
4.3 Navigation
Gregersen et al (1987) quote a study that examines the effects of reservoir sedimentation on navigation in the Panama Canal. The Lake Alajuela Reservoir
44 supplies water for the operation of locks in the Panama Canal. The study examines the loss in live reservoir storage that occurs due to sedimentation; it is estimated that live storage accounts for 65% of the reservoirs total storage capacity. The study assumes that sedimentation will reduce the supply of water to operate the locks in the Panama Canal. The cost of sedimentation is calculated in terms of the cheapest alternative method to maintain a navigable route through the Panama Canal; the least cost alternative proposed by the Panama Canal Company was to deepen the ship canal by three feet using dredgers. It is estimated that this would cost $ 27.5 million and would increase the water storage capacity of the canal by 370 million m 3 . These costings are used to value the impact of sedimentation of Lake Alajuela Reservoir on water yield for the Panama Canal.
The cost of siltation of waterways is also estimated by Magrath and Arens (1989) for their study of Java. Figures for national expenditure on dredging are used to estimate the costs of siltation. The authors are unable to break the data down into the costs of dredging navigable channels and dredging harbours. As a result, the study simply uses high and low estimates of the costs of siltation for the analysis.
4.4 Other off-site effects
Other off-site effects of erosion and conservation include impacts on water quality, deforestation and road maintenance.
Abelson (1979) examines water quality as an unquantifiable impact of a soil conservation project in Australia. The studys discussion illustrates the array of effects that a soil conservation project can have on one single environmental factor. The increase in livestock numbers that occurs under the project could increase bacterial levels in the water supply and so reduce the quality of drinking water. Furthermore, the agricultural practices promoted by the project increase water infiltration and so could increase leaching of salts into the water supply, further impairing the quality of water. Conversely the project decreases run-off which reduces water turbidity and decreases the need for a water filtration plant. The reduction in turbidity could, however, increase the risk of algal blooms as a result of greater penetration of sunlight into the water. The risk of algal blooms could also increase as a result of the rise in the number of livestock causing greater phosphate and nitrate levels in the water. After consideration of all these possible impacts the author decides that, on balance, the project could marginally decrease the quality of water in the area.
The effects of erosion-induced land degradation on forest encroachment are examined by Attaviroj (1990). In Northern Thailand, land degradation has created
45 pressure to expand the area under agricultural production; this has been achieved through clearing land under forestry. The study estimates that 200,000 rai of forest land are cleared each year to "replace" agricultural land that has become unproductive. The opportunity cost of deforestation is valued in terms of the income foregone from timber production, estimated as 330 million baht/year.
Finally, the impacts of a soil conservation project on road maintenance are estimated by Abelson (1979). The conservation project in Australia reduces flood damage caused to roads. As a result, the project decreased expenditure on road maintenance from $ 30,000/year in the early 1960s to $ 15,000 in 1975. It is assumed that without the project, expenditures would not have fallen to this level until 2010.
46 5. The Economic Analysis of Soil Erosion and Conservation
The economic impacts of soil erosion and conservation can be appraised through the use of financial analysis and cost benefit analysis. Studies employ one or both of these forms of analysis and carry them out at different levels: the farm unit, village, sub-catchment, catchment, administrative district, region, or national level. This section reviews the use of financial, cost benefit and cost-effectiveness analysis for the economic appraisal of soil erosion and conservation. Measures that can be adopted to address distributional issues and risk and uncertainty are also examined.
The economic valuation of soil erosion and conservation generally involves comparison against a base case scenario; the details and respective merits of three different base cases are described. The value of soil erosion can be calculated relative to a base case in which erosion is assumed not to occur. This is a somewhat artificial scenario; soil erosion is a naturally occurring process that takes place even on land under natural vegetation. As a result it can be argued that use of this base case results in over-valuation of the effects of erosion. A more representative base case is land under natural vegetation, though this does not represent a realistic alternative form of land use. Concerns about soil erosion are unlikely to result in the transfer of significant areas of land from crop production to natural vegetation. A more realistic scenario is crop production with appropriate soil conservation technologies, though this requires use of a site-specific erosion rate. In summary, the specifications of the base case used in an economic analysis can have significant effects on the results. For the sake of simplicity, studies generally adopt the zero-erosion scenario as the base case for the valuation of soil erosion. The impacts of conservation tend to be assessed through comparison with the scenario of crop production without conservation.
It is generally assumed that the specifications of the base case scenario remain unchanged over the period of the analysis. In some instances this assumption does not hold and the base case should be adjusted accordingly. One such example is the adoption of improved farming practices (e.g. use of hybrid seed and fertilisers) by farmers independently from the implementation of soil con-servation. A conservation project might promote and accelerate the adoption of practices that would have been adopted by farmers anyway but at a slower rate in the absence of the project. Abelson (1979) assumes that under a soil conser-vation project in Victoria, Australia, the project's conservation works and land improvements were carried out by all farmers over the life of the project, 1965 to 1970. In the absence of the project it is assumed that all farmers would still have carried out these
47 activities but the land improvements would not have been completed until 1990 and the soil conservation works not until 2005.
The following sections examine the different methods that can be adopted for the economic analysis of soil erosion and conservation.
5.1 Financial analysis
Financial analysis examines the monetary costs and benefits (or returns) of soil erosion and conservation to the private agent or individual. It comprises only those costs and benefits associated with goods and services that can be exchanged in the market place. These are valued at prevailing market prices with no adjustments made for any distortions in the market. Crop production used for household consumption is generally valued at market prices and household labour may be valued at the market wage rate or alternatively its cost may be excluded (see below and Section 5.2.1).
Financial analysis can be carried out from the perspective of a project, such that it covers all project activities, expenditures and returns, or from the perspective of a farmer, in terms of the financial costs and benefits that accrue to the individual. The analysis involves the construction of a budget for an enterprise or the entire farm unit, and calculation of the net returns and/or profit. These indicators can be used to appraise the financial impacts of soil erosion and to assess the financial viability and relative financial merits of different soil conservation scenarios. However, it can be argued that the use of net returns and profit as decision criteria for smallholder farmers is inappropriate. Maximisation of financial returns may not be the primary objective of such farmers; the achievement of household subsistence requirements and minimisation of risk may have greater priority.
The returns to labour for soil conservation work can be calculated by excluding labour costs from the financial analysis of soil conservation. Returns per man day can be compared with prevailing wage rates to assess whether the net benefits of conservation exceed the opportunity cost of labour (see Sections 5.2 and 5.2.1) or the cost of hired labour. Holmberg (1990) takes this approach one step further and examines the marginal returns to labour for different standards of construction and maintenance of terraces and cut-off drains. The standards to which farmers implement the soil conservation measures are divided into no conservation, average and superior conservation. The study finds that the adoption of both average and superior conservation by farmers that have no conservation results in marginal returns to labour "far above the levels of payment for casual and other unskilled labour in the area" (Holmberg, 1990).
48
The study concludes that "it is profitable for farmers who are working partly as casual labourers to concentrate on improving their own farms instead. There is, however, one drawback: the farmers will not receive any payment before the next harvest." Likewise, the cash and credit required by farmers to fund construction of the soil conservation measures may not be available.
