Assessing the costs and benefits of improved land management practices in three
watershed areas in Ethiopia
Tesfaye, Abonesh; Brouwer, Roy; van der Zaag, Pieter; Negatu, Workneh
DOI
10.1016/j.iswcr.2016.01.003
Publication date
2016
Document Version
Final published version
Published in
International Soil and Water Conservation Research
Citation (APA)
Tesfaye, A., Brouwer, R., van der Zaag, P., & Negatu, W. (2016). Assessing the costs and benefits of
improved land management practices in three watershed areas in Ethiopia. International Soil and Water
Conservation Research, 4(1), 20-29. https://doi.org/10.1016/j.iswcr.2016.01.003
Important note
To cite this publication, please use the final published version (if applicable).
Please check the document version above.
Copyright
Other than for strictly personal use, it is not permitted to download, forward or distribute the text or part of it, without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license such as Creative Commons. Takedown policy
Please contact us and provide details if you believe this document breaches copyrights. We will remove access to the work immediately and investigate your claim.
This work is downloaded from Delft University of Technology.
H O S T E D B Y
Original Research Article
Assessing the costs and bene
fits of improved land management
practices in three watershed areas in Ethiopia
Abonesh Tesfaye
a,n, Roy Brouwer
a, Pieter van der Zaag
b,c, Workneh Negatu
d aDepartment of Environmental Economics, Institute for Environmental Studies, VU University Amsterdam, De Boelelaan 1087, 1081 HV Amsterdam, The Netherlands
b
UNESCO– IHE, Institute for Water Education, Delft, The Netherlands
c
Delft University of Technology, Delft, The Netherlands
d
Addis Ababa University, College of Development Studies, Centre for Rural Development, Addis Ababa, Ethiopia
a r t i c l e i n f o
Article history:
Received 26 August 2015 Received in revised form 11 January 2016 Accepted 15 January 2016 Available online 17 February 2016 Keywords:
Soil conservation Cost–benefit analysis
Cobb–Douglas production function Blue Nile
Ethiopia
a b s t r a c t
Unsustainable land use management and the resulting soil erosion are among the most pervasive pro-blems in rural Ethiopia, where most of the country’s people live, jeopardizing food security. Despite various efforts to introduce soil conservation measures and assess their costs and benefits, it is unclear how efficient these measures are from an economic point of view in securing food production. This paper examines the costs and benefits of three soil conservation measures applied in the country in three different rural districts facing different degrees of soil erosion problems using survey data collected from 750 farm households. A production function is estimated to quantify the costs and benefits of more sustainable land use management practices. We show that the soil conservation measures significantly increase productivity and hence food security. Comparing the costs and benefits, the results indicate that implementing soil conservation measures would benefit farm communities in the case study areas through increased grain productivity and food security.
& 2016 International Research and Training Center on Erosion and Sedimentation and China Water and Power Press. Production and Hosting by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
Contents
1. Introduction . . . 20
2. Study area . . . 21
3. Methodological approach . . . 22
4. Data collection and main assumptions . . . 23
5. Sample characteristics . . . 24
6. Estimated production function . . . 25
7. Cost–benefit analysis . . . 26 8. Discussion . . . 27 9. Conclusions . . . 27 Acknowledgements . . . 28 References . . . 28 1. Introduction
Soil erosion and the resulting agricultural land degradation are the most severe environmental problem in the Ethiopian high-lands (Amsalu & de Graaff, 2007;Pender, Gebremehedhin, Benin, & Ehui, 2001; Shiferaw & Holden, 1999; Tefera & Sterk, 2010), jeopardizing the sustainability of agricultural production and Contents lists available atScienceDirect
journal homepage:www.elsevier.com/locate/iswcr
International Soil and Water Conservation Research
http://dx.doi.org/10.1016/j.iswcr.2016.01.003
2095-6339/& 2016 International Research and Training Center on Erosion and Sedimentation and China Water and Power Press. Production and Hosting by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
nCorresponding author.
E-mail addresses:abonesh.tesfaye@gmail.com(A. Tesfaye),
roy.brouwer@vu.nl(R. Brouwer),p.vanderzaag@unesco-ihe.org(P. van der Zaag),
wnegatu@yahoo.com(W. Negatu).