Lindgren (1988) adopts the same approach for the financial analysis of a soil conservation project in Kenya. Three scenarios are examined: no conservation, partial adoption of the soil conservation package and complete adoption of the package. Returns per man day are calculated for each scenario and are only slightly above the casual labour wage rate for the complete adoption of the package. The study also examines the marginal returns to labour for partial or complete adoption of the soil conservation package by a farmer with no conser- vation (see Table 7). The marginal returns to labour for partial and complete adoption are calculated to be 2.5 and 4 times the wage rate for casual labour respectively, which indicates the economic attractiveness of the soil conser-vation package to farmers in the study area.
Table 7: Returns to Labour for a Soil Conservation Project in Machakos District, Kenya (Wage rate for casual labour: 20 Kenyan Shillings per man day)
No conservation Partial adoption of the conservation package a
Complete adoption of the conservation package b
Net return per manday (Kenyan Shillings)
5
11
21 Marginal return per manday for adoption by a farmer with no conservation (Kenyan Shillings)
-
50
82
Source: Lindgren, 1988. a construction of terraces on 0.95 ha of the farmer's land, cultivation of soil conser- vation crops, the planting of 50% of the recommended quantity of fodder grass on terrace edges and the planting of 25% of the recommended quantity of trees. b construction of terraces on 0.95 ha of the farmer's land, cultivation of soil conservation crops, the planting of the recommended quantity of fodder grass on terrace edges and the planting of the recommended quantity of trees. 5.2 Cost Benefit Analysis
49 Cost benefit analysis is the most common form of analysis used for the economic appraisal of soil conservation projects. It should examine all the costs and benefits that accrue to and result from a project, experienced on a national (or even international) level, including those that can not be quantified or valued. The analysis incorporates the effects of a project experienced directly and indirectly by individuals as well as those experienced by society as a whole. It should include all the costs and benefits of a project; those that can not be quantified should be qualitatively assessed or at least listed. In practice the quantitative component of cost benefit analysis is limited to certain costs and benefits and many impacts of the project tend to be completely ignored.
In cost benefit analysis the valuation of costs and benefits is carried out from a national resource perspective in terms of opportunity costs. The opportunity cost of project inputs represents the value of the resources in their next best alternative use ie: the value foregone by their use in the project. The opportun-ity cost value of project outputs represents the resource cost of obtaining the outputs from the next best alternative source. Opportunity costs can be derived from market prices but only if a market exists and is relatively competitive and unrestricted; even then, market prices must be adjusted for the effects of any market distortions (e.g. taxes and subsidies). When market prices cannot be used to value costs and benefits, the opportunity costs can be inferred, for example from the price of substitutes, the resource cost, or by the use of hedonic pricing or contingent valuation. For the purposes of cost benefit analysis the costs and benefits of a project are valued in real terms (ie: adjusted for inflation) and in a single currency unit. The distribution of these costs and benefits over time is accounted for by the use of an appropriately valued discount rate.
The following section examines the valuation of costs and benefits for use in cost benefit analysis. This is followed by a discussion on the choice of discount rate and time horizon for analysis. A review is then given of the terms calculated in cost benefit analysis that can be used to assess the economic impacts of soil erosion and conservation.
5.2.1 Valuation of Costs and Benefits The different approaches that can be adopted for valuation of the costs of soil erosion and the benefits of soil conservation are examined in Sections 2, 3 and 4.
Cost benefit analysis requires valuation in terms of opportunity cost and for these purposes goods and services can be divided into three groups: Goods that can be or are internationally traded, otherwise known as traded goods e.g. fertiliser, seed, rice and maize). These are valued at
50 their world market prices adjusted for internal transport and distribution costs (ie: at their import parity prices). Domestic factors of production. Land and labour are valued in terms of their marginal value product. Land may be valued at its market price, its market rent, or in terms of its productive capability, according to the form of land use and the competitiveness of the market. Further details on the valuation of labour are provided below. Goods and services that are not and cannot be internationally traded. The costs of non-traded items (e.g. electricity) are broken down into their traded and domestic components and valued accordingly. Non-traded items that comprise entirely of domestic costs can be valued at their market prices if the market is competitive. Alternatively, inputs (e.g. draught animals and bricks) are valued at their marginal value product and outputs (e.g. vegetables and milk) are valued at their marginal social benefit (the consumers willingness to pay).
In some instances it may be necessary to adjust the opportunity costs of trade-able items for the over-valuation of foreign exchange which results, for example, from import tariffs and export subsidies. This adjustment can be made through the use of a shadow foreign exchange rate or appropriate conversion factors. Further details are provided in standard texts such as Gittinger (1982) and ODA (1988).
The costs of soil conservation are relatively easy to value as most result from the use of resources that are exchanged in the market place. They can be divided into those costs that are met by the farmer and those that are met by the project and/or the government. In general, the farmer is responsible for the costs of the adoption, construction and maintenance of soil conservation technologies. Additional expenditure may also be required on inputs such as tools, seeds, fertilisers, labour and draught power, needed to implement conservation technologies. Further costs may also be incurred by the farmer for additional labour to harvest the higher yields resulting from soil conservation. The increase in yields may, however, be counterbalanced by a reduction in crop area due to the land occupied by conservation technologies.
The costs of soil conservation incurred by projects and governments include the costs of project planning, extension services to farmers and staff training. Projects and/or governments may provide seed, seedlings, fertilisers and tools to farmers for the purposes of soil conservation and in some instances they may fund the construction and maintenance of soil conservation measures.
51 In cost benefit analysis there is often some adjustment made to the valuation of labour inputs. In the case of skilled and semi-skilled labour it can generally be considered that the market wage rate reflects the opportunity cost. However, unskilled labour (which includes farm labour) tends to experience unemploy-ment in developing countries. As a result, the opportunity cost of unskilled labour ie: the value of production that could be achieved by the labour elsewhere in the economy, can be valued at zero. This is justified from the perspective of society, as the alternative is for the labour to be unemployed. From the perspective of the individual unskilled worker there is, however, a cost of their labour in terms of the disutility of work, the opportunity cost of the loss of time spent on household production and the opportunity cost of the loss of leisure time. Studies often value the opportunity cost (also known as the shadow wage rate) of unskilled labour at zero, though this fails to take into account any of the costs experienced by the individual worker.
The approaches adopted in several studies for the valuation of labour are given below. Bojs study based in Lesotho (1991a) includes both a financial and an economic analysis. The financial analysis values unskilled labour inputs in terms of the payments made to individuals by the project. These payments are made in cash or kind (hybrid seed, fertiliser and tree seedlings) and both are worth Maloti 3.5 per day. For the economic analysis the opportunity cost of unskilled labour is assumed to be equal to the local wage rate for hired casual agricultural labour. Boj justifies this on the grounds that it reflects unemployment in the study area and the disutility of work to individuals, which is broken down into the loss of leisure time and the loss of production for household consumption. The study uses relevant market wage rates to reflect the opportunity costs of skilled and semi- skilled labour, with the valuation of the latter justified by the employment opportunities for semi-skilled labour in South Africa.
In the study of a soil conservation project in Kenya, Lindgren (1988) takes the seasonality of labour into account in valuating the opportunity cost of farmers labour. In the study area both the construction and maintenance of soil conservation measures occur in the off-season. At this time of year there is un- employment of unskilled labour, so farmers labour inputs for soil conservation works are valued at zero. The author suggests that this is the appropriate value to use in the cost benefit analysis but from the perspective of the farmers their labour inputs have a cost in terms of the leisure time foregone.
Wiggins and Palma (1980) incorporate seasonality into the resource value of labour for their study in El-Salvador. The authors estimate the opportunity cost of labour in terms of the "sum paid to the last worker hired - that is, the lowest wage rates recorded in the [river] basin at that time." It is assumed that the con-servation
52 measures examined in the study require labour inputs only in the slack season. Use of the lowest price paid for labour indicates the compensation required by workers for the leisure time and work on their own land foregone.
Not all of the costs and benefits of soil erosion and conservation can be easily quantified and/or valued. The "unquantified" costs and benefits should, however, always be identified and listed as part of a cost benefit analysis. In some cases it may be possible to carry out a qualitative assessment of the unquantified impacts. Such an approach is adopted by Abelson (1979) for a project in Australia; the unquantified effects of the project (on flood mitigation, water quality, research value and soil fertility) are identified and divided into costs and benefits. These are then qualitatively assessed and balanced against each other in order to identify whether the net impact is positive or negative.
5.2.2 Discount rate and time horizon Specification of the discount rate and time horizon for cost benefit analysis should be carried out simultaneously as the two interact in their effect on the results. The discount rate is a compound rate of interest applied in its reciprocal form. It has little effect on the value of immediate costs and benefits but the greater the value of the discount rate, the smaller the impact of long term costs and benefits on the results of analyses. The effects of discounting are especially important for projects that involve immediate expenditures but have benefits that do not accrue until some time into the future.
Selection of the discount rate should be governed by the perspective of the study, be it from the stance of the government, donor, or farmer. If an investment is to be financed using public funds (by a host government or donor government) the social rate of discount should be used. The social discount rate reflects the opportunity cost of capital in the public sector (also known as the accounting rate of interest (ODA, 1988)); its value is specific to an individual country and is often available from the national planning authority. The cost benefit analysis of projects financed by the public sector should always be carried out using the social discount rate, even if analysis is carried out at the farm level. It is important to note that analysis at the farm level is distinct from analysis from the farmers perspective. Cost benefit analysis can be carried out from the farmers perspective only if the project is privately financed by the farmer. The farmers discount rate is then used which can be based on the interest rate payable by the farmer on an appropriate bank loan. The farmers discount rate is strongly influenced by the farmers level of risk aversion and so can be quite specific to an individual. In general, smallholder farmers have relatively high rates of time preference, or discount rates, and short planning horizons. In contrast, projects funded by the public sector have relatively low discount rates and long planning horizons. For
53 details on the selection of discount rates and time periods for cost benefit analysis the reader is referred to standard texts on economic appraisal such as ODA (1988).
The discount rates and time periods used in a selection of cost benefit analyses of soil erosion and/or conservation are given in Table 8. It is interesting to note the wide range in the discount rates used (1% to 20%) and in the time periods (5 years to 100 years). The table should be read bearing in mind that the choice of discount rate is affected by the country in which the analysis is based and the perspective from which the analysis is carried out (the farmer, project or government) and that the discount rate and time period are inter-dependent.
Table 8: Discount Rates and Time Periods used in Cost Benefit Analyses of Soil Erosion and Conservation in a Selection of Studies
Study Location Discount rate Time period of analysis Abelson (1979) Australia 8% a 30 years Attaviroj (1990) Thailand 5, 10 & 15% 15 years Bishop & Allen (1989) Mali 10% 10 years Boj (1991a) Lesotho 1% b & 10% c 50 years d
Briones (1985) Philippines 10% 50 years Brooks et al (1982) Morocco 10% 50 years Cuesta (1994) Cost Rica 20% 50 & 100 years Fleming (1983) & Hufschmidt et al (1983) Nepal 10% 20 years Kim & Dixon (1984) Korea 10% 15 years Lindgren (1988) Kenya 4% e & 1% f 20 years g
Magrath (1990a) India 10% 30 years Mitchell et al (1980) US 8% 20 years Mwakalagho (1986) Malawi 10% & 13.5% 20 years Samarakoon & Abeygunawardena (1992)
Sri Lanka
10%
5 years Swanson & Harshbarger (1964) US 5% & 20% 50 years Valds P. (1994) Honduras 20% 100 years Vsquez & Santamara (1994) Panama 20% 100 years White & Jickling (1994) Haiti 20% 50 years Wiggins and Palma (1980) El Salvador 10% h 30 years i
a the "inflation-free opportunity cost of capital". b the social opportunity cost of capital. c the discount rate most commonly used in international literature for the evaluation of soil conservation projects.
54 d justified on the grounds that if the project is not economically viable over a 50 year period it should not be implemented. The extension of the period of analysis will make little difference if the analysis is carried out at the higher discount rate (10%). However, this is not true for the lower discount rate of 1%. e to account for real interest rates and the surplus of donor funding in Kenya. f to account for high population growth rates, slow growth of the economy and the permanence of soil erosion. g expected life of the soil conservation structures. h based on the commercial rate of interest from 1969 to 1973 (6.2%), plus 2% to account for subsidised credits, plus 2% to account for unsatisfied demand. i justified on the grounds that the soil conservation measures should last for at least 30 years, but further extension of the time period would reduce the reliability of the price forecasting and, as a result of the effects of discounting, would have little effect on the results.
5.2.3 Measures of the economic impacts of soil erosion and conservation The results of cost benefit analyses can be expressed in terms of net present value, internal rate of return, benefit cost ratio and payback period. The respective merits of these different indicators are discussed in the literature on project appraisal (Gittinger, 1982 and ODA, 1988). Studies generally use a combination of measures to assess the economic viability of conservation projects. Comparisons can be made between different soil conservation tech-nologies in terms of the net cost of achieving a unit reduction in soil erosion (or a unit reduction in crop yield loss), or the economic cost of achieving a target maximum rate of soil erosion (or crop yield loss). This is the basis of cost effectiveness analysis, described in greater detail in Section 5.4
The economic significance of the impacts of erosion can be indicated by comparing the value of erosion with other economic indicators. At the farm level the cost of erosion can be compared with total expenditure on fertilisers or net returns to agricultural production. For studies on a regional or national scale, the cost of erosion can be compared with indicators such as the regional or national value of agricultural production, regional or national expenditure on fertilisers or the gross domestic product.
5.3 Partial and marginal analysis
The economic analysis of soil conservation projects can be simplified through the use of partial analysis. This requires quantification of only the major costs and
55 benefits that accrue to a project. If the project is found to be economically viable from such analysis it may be unnecessary to quantify the more intangible costs and benefits, though these should at least be listed. Brooks et al (1982) examine the benefits of a watershed management project in Morocco in terms only of the fruit production from trees planted as part of the project and the downstream benefits for irrigation. The authors note that the analysis excludes a number of additional project benefits, such as reduction in the loss of hydro-electricity generation capacity, reduction in the loss of municipal and industrial water supplies and increased flood control capacity. However partial analysis demon-strates that the project is economically viable without the inclusion of these additional benefits.
The data requirements for economic analysis of soil conservation projects may be further reduced through the use of marginal analysis. This involves economic analysis only of the additional costs and additional benefits that result from implementation of the project. Marginal analysis therefore removes the require- ments for data on a base case scenario and the prevailing farming system.