Peer review under responsibility of International Research and Training Center on Erosion and Sedimentation and China Water and Power Press.
ultimately national food security (Kassie, Holden, Kohlin, & Bluff-stone, 2008;Sonneveld & Keyzer, 2002). The on-site effects are a major source of concern since they threaten the livelihoods of a majority of the country's population. The highlands of Ethiopia cover 40 percent of the country’s land mass and are home to al-most 88 percent of its human population and 70 percent of the total livestock population (Ayele, 1999). The causes underlying land degradation are a combination of climate conditions and extreme weather events, such as heavy rainfall and droughts, population pressure, unsustainable agricultural land use practices such as overgrazing, cultivation of steep slopes, and no or limited fallow periods (Geist & Lambin, 2004), and lack of institutions to enact regulations or laws that enhance sustainable land manage-ment practices (FAO, 2011).
The problem is transboundary in nature particularly in the upper Blue Nile basin where soil and excessive runoff that leave the boundary of individual farms cause off-site or off-farm impacts to reservoirs, irrigation schemes and waterways downstream across political borders. An example is the sedimentation of the Gezira irrigation scheme in Sudan due to massive erosion from the Upper Blue Nile river basin.Ahmed (2003) reported that the sediment load of the Blue Nile at the border at El Diem (120 km upstream of the El-Roseires Dam) is 140 million tons per year, causing man-agement difficulties of irrigation canal networks in the Gezira scheme and consuming more than 60 percent of the total costs of the operation and maintenance in sediment clearance.
In the Ethiopian highlands, topsoil loss due to soil erosion is estimated to be 1.5 billion tons per year (Taddese, 2001), and average annual soil loss from cultivated land is 42 t/ha (Hurni, 1993). This is very high compared to other countries worldwide (Pimentel, 2006). Total estimated soil erosion in the US, for ex-ample, a country 9 times the size of Ethiopia is 3 billion t/year (Carnell, 2001 cited inPimentel, 2006). The estimated soil forma-tion rate in Ethiopia is less than 2 t/ha/year, which is very low compared to the estimated soil erosion rates (Hurni, 1983). Worldwide soil erosion rates are highest in Asia, Africa and South America, averaging 30–40 t/ha/year, and lowest in the United States and Europe, averaging about 17 t/ha/year (Barrow, 1991). Studies conducted in the Amhara region confirm that soil loss due to ero-sion has a significant impact on the decline of crop yield and loss of agricultural land (e.g.Ludi, 2002; Shiferaw & Holden, 1999; Son-neveld, 2002). In order to mitigate the problem of soil erosion, the regional government and non-governmental organizations have supported various efforts to introduce soil conservation measures.
A number of studies exist that investigated the costs and ben-efits of soil conservation measures in east Africa in general and Ethiopia in particular. However, the empirical evidence base is ambiguous. There does not seem to be a straightforward answer to the question whether soil conservation measures are economically efficient, that is, whether the benefits of using soil conservation measures outweigh the cost of these measures.Tenge, De graaff,
and Hella (2005)for example, found that in Tanzania the costs of
establishing bench terraces, grass strips and fanya juu bunds exceed the returns in the initial two years. However in the long term, the three soil and water conservation measures are profitable to farmers on gentle to moderate slopes and with low to medium opportunity costs of labour. It was also found that soil and water conservation measures are notfinancially attractive to most farmers with off-farm activities and other sources of income. In Kenya
Kauffman et al. (2014)estimated the effect of 11 soil conservation measures on soil erosion and three ecosystem services that is food production, water availability and energy production acting as provisioning services. Modelling indicated that the three ecosystem services could be improved, as compared with the base level, by up to 20 percent by introducing appropriate conservation measures with benefit/cost relations of around 7. However, farmers were
unable to make the necessary investments and much effort and many institutional studies were needed to achieve progress towards implementation. Whereas in Ethiopia studies by Gebremedhin,
Swinton, and Tilahun (1999), Shiferaw and Holden (2001), Ludi
(2002)andKassie et al. (2008)report that combined soil and water conservation measures benefit farmers only in low rainfall areas as these measures primarily serve the purpose of water harvesting in such areas. The research carried out byBekele (2005)andKassie et al. (2008)on the other handfind that in high rainfall areas soil conservation measures only become profitable if the land lost be-cause of the construction of these measures on the land such as bunds is compensated through the planting of grass for livestock fodder and trees for fuel and fruits on these bunds. These studies employed a variety of different approaches, such as ANOVA, sto-chastic dominance analysis, matching methods, and damage cost functions to estimate the costs and benefits of soil conservation measures. In the case of ANOVA, group means are compared based on estimated crop yields on plots with and without soil conserva-tion measures and tested for their statistical significance. Stochastic dominance analysis compares and ranks the expected net returns from crop production with and without soil and water conservation measures to assess the most likely profitable plot treatment. Matching is a statistical technique which is used to evaluate the effect of a treatment by comparing the treated and the non-treated units in an observational study or quasi-experiment (i.e. when the treatment is not randomly assigned). The goal of matching is, for every treated unit, tofind one (or more) non-treated unit(s) with similar observable characteristics against whom the effect of the treatment can be assessed. By matching treated units to similar non-treated units, matching enables a comparison of outcomes among treated and non-treated units to estimate the effect of the treatment without reduced bias due to confounding. Hence, matching methods examine how crop yields and productivity in-dicators on plots with and without soil conservation measures differ based on a search procedure to match comparable plots focusing on key plot and climate characteristics such as soil conditions and precipitation. Damage cost functions estimate the monetary value of the loss of crop yield based on soil erosion rates on plots without soil conservation measures. Unlike the different methods reviewed above, one methodological similarity to our research was a study by
Kato, Ringler, Yesuf, and Bryan (2011)who applied the Just and Pope framework using a Cobb– Douglas production function to explore the effect of soil and water conservation technologies on crop yields in different regions and rainfall zones in Ethiopia. Their result in-dicates that soil and water conservation investments perform dif-ferently in different rainfall areas and regions of the country.