5.4 Cost effectiveness analysis
Cost effectiveness analysis can be used to determine the least cost conservation technology for achieving a target reduction in soil erosion or crop yield loss. The analysis calculates the total discounted cost of different conservation tech- nologies, all of which are assumed to reduce soil erosion to a specific target level. The cost-effective technology is that with the lowest total discounted cost, though the benefits of the technology may not necessarily exceed the costs. The advantage of this method is that valuation of the benefits of soil conservation (ie: the costs of soil erosion) is not necessary; it assumes that the benefits of all of conservation technologies are the same. In practice most technologies differ in their benefits: in the extent to which they reduce erosion, in the provision of animal fodder, fuel wood and in the reduction of costs of production. Supplementary partial cost benefit analysis ensures that the full merits of the conservation technologies are taken into account and that the most economically attractive option is identified.
5.5 Distributional issues
In its basic form cost benefit analysis does not take into account the distribution of the costs and benefits of projects. Certain groups of individuals may be dis- advantaged by a project, or the distribution of benefits may not conform with the
56 priorities for local development. The conventional method for incorporating distributional issues into cost benefit analysis is distributional weighting; the costs and benefits experienced by poor and disadvantaged groups of individuals may be given a greater weighting than those experienced by the better-off. Distributional weighting is, however, fraught with problems in terms of the actual distribution of costs and benefits between different groups and specifi-cation of the weights to be used. Alternative mechanisms adopted by several studies to incorporate distributional issues into their economic analyses are given below.
Abelson (1979) examines the distributional aspects of a soil conservation project in Australia by carrying out partial cost benefit analysis at four different levels. Analyses are carried out at the level of the farmer, the Government, the local region and the nation as a whole. Each analysis examines only the project costs and benefits experienced by the specific group of interest. The costs and benefits for each analysis are listed in Table 9 along with their associated present values.
An alternative approach for examining distributional issues is the analysis of cost and benefit sharing. For a study in India, Magrath (1990a) suggests that there are two possible scenarios for sharing the costs of soil conservation: sharing between the farmer and the landlord; and sharing between the farmer and the Government (through the payment of grants or subsidies). Magrath calculates internal rates of return for different combinations of cost and benefit sharing for the two conservation measures examined in the study. The internal rates of return are plotted on a graph to give iso-return curves for different combinations of cost and benefit sharing ie: combinations of cost and benefit sharing that give equal rates of return. The graph is used to assess the proportion of benefits a farmer needs to receive for it to be economically viable for the farmer to provide a specific proportion of the funding for soil conservation. The greater the gradient of the iso- return curves the greater the sensitivity of the investment to cost/benefit sharing.
Table 9: Distribution of the Benefits and Costs between Different Groups of Beneficiaries for the Eppalock Catchment Soil Conservation Project, Victoria, Australia
Group of beneficiaries Costs (total present value a ) Benefits (total present value a ) Expenditures on land improvements Returns from increased livestock
57 Farmers and expenditures related to increased livestock holdings ($ 1.29 m) production and the enhanced environ- mental value of the land ($ 2.86 m)
Government Expenditures on the construction and maintenance of soil conservation measures, project planning and extension, and the super phosphate subsidy ($ 1.10 m) Increased tax revenue from farmers, increased revenue from water sales and reduced road maintenance costs ($ 1.14 m)
Local region
Unspecified Stimulatory effect on the regional economy of increased returns to agriculture and to the local authority ($ 3.57 m) b
Nation
Unspecified National surplus resulting from increased livestock production and the surplus that accrues to foreign consumers from the reduction in wool prices due to increased production ($ 2.91 m)
Source: Abelson, 1979.
a calculated over a 30 year period discounted to 1960-61 at an 8% discount rate. b calculated by multiplying total net after-tax returns to farmers and the local authority by a suitable rural regional multiplier.
Day et al (1992) examine the different impacts of erosion on a farming community through economic analyses from two different perspectives adopted by farmers. The study examines a scenario in which farmers value the permanent long term loss in crop productivity that results from soil erosion and an alternative scenario in which they do not. In the former case, the cost of erosion in a single year is valued in terms of the present value of the loss in yields that occurs over the following ten years. In the latter case of the myopic farmer, value is placed only on the present loss in yields and not on any future losses. Interestingly there is little difference in the results for the two different scenarios, though this finding is highly dependent on the assumptions used in the analysis.
To a limited extent the distribution of costs and benefits can be investigated through accompanying full (national-scale) cost benefit analysis (which implicitly adopts the perspective of society) with financial analysis from the perspective of the farmer. This is the approach employed by Boj (1991a) for a study in Lesotho. Wiggins and Palma (1980) take it one step further and use three different economic analyses to investigate distributional issues for their study in El- Salvador. Financial analysis of erosion and conservation from the perspective of the national cash flow investigates the impacts on the national economy. Financial
58 analysis at the farm level determines the impacts on the farmer and full cost benefit analysis assesses the effects of erosion and conservation on society as a whole.
5.6 Risk and uncertainty
Risk and uncertainty can be integrated into both financial and cost benefit analysis. Risk can be incorporated into the values of variables by adjusting the data used to reflect their associated probabilities. Uncertainty, for which the associated probability is (by definition) unknown, is taken into account through sensitivity analysis. This indicates the resilience of the economic viability of a project to changes in the values of the data used. Sensitivity analysis should include all key variables and use values that either represent the likely best and worst case scenarios, or that are hypothetical e.g. 10% more or less than the expected value.
In addition to sensitivity analysis, the susceptibility of a project to risk and uncertainty can be roughly assessed in terms of the size of the net present value. The lower the net present value the more sensitive the economic viability of the project to changes in values of the data. For soil conservation works funded by individual farmers, the vulnerability to risk and uncertainty can be roughly assessed by comparing the cost of investment with net farm income. Broadly speaking, the smaller the size of the investment relative to net farm income, the lower the vulnerability to risk and uncertainty.
This review of economic analyses of soil erosion and conservation indicates the overwhelming problems encountered with the availability and accuracy of data. In the absence of accurate data, sensitivity analysis should be used to assess the robustness of the results of analyses to changes in the data. Detailed sensitivity analysis is incorporated into some studies but others include limited or even no sensitivity analysis. Details of the sensitivity analyses carried out in a selection of studies for both financial and cost benefit analyses are summarised in Table 10.
59 Table 10: A Summary of the Sensitivity Analysis carried out in a selection of Economic Analyses of Soil Erosion and Conservation a
Study Variables subjected to sensitivity analysis Abelson (1979) price levels, stocking densities, discount rates and time periods for analysis Attaviroj (1990) adoption rates of one of the soil conservation packages examined Bishop & Allen (1989) effects of erosion on crop yields, period of analysis (or period until fallowing), discount rate
Boj (1991a) project performance after handover to the local authorities, income from fruit production, rate of crop yield decline, future trends in grain prices, the weighting of benefits accruing to Lesotho farmers Briones (1985) project costs, reservoir sedimentation rate Brooks et al (1982) +/- 25% for all costs and benefits Day et al (1992) None Cuesta (1994) time period Grohs (1992) rate of erosion-induced decline in crop yields Hernndez (1992) rate of erosion-induced decline in crop yields, rate of soil erosion, costs of implementing and maintaining soil conservation measures Kalikandra & Hoekstra b levels of response of crop yields to increased nutrient and soil moisture levels Lu (1994) time period, discount rate
Lindgren (1988) discount rate, increases in yields resulting from soil conservation, rate of yield decline in the absence of soil conservation, acreage terraced per farm, future trends in crop prices, time period, combination of the lowest of all the above figures to give a worst case scenario Magrath (1990a) increases in yields resulting from soil conservation, erosion-induced yield decline, life expectancy of conservation measures, time period Samarakoon & Abeygunawardena (1992) None Swanson & Harshbarger (1964) discount rate, time period
Valds P. (1994) +/- 10, 25 and 50% for each of: rate of decline in yield without conservation, rate of increase in yield with conservation, cost of construction and maintenance of soil conservation measures Vsquez & Santamara (1994) +/- 10, 25 and 50% for each of: cost of clearing land, number of years continuous cultivation with soil conservation, cost of construction and maintenance of conservation measures White & Jickling (1994) erosion rate, costs of construction and maintenance of conservation measures
Wiggins & Palma (1980) worst case scenario for each of the three analyses (national, farm, social) in terms of: soil erosion rate, rate of yield decline resulting from soil erosion, initial increase in yields resulting from conservation, costs of conservation, crop yields and prices, cost of labour, discount rate
a carrying out all forms of economic analysis, not just cost benefit analysis b as described by Dixon et al (1990)
60 6. Discussion
The economic analysis of soil erosion and conservation adopts numerous different approaches. There is general consensus about the method of economic appraisal; analyses generally use cost benefit analysis and/or financial analysis. However, there is a wide diversity in approaches for valuing the costs and benefits of erosion and conservation.
The on-site impacts of erosion are generally valued in terms of its effects on soil properties or crop yields. The former tends to be valued in terms of the "replacement cost" of losses in soil nutrient content. However, this is a theoretical approach to valuation; in practice all the soil nutrients lost through erosion cannot be and are not replaced. The replacement cost also has little bearing on the economic value of soil as a resource. Valuation of erosion in terms of loss in crop yields examines the effects of erosion on the productive capacity of the soil; this is the cost of erosion generally experienced by farmers.
However, quantifying the effects of erosion on yields is highly problematic. It is difficult to isolate the effects of erosion from other factors that influence yields. There is also a lack of data on the relationship between erosion and yield, which is anyway highly situation-specific. Other less well established approaches for valuing the on-site effects of soil erosion include hedonic pricing and contingent valuation.
The on-site benefits of soil conservation technologies are generally valued in terms of the reduction in the costs of erosion. This involves quantifying the extent to which the conservation technologies reduce the effects of erosion as well as valuation of the effects of erosion. The majority of studies examine the benefits of conservation in terms of the reduction in crop yield loss but data on this relationship are situation specific and limited in their availability. A few studies use field data, but where it is not available studies use data from empirical models, from other studies or even hypothetical data. Other effects of conser-vation include reduction in crop area under cultivation, reduction in costs of crop production, and increase in livestock production.
The downstream effects of soil erosion are numerous but are examined in relatively few studies. The majority of analyses that incorporate the off-site impacts of erosion examine only the effects of reservoir sedimentation. The latter is generally valued in terms of the reduction in hydro-electricity generation and water supplies for irrigation. Quantifying and valuing the off-site effects of erosion is far from easy and is made more difficult if the study area does not cover an entire catchment or sub-catchment. For the purposes of cost benefit analysis the
61 off-site effects of erosion and conservation should, however, always be included; those effects that cannot be readily quantified or valued can be qualitatively assessed or should at least be listed.
The methodologies for financial and cost benefit analysis of soil erosion and conservation are straight forward, once all the benefits and costs have been valued. The amount of data required can be reduced by the use of partial analysis which quantitatively assesses only the major costs and benefits, or the use of marginal analysis. The economic appraisal of conservation technologies using cost effectiveness analysis removes the need to value the benefits of soil conservation but can under-value the merits of the different conservation technologies.
The overwhelming problem encountered in the economic analysis of soil erosion and conservation is the lack of data on the impacts of erosion and conservation, particularly in developing countries. As a result, studies are forced to adopt the best possible approach for the valuation of erosion and conservation given the available data. Economic analyses are often based on data extrapolated from other geographical areas or that are, at best, estimates. The results of the analyses are significantly influenced by both the data used and the valuation approaches adopted. It is therefore essential that the economic analysis of erosion and conservation is accompanied by explicit details of the assumptions, the basis on which they were made, and sensitivity analysis which assesses the effects of the assumptions on the results of the analysis. For economic analyses to be accurate or even just representative, further data are required on the impacts of soil erosion and conservation. In the absence of such data, economic analyses must be accompanied by the necessary caution signs to ensure that results that are based on estimates are interpreted only as estimates and do not become engraved in stone.
62 References
Abelson, P. (1979) Cost benefit analysis and environmental problems. Saxon House, Farnborough.
Alegre, J.C., Felipe-Morales, C. & La Torre, B. (1990) Soil erosion studies in Peru. Journal of Soil and Water Conservation (May-June) 417-420.
Anderson, J.R. & Thampapillai, J. (1990) Soil conservation in developing countries: project and policy intervention. Policy & Research Series No. 8, The World Bank, Washington DC.)
Attaviroj, P. (1990) Soil erosion and land degradation in the Northern Thai Uplands in J.A. Dixon, D.E. James & P.B. Sherman (eds) Dryland management: economic case studies. Earthscan, London pp 18-33.
Barbier, E.B. (1990) The farm-level economics of soil conservation: the uplands of Java. Land Economics 66 (2) 199-211.
Barrow, C.J. ( ) Land degradation - development and breakdown of terrestrial environments. Cambridge University Press, Cambridge.
Benson, V.W., Rice, O.W., Dyke, P.T., Williams, J.R. & Jones, C.A. (1989) Conservation impacts on crop productivity for the life of the soil. Journal of Soil & Water Conser- vation (Nov-Dec) 601-604.
Bishop, J. (1992) Economic analysis of soil degradation. Gatekeeper Series No. 92-01. London Environmental Economics Centre, London.
Bishop, J. & Allen, J. (1989) The on-site costs of soil erosion in Mali. Environment Working Paper No. 21, World Bank, Washington DC.
Bishop, J., Aylward, B. & Barbier, E.B. (1991) Guidelines for applying environmental economics in developing countries. Gatekeeper Series No. LEEC 91-02, International Institute for Environment & Development, London.
Blaikie, P. (1985) The political economy of soil erosion in developing countries. Longman Scientific & Technical, Harlow.
Blyth, M. & McCallum, A. (1987) Onsite costs of land degradation in agriculture and forestry in A. Chisolm & R. Dumsday (eds) Land degradation: problems and policies. Cambridge University Press, Cambridge pp 79-98.
Boj, J. (1986a) A method for economic appraisal/evaluation of conservation projects in Southern African Development Coordination Conference (SADCC) Cost-benefit analysis of soil and water conservation projects - report from a seminar held April 21-25 in Maseru. Report No. 4, SADCC pp 16-21. Boj, J. (1986b) Lessons from ten CB studies of conservation projects in Southern African Development Coordination Conference (SADCC) Cost-benefit analysis of soil and water
63 conservation projects - report from a seminar held April 21-25 in Maseru. Report No. 4, SADCC pp 32-34.
Boj, J. (1989) Sustainable use of land in developing countries. Reprint series No. 12, Department of International Economics & Geography, Stockholm School of Economics (Stockholm Studies in Natural Resources Management No. 1 pp 39-51).
Boj, J. (1990) Case study: benefit-cost analysis of soil conservation in Maphutseng, Lesotho. Reprint series No. 23, Department of International Economics & Geography, Stockholm School of Economics (in J.A. Dixon, D.E. James & P.B Sherman (eds) The economics of dryland management. Earthscan, London pp 250-288).