The main objective of this study is to inform land use policy in Ethiopia based on the estimation of a Cobb–Douglas production function using a stratified rural household survey and farmers’ self-reported costs and benefits of soil conservation measures. The functional relationships embodied in the estimated production function help us to identify the direct contribution of the soil conservation measures to agricultural productivity by isolating their effect from other factors. Moreover, while some work has already been done in estimating the costs and benefits of soil conservation measures at farm household level, there has been no attempt to address the costs and benefits of these measures at the wider watershed level. Hence, this study tries to fill this in-formation gap by estimating the costs and benefits of soil con-servation measures in the whole Gedeb watershed in Ethiopia.
2. Study area
The Blue Nile basin is the second largest basin in Ethiopia comprising 17 percent of the surface area (176,000 km2) (Conway,
2000). The basin drains a large portion of the central and south-western Ethiopian highlands. Amhara, Oromiya and Benishangul-Gumuz are the three most important administrative regions lo-cated in the basin.
There is high sheet erosion in the basin due to the steep slope and the high rain fall especially around Mount Choke (Hydrosult Inc et al., 2006). The case study was conducted in three different districts in the Gedeb watershed in the Upper Blue Nile basin
(Fig. 1). Senan is a high land (43500 MASL), Gozamin is located
between the low and highlands at an altitude of 2000–2500 MASL, and Machakel is a lowland (1500 MASL) (Hurni, 1998). The three locations are found at different slope gradients: Senan district is between 15 and 50 percent, whereas Gozamn and Machakel dis-tricts are between 8 and 15 percent slope gradient on average.1
The Gedeb watershed was selected as the case study area, be-cause it is highly degraded due to soil erosion (Emrie, 2008), and covers different altitudes with varying erosion rates. The Gedeb watershed, located in the Amhara regional state approximately 300 km north-west from Addis Ababa, covers a total area of 871 km2 with an estimated population of 495 thousand (CSA,
2007). The area has humid to sub-humid climatic conditions with mean annual rainfall ranging from 920 to 1649 mm. The average temperature varies between 7.5 and 22.5°C (MoA, 2000). The soil types of the watershed vary from Humic Nitosols to Chromic Lu-visols (MoA, 2000). Agriculture is the way of life for more than 80 percent of the households living in the watershed. The farming
system is mixed crop–livestock subsistence farming. The main crops grown in the area include tef (Eragrostis tef), wheat, barley, potato and Engedo (Avena sativa) and almost all farm households also own some livestock (e.g. chicken, goats and cows).
The main soil conservation measures used in the study area are soil, stone and fanya juu bunds (Herweg & Ludi, 1999). Soil and stone bunds are ridges and ditches made of soil or stone, dug across the slope along the contour. Fanya juu is a type of terrace adopted from Kenya. In Swahili, fanya means ‘throw’ while juu means‘up’. It thus means, ‘throwing up the slope’ as opposed to ‘throwing down the slope’ in the conventional soil bunds con-struction. With fanya juu, less cultivable land is taken up by the structure and benching is faster than the conventional soil bunds. However, fanya juu requires more labour input (Desta, Kassie, Benin, & Pender, 2000). According to the information provided by the farmers interviewed in the case study area, these soil con-servation measures are implemented on all types of cropland based on the technical guidance offered by local development agents. No policies exist to prioritize more or less degraded or fertile lands in the case study area.