Boj, J. (1991a) The economics of land degradation: theory and applications to Lesotho. The Economic Research Institute, Stockholm School of Economics.
Boj, J. (1991b) Economics and land degradation. Reprint series No. 57, Department of International Economics & Geography, Stockholm School of Economics (Ambio 20 (2) 75-79).
Boj, J. (1991c) Economic analysis of environmental impacts. Reprint series No. 70, Department of International Economics & Geography, Stockholm School of Economics (in C. Folke & T. Kberger (eds) Linking the environment and the economy: essays from the Eco-Eco Group. Kluwer Academic Publishers, Doedrecht pp 43-59).
Boj, J. (1994) The costs of land degradation from a national perspective: an assessment of African evidence. Paper prepared for the 8th International Soil Conservation Conference, December 4-8 1994, New Delhi, India. Environmentally Sustainable Development Division, the World Bank, Washington DC.
Briones, N.D. (1987) Example of estimating the costs of erosion in the Lower Agno River Watershed in the Philippines. Annex 3 in Gregersen, H.M., Brooks, K.N., Dixon, J.A. & Hamilton, L.S. (1987) Guidelines for economic appraisal of watershed management projects. FAO Conservation Guide No. 16, FAO, Rome.
Brooks, K.N., Gregersen, H.M., Bergluna, E.R. & Tayaa, M. (1982) Economic evaluation of watershed projects - an overview of methodology and application. Water Resources Bulletin 18 (20) 245-250. (Annex I in Gregersen, H.M., Brooks, K.N., Dixon, J.A. & Hamilton, L.S. (1987) Guidelines for economic appraisal of watershed management projects. FAO Conservation Guide No. 16, FAO, Rome).
Brown, L.R. (1981) Eroding the base of civilisation. Journal of Soil & Water Conservation (Sept-Oct) 255-260.
Brown, L.R. & Wolf, E.C. (1984) Soil erosion: quiet crisis in the world economy. Worldwatch Paper No. 60, Worldwatch Institute, Washington DC.
Colacicco, D., Osborn, T. & Alt, K. (1989) Economic damage from soil erosion. Journal of Soil & Water Conservation (Jan-Feb) 35-39.
64 Cuesta, M.D. (1994) Economic analysis of soil conservation projects in Costa Rica in E. Lutz, S. Pagiola, & C. Reiche (eds) Economic and institutional analyses of soil conservation projects in Central America and the Caribbean. World Bank Environment Paper No. 8, The World Bank, Washington DC pp 40-52.
Day, J.C., Hughes, D.W. & Butcher, W.R. (1992) Soil, water and crop management alternatives in rainfed agriculture in the Sahel: an economic analysis. Agricultural Economics 7 262- 287.
Doolette, J.B. & Smyle, J.W. (1990) Soil and moisture conservation technologies: review of literature in J.B. Doolette & W.B. Magrath (eds) Watershed development in Asia - strategies and technologies. Technical Paper No. 127, World Bank, Washington DC pp 35-70.
Dumsday, R.G. & Seitz, W.D. (1985) A model for quantifying incentive payments for soil conservation in cropping regions subject to water erosion in S.A. El-Swaify, Moldenhauer, W.C. & Lo, A (eds) Soil erosion and conservation. Soil Conservation Society of America pp 296-305.
El-Swaify, S.A. (1993) Soil erosion and conservation in the humid tropics in D. Pimentel (ed) World soil erosion and conservation. Cambridge University Press, Cambridge pp 233- 255.
Fleming, W.M. (1979) Environmental and economic impacts of watershed conservation on a major reservoir project in Ecuador in M.M. Hufschmidt & E.L. Hyman (eds) (1979) Economic approaches to natural resource and environmental quality analysis. Proceedings of a Conference on Extended Benefit-Cost Analysis, September 19-26, East-West Center, Honolulu, Hawaii. Tycooly International, Dublin pp 292-297.
Fleming, W.M. (1983) Phewa Tal catchment management program: benefits and costs of forestry and soil conservation in Nepal in L.W. Hamilton (ed) Forest and watershed development and conservation in Asia and the Pacific. Westview Press, Boulder, Colarado pp 217-288 vide Gregersen et al, 1985.
Gittinger, J.P. (1982) Economic analysis of agricultural projects. EDI Series in World Development, Johns Hopkins Press for the World Bank, Baltimore.
Gregersen, H.M., Brooks, K.N., Dixon, J.A. & Hamilton, L.S. (1987) Guidelines for economic appraisal of watershed management projects. FAO Conservation Guide No. 16, FAO, Rome.
Grohs, F. (1992) Monetarising environmental damages, a tool for development planning? A case study of soil erosion in Zimbabwe. Paper presented at the Second Meeting of the International Society for Ecological Economics (ISEE) "Investing in natural capital. A prerequisite for sustainability", 3-6 August, Stockholm, Sweden.
Hernndez, J.A. (1994) Economic and institutional analysis of soil conservation practices in the Dominican Republic in E. Lutz, S. Pagiola, & C. Reiche (eds) Economic and institutional
65 analyses of soil conservation projects in Central America and the Caribbean. World Bank Environment Paper No. 8, The World Bank, Washington DC pp 89-97.
Holmberg, G. (1990) An economic evaluation of soil conservation in Kitui District, Kenya in J.A. Dixon, D.E. James & P.B. Sherman (eds) Dryland management: economic case studies. Earthscan, London pp 56-71.
Hudson, N. (1981) Soil conservation. B.T. Batsford, London.
Hudson, N. (1988) Soil conservation strategies for the future. Paper presented to the 5th International Soil Conservation Conference, January 18-29, Bangkok, Thailand.
Hufschmidt, M.M, James, D.E., Meister, A.D., Bower, B.T., Dixon, J.A. (1983) Environment, natural systems and development: an economic valuation guide. Johns Hopkins University Press, Baltimore pp 175-183 vide Gregersen et al, 1985.
Hurni, H. (1985) Erosion-productivity-conservation systems in Ethiopia. Paper presented for IVth International Conference on Soil Conservation, November 3-9, Maracay, Venezuela.
Inthapan, P. & Boonchee, S. (1988) Conservation cropping system of the Highland of Northern Thailand. Paper presented at the 5th Seminar on Farming Systems, April 4-7, Kempeangsan, Thailand.
Johnson, L.C. (1987) Soil loss tolerance: fact or myth? Journal of Soil & Water Conservation (May-June) 155-160.
Jones, A.J., Selley, R.A. &111 Miekle, L.N. (1990) Cropping and tillage options to achieve erosion control goals and maximum profit on irregular slopes. Journal of Soil & Water Conservation (Nov-Dec) 649-653.
Kigoda, M.A. (1988) Essential economic and design tools in the field of soil and water conservation and land utilisation (SWCLU) in Southern African Development Coordination conference (SADCC) Economics of conservation - proceedings from a workshop held March 1-8, Maseru. SADCC, Maseru pp 5-21.
Kim, S.H. & Dixon, J.A. (1984) Economic valuation of environmental quality aspects in Ahmad, Y.F. (ed) Environmental decision-making, vol. II.
King, D.A. & Sinden, J.A. (1988) Influence of soil conservation on farm land values. Land Economics 64 (3) 242-255.