3. Methodological approach
In order to inform future land use policy, a cost–benefit analysis (CBA) was carried out (e.g.Dinwiddy & Teal, 1996;Gittinger, 1982;
Pearce, 1987), in which the privatefinancial consequences of
im-plementation of the three main soil conservation measures were estimated and compared to a situation in which farmers take no soil conservation measures. We focus on the private costs and
Fig. 1. Gedeb watershed.
1
The information about the slope of the area was obtained from the officials of Senan District Agricultural and Rural Development Office.
A. Tesfaye et al. / International Soil and Water Conservation Research 4 (2016) 20–29 22
benefits at farm household level to test the impact of soil con-servation measures on agricultural output in the case study area. Sustainable land management using the soil conservation mea-sures is also expected to have a number of additional, broader socio-economic consequences, such as reduced soil runoff into the Blue Nile River and associated sedimentation andflood risks, but these were not included in the CBA. The central hypothesis we aim to test here is whether the private costs outweigh private benefits when taking soil conservation measures, as often is suggested, but about which the empirical evidence base is ambiguous. We are foremost interested in identifying the private costs and benefits to farmers to support policy and decision-making towards sustain-able land use management. If the benefits of soil conservation measures to the farmer can be shown to be higher than their costs, this should induce farmers to implement such measures. Existing institutional-economic conditions (e.g. lack of land use rights) may still prohibit the adoption of soil conservation measures, justifying a land use policy intervention, but not taking these measures based on economic efficiency considerations can then be ruled out. Another reason for a public policy intervention would be the ex-tent of the avoided damage costs due to reduced siltation and flood risks downstream. However, these damage costs are cur-rently unknown, hence the focus here on the private costs and benefits only.
In order to inform the CBA, a Cobb–Douglas (CD) production function was estimated. This production function is widely used to describe the technical relationship between the inputs and out-puts of a production process (e.g.Coelli, Rao, & Battese, 1999). In this study we regress individual farm grain2 yield productivity
(output) measured in kilograms per hectare per year on standard input factors for grain cultivation, such as land, labour, capital, consumables such as fertilizer use, and whether or not farmers take specific soil conservation measures.3The general functional
form is presented as follows:
∏
= ( )
Y Xibi 1
where Y is the grain yield in kilograms per hectare per year, Xiare
the factor inputs (also measured per hectare per year), the effects of which on productivity are estimated through the associated coefficients bi. The log-transformed functional form is
∑
( ) =Y b ( )X ( )
ln iln i 2
The estimated coefficients for the input factors can be used to assess the marginal effect of the specific input factors on crop productivity. Given the fact that farm households provide their own unpaid labour as input, the estimated CD production function allows derivation of a shadow wage rate, which reflects the mar-ginal product of labour (e.g.Gittinger, 1982;Jacoby, 1993). In the case of the soil conservation measures, dummy variables are in-cluded for the specific measures farmers take to protect their land against erosion. The baseline category here is a control group of farmers who do not implement soil conservation measures. Due to their log-transformation, the regression coefficients can be inter-preted as the relative (percentage) change in crop productivity as a result of the implementation of one of the soil conservation measures.
The estimated coefficients are subsequently used to calculate the potential increase in yields for farmers who do not (yet) take any soil conservation measures in vulnerable zones throughout the Gedeb watershed. The definition of a vulnerable zone is based
here onHurni (1986), i.e. areas where soil losses exceed soil for-mation rates of 12.5 t/ha/year. Based on inforfor-mation from Agri-cultural and Rural Development District Offices ofGozamn, Senan,
and Machakel (2009), 104.9 thousand hectares of cropland in the
Gedeb watershed are in vulnerable zones that are sensitive to soil erosion and in urgent need of protection.
4. Data collection and main assumptions
The estimation of the CD production function and the CBA are based on face to face interview conducted with the local language using enumerators recruited from the area in July 2009 from 750 farm households living in the three districts in the Gedeb wa-tershed Gozamn, Machakel and Senan. A stratified random sam-pling technique was used to recruit sample respondents. The three different districts werefirst selected based on their altitude and soil erosion problems, followed by a random selection of villages within each district, and households within each village.