Kiome, R.M. & Stocking, M. ( ) Rationality of farmer perception of soil erosion. Paper for Special Issue of Global Environmental Change, (ed) H. Brookfield. Kirby, M. & Blyth, M. (1987) An economic perspective on government intervention in land degradation in A. Chisolm & R. Dumsday (eds) Land degradation: problems and policies. Cambridge University Press, Cambridge pp 213-222.
66 Lal, R. (1981) Soil erosion problems on alfisols in Western Nigeria VI: effects of erosion on experimental plots. Geoderma 25 215-130 vide Boj, 1991a.
Lal, R. (1987) Effects of soil erosion on crop productivity. CRC Critical Reviews in Plant Sciences 5 (4) 303-367.
Lee, M.T., Narayanan, A.S., Swanson, E.R. (1974) Economic analysis of erosion and sedimentation, Seven Mile Creek Southwest Branch Watershed. Department of Agricultural Economics, Agricultural Research Station, University of Illinois.
Lindgren, B.M. (1988) Economic evaluation of a soil conservation project in Machakos District, Kenya: a minor field study. Arbetsrapport 95, Working Paper, International Rural Development Centre, Swedish University of Agricultural Sciences, Uppsala.
Lockeretz, W. (1981) Crop residues for energy: comparative costs for the farmer, the energy facility and the public. Energy in Agriculture 1 71-89 vide Boj, 1991a.
Lu, Y. & Stocking, M. (1994) A simulation model for decision-making in soil conservation - design and validation for use on the Loess Plateau, China. Paper presented at the 8th International Soil Conservation Conference, December 4-8, New Delhi, India.
Lutz, E., Pagiola, S. & Reiche, C. (1994) Lessons from economic and institutional analyses of soil conservation projects in Central America and the Caribbean in E. Lutz, S. Pagiola, & C. Reiche (eds) Economic and institutional analyses of soil conservation projects in Central America and the Caribbean. World Bank Environment Paper No. 8, The World Bank, Washington DC pp 3-18.
Lyles, L. (1975) Possible effects of wind erosion on soil productivity. Journal of Soil & Water Conservation (Nov-Dec) 279-283 vide Boj, 1991a.
Magrath, W.B. (1990a) Economic analysis of soil conservation technologies in J.B. Doolette & W.B. Magrath (eds) Watershed development in Asia - strategies and technologies. Technical Paper No. 127, World Bank, Washington DC pp 71-96.
Magrath, W.B. (1990b) Economic analysis of off-farm soil conservation structures in J.B. Doolette & W.B. Magrath (eds) Watershed development in Asia - strategies and technologies. Technical Paper No. 127, World Bank, Washington DC pp 97-108.
Magrath, W. & Arens, P. (1989) The costs of soil erosion on Java: a natural resource account- ing approach. Environment Department Working Paper No. 18, Environment Department, World Bank, Washington DC.
Magrath, W.B. & Doolette, J.B. (1990) Strategic issues in watershed development in J.B. Doolette & W.B. Magrath (eds) Watershed development in Asia - strategies and technologies. Technical Paper No. 127, World Bank, Washington DC pp 1-34. Massee, T.W. & Waggoner, H.O. (1985) Productivity losses from soil erosion on dry cropland in the intermountain area. Journal of Soil & Water Conservation (Sept-Oct) 447-450.
67 McLaughlin, L. (1993) A case study of Dingxi County, Gansu province, China in D. Pimentel (ed) World soil erosion and conservation. Cambridge University Press, Cambridge pp 87-107.
Mills, W.C, Thomas, A.W. & Langdale, G.W. (1981) Conservation tillage and season effects on soil erosion risk. Journal of Soil and Water Conservation (Nov-Dec) 457-460.
Miranowski, J.A. & Hammes, B.D. (1984) Implicit prices of soil characteristics for farm land in Iowa. American Journal of Agricultural Economics (Dec) 745-749.
Mitchell, J.K., Brach, J.C. & Swanson, E.R. (1980) Costs and benefits of terraces for erosion control. Journal of Soil & Water Conservation (Sept-Oct) 233-236.
Moore, I.D. & Wilson, J.P. (1992) Length-slope factors for the revised Universal Soil Loss Equation: simplified method for estimation. Journal of Soil and Water Conservation 47 (5) 423-428.
Munasinghe, M. & Lutz, E. (1993) Environmental economics and valuation in decision making in M. Munasinghe (ed) Environmental economics and natural resource management in developing countries. Committee of International Development Institutions of the Environment, the World Bank, Washington DC pp 17-72.
Mwakalagho, R.J.M. (1986) A CB study of soil conservation in Malawi in Southern African Development Coordination Conference (SADCC) Cost-benefit analysis of soil and water conservation projects - report from a seminar held April 21-25 in Maseru. Report No. 4, SADCC pp 35-42.
Mwansa, K., Mbewe, C.M. & Chiti, R. (1990) Lusaka Rural Agro-Forestry Project: a case study of economics and conservation in South African Development Coordination Conference (SADCC) Economics of conservation - proceedings from a second workshop held November 28 - December 2 in Manzini, Swaziland. Report No. 23, SADCC pp 156-167.
Navas, J.R.H., Melndez, R.L. & Alferes, J.B. (1994) Soil conservation projects in El Salvador in E. Lutz, S. Pagiola, & C. Reiche (eds) Economic and institutional analyses of soil conservation projects in Central America and the Caribbean. World Bank Environment Paper No. 8, The World Bank, Washington DC. pp 53-57.
Nobe, K.C. & Seckler, D.W. (1979) An economic and policy analysis of soil-water problems in the Kingdom of Lesotho. LADSA Report No. 3, Ministry of Agriculture and Department of Economics, Colorado State University vide Boj, 1991a.
Norse, D. & Saigal, R. (1992) National economic cost of soil erosion in Zimbabwe in M. Munasinghe (ed) Environmental economics and natural resource management in developing countries. Committee of International Development Institutions of the Environment, the World Bank, Washington DC pp 229-241.
68 Obando E., M. & Montlvan, D. (1994) Technical and economic analysis of a soil conservation project in Nicaragua in E. Lutz, S. Pagiola, & C. Reiche (eds) Economic and institutional analyses of soil conservation projects in Central America and the Caribbean. World Bank Environment Paper No. 8, The World Bank, Washington DC pp 75-80.
Overseas Development Administration (ODA) (1988) Appraisal of projects in developing countries: a guide for economists. HMSO, London.
Pagiola, S. (1992) Estimating the yield effect of soil degradation. Soil & Water Conservation 47 (6) 486-490.
Pagiola, S. (1994) Cost benefit analysis of soil conservation in E. Lutz, S. Pagiola, & C. Reiche (eds) Economic and institutional analyses of soil conservation projects in Central America and the Caribbean. World Bank Environment Paper No. 8, The World Bank, Washington DC pp 21-39.
Parkinson, J. (1990) Evaluation of soil conservation measures. Prepared for the Africa Section of the Overseas Division, AFRC Engineering.
Pearce, D., Markandya, A. & Barbier, E.B. (1989) Blueprint for a green economy. Earthscan, London.
Pimentel, D., Harvey, C., Resosudarmo, P., Sinclair, K., Kurz, D., McNair, M., Crist, S., Shpritz, L. Fitton, L., Saffouri, R. & Blair, R. (1995) Environmental and economic costs of soil erosion and conservation benefits. Science 267 (24th February) 1117-1123.
Ponzi, D. (1993) Soil erosion and productivity; a brief review. Desertification Control Bulletin 22 36-44.
Portillo, E.M. (1989) Research priorities and the economics of soil and water management in J. kerman. & A. Pain (eds) (1989) Objectives, strategies and techniques in applied research on soil and water management - proceedings from a workshop held 26-29 September in Gaborone, Botswana. Southern African Development Coordination Conference, Report No. 21, Maseru.
Prato, T. & Wu, S. (1991) Erosion, sediment, and economic effects of conservation compliance in an agricultural watershed. Journal of Soil & Water Conservation (May-June) 211-214. Prem Kumar, P.D. (1994) Farmers are engineers - indigenous soil and water conservation practices in a participatory watershed development programme. Pidow Myrada, Gulbarga, Karnataka, India.
Purvis, A., Hoehn, J.P., Sorenson, V.L. & Pierce, F.J. (1989) Farmers' response to a filter strip program: results from a contingent valuation survey. Journal of Soil & Water Conservation (Sept-Oct)501-504.
Rosado P., P., Barrientos C., L.E., Lima L., S.A. (1994) Soil erosion control efforts in Guatemala in E. Lutz, S. Pagiola, & C. Reiche (eds) Economic and institutional analyses of soil conservation projects in Central America and the Caribbean. World Bank Environment Paper No. 8, The World Bank, Washington DC pp 58-62.
69
Rosenberry, P., Knutson, R. & Harmon, L. (1980) Predicting the effects of soil depletion from erosion. Journal of Soil & Water Conservation (May-June) 131-134.
Rydgren, B. (1993) Environmental impacts of soil erosion and soil conservation: a Lesotho case study. Doctoral thesis, UNGI Rapport Nr 85, Imstiute of Earth Sciences, Uppsala University, Sweden.
Samarakoon, S.M.M. & Abeygunawardena, P. (1992) An economic assessment of on-site effects of soil erosion in potato lands in Nuwara Eliya District of Sri Lanka. Unpublished paper, University of Peradeniya, Kandy, Sri Lanka.
Seckler, D. (1987) Economic cost and benefits of degradation and its repair - issues in the economic valuation of soil and water conservation programmes in P. Blaikie & H. Brookfield (eds) Land degradation and society. Methuen, London pp 84-96.
Sfeir-Younis, A. ( ) Economic aspects of soil conservation programmes in LDCs in Polders of the world: final report pp 187-220.
Siebert, S.F. & Belsky, J.M. (1990) Bench terracing in the Kerinchi uplands of Sumatra, Indonesia. Journal of Soil & Water Conservation (Sept-Oct) 559-562.
Sinden, J.A., Sutas, A.R. & Yapp, T.P. (1990) Damage costs of land degradation: anAustralian perspective in J.A. Dixon, D.E. James & P.B. Sherman (eds) Dryland management: economic case studies. Earthscan, London pp 265-282.
Singh, G., Babu, R., Narain, P., Bhushan, L.S. & Abrol, I.P. (1992) Soil erosion rates in India. Journal of Soil and Water Conservation 47 (1) 97-99.
Stocking, M. (1984) Erosion and soil productivity - a review. Manuals & Reports Series, No. 22, School of Development Studies, East Anglia. Consultant's Working Paper No. 1, Soil Conservation Programme, Land & Water Development Division, FAO, Rome.
Stocking (1987) Measuring land degradation in P. Blaikie & H. Brookfield (eds) Land degradation and society. Methuen, London pp 49-63.
Stocking, M. (1988) Quantifying the on-site impact of soil erosion in S. Rimwanich (ed) Land conservation for future generations. Proceedings of the 5th International Soil Conservation Conference, January 18-29, Bangkok, Thailand. Food and Agriculture Organisation of the United Nations & Overseas Development Group, University of East Anglia. Stocking, M. (1995) Soil erosion and land degradation in T. O'Riordan (ed) Environmental science for environmental management. Longman, Harlow pp 223-242.
Stocking, M. & Pain, A. (1983) Soil life and the minimum soil depth for productive yields: developing a new concept. Discussion Paper No. 150, School of Development Studies, University of East Anglia, Norwich vide Boj, 1991a.
70 Swanson, E.R. & Harshbarger, C.E. (1964) An economic analysis of effects of soil loss on crop yields. Journal of Soil & Water Conservation (Sept-Oct) 183-186.
Tagwira, F. (1992) Soil erosion and conservation techniques for sustainable crop production in Zimbabwe. Journal of Soil and Water Conservation (Sept-Oct) 371-374.
Tejwani, K.G. (1981) Watershed management as a basis for land development and management in India in R. Lal & E.W. Russell (eds) Tropical agricultural hydrology. John Wiley & Sons Ltd, Chichester pp 239-255.
Troeh, F.R., Hobbs, J.A. & Donahue, R.L. (eds) (1980) Soil and water conservation for productivity and environmental protection. Prentice Hall, New Jersey.
Upstill, G. & Yapp, T. (1987) Offsite costs of land degradation in A. Chisolm & R. Dumsday (eds) Land degradation: problems and policies. Cambridge University Press, Cambridge pp 99-107.
Valds P., A. (1994) Economic analysis of soil conservation in Honduras in E. Lutz, S. Pagiola, & C. Reiche (eds) Economic and institutional analyses of soil conservation projects in Central America and the Caribbean. World Bank Environment Paper No. 8, The World Bank, Washington DC pp 63-74.
van der Pol, F. (1992) Soil mining - an unseen contributor to farm income in southern Mali. Bulletin 325, Royal Tropical Institute, Amsterdam.
Vsquez, T. & Santamara, J. (1994) Economic and institutional analysis of soil conservation at the farm level in Cocl, Panama in E. Lutz, S. Pagiola, & C. Reiche (eds) Economic and institutional analyses of soil conservation projects in Central America and the Caribbean. World Bank Environment Paper No. 8, The World Bank, Washington DC.
Vittal, K.P.R., Vijayalakshmi, K. & Rao, U.M.B. (1990) The effect of cumulative erosion and rainfall on sorghum, pearl millet and castor bean yields under dry farming conditions in Andra Pradesh, India. Experimental Agriculture 26 429-439 vide Boj, 1991a.
White, T.A. & Jickling, J.L. (1994) An economic and institutional analysis of soil conservation in Haiti in E. Lutz, S. Pagiola, & C. Reiche (eds) Economic and institutional analyses of soil conservation projects in Central America and the Caribbean. World Bank Environment Paper No. 8, The World Bank, Washington DC pp 98-106.
Wiggins, S.L. (1981) The economics of soil conservation in the Acelhaute River Basin, El Salvador in R. Morgan (ed) Soil conservation: problems and prospects. J.Wiley, Chichester, pp 399-418. Wiggins, S.L. & Palma, O.G. (1980) Acelhuate River Catchment Management project, El Salvador - Cost benefit analysis of soil conservation. Project Report No. 104, Land Resources Development Centre, Overseas Development Administration, Surbiton.
Winpenny, J.T. (1991) Values for the environment - a guide to economic appraisal. Overseas Development Institute, HMSO, London.
71 Woolman, M.G. (1985) Soil erosion and crop productivity: a worldwide perspective in R.F. Follett & B.A. Stewart (eds) Soil erosion and crop productivity. American Society of Agronomy, Winchester pp 9-21.