Out of the 750 farmers (250 in each district), 500 reported to have implemented one of the main soil conservation measures: soil, stone or fanya juu bunds on their land over the pastfive years (2004–2008). These 500 farmers were asked to report the differ-ent types of costs related to the soil conservation measures, in-cluding the opportunity costs of the land needed to construct the bunds, the costs of equipment and maintenance costs over those 5 years. Not all farmers reported their opportunity costs of taking soil conservation measures. The opportunity costs of forgone crop production were therefore calculated by the authors based on the average slope of the farm lands in the different districts. This latter information was provided by the Agricultural and Rural Develop-ment District Offices. The steeper the slope, the smaller the spa-cing between the bunds and hence the higher the share of the area needed to implement the measures. The slope gradient for Senan is between 15 and 50 percent, while Gozamn and Machakel have an average slope gradient of 8–15 percent. Based on existing guidelines4 for terrace construction (Hurni, 1986; Shiferaw &
Holden, 1999), land and corresponding crop losses due to the
construction of soil conservation bunds were assumed to be equal to 15 percent of the area size in Senan and 8 percent in Machakel
Table 1
Input data for the estimation of the costs and benefits of soil conservation measures in the Gedeb watershed.
Variable Measurement unit
Investment cost of soil bunds construction USD 28.55/ha
Investment cost of stone bunds construction USD 32.65/ha
Investment cost of fanya juu bunds construction USD 86.66/ha
Maintenance cost of soil bunds USD 5.19/ha/year
Maintenance cost of stone bunds USD 1.73/ha/year
Maintenance cost of fanya juu bunds USD 6.05/ha/year
Yield increment from implementation of soil bunds 18.77 kg/ha/year
Yield increment from implementation of stone bunds 11.00 kg/ha/year
Yield increment from implementation of fanya juu bunds 25.46 kg/ha/year
Area of crop land in need of soil protection in the Gedeb watersheda
104,917 ha
Average market price of grain (wheat)b USD 57.70/100 kg
Average yield in the three Gedeb watershed districts 10975 kg/ha/year
Note: All the information is based on the survey data, with the exception of data in footnotes given below.
aAgricultural and Rural Development District Offices of Gozamn, Senan and
Machakel (2009).
b
FAO (2009).
2
Grain yield here refers to wheat grain. Wheat is selected since it is one of the major crops in the area.
3
In order to select the relevant variables a stepwise regression procedure is used which is a combination of backward elimination and forward selection.
4This guideline is used to crosscheck the accuracy of the replies given by
re-spondents in relation to input they reported to use for the implementation of soil conservation measures.
and Gozamn. We assume here that no income is generated from the grass strips on the bunds, which is a relatively strong as-sumption compared to the existing literature (e.g.Gebreselassie, Amdemariam, Haile, & Yamoah, 2009), but gives us a conservative estimate of the benefits involved.
The reported cost of equipment consisted of the amount of money spent on the purchase of a shovel. The shovel is the only instrument used for the construction and maintenance of soil conservation measures in the study area. Because the shovel serves various other purposes in farming too and is not only used for soil conservation measures, only 50 percent of the total pur-chase cost were attributed to the soil conservation costs.
The CD regression coefficients for the soil, stone and fanya juu bunds provide information about the percentage change in pro-ductivity due to the implementation of these measures. The a priori expectation is that this change is positive. This percentage increase is thenfirst multiplied by the average yield in each of the three districts where the data were collected and secondly by the average price of grain in each of the three districts. The monetized revenues of the increase in yield due to the implementation of the three soil conservation measures were assumed to remain con-stant throughout the planning horizon.
Finally, the present value of the costs and benefits were
estimated using a social discount rate of 8 percent, the minimum official borrowing rate in the Commercial Bank of Ethiopia in 2009, and a planning horizon of 12, 8 and 27 years for soil, stone and fanya juu bunds respectively. Information regarding the lifetime of the three soil conservation measures was provided by the Agri-cultural and Rural Development District Offices of Gozamn, Senan and Machakel.
The comparison of the Net Present Value (NPV) of the three measures is based on the longest life time period of 27 years. In order to make the three measures comparable over this time period, farmers face investment costs for soil bunds twice and three times for stone bunds.
The main input data into the cost–benefit analysis and their information sources are summarized inTable 1.
5. Sample characteristics
The main sample characteristics are presented inTable 2. More than 90 percent of the respondents were male household heads with an average age of 45 and family size of 5. Comparing households with and without soil conservation measures, the average age of farmers implementing soil conservation measures was slightly, but significantly higher (46) compared to the group not taking any measures (42). Both groups have the same average family size. There is a statistically significant difference between households with and without soil conservation measures with regards to land size. Those taking measures tend to have slightly bigger land holdings (0.99 ha) compared to those without mea-sures (0.85 ha). The former also have significantly more labour available to take soil conservation measures than the latter. Moreover, a statistically significant difference in yield per hectare per year can be observed between the two groups, even though both groups use the same amount of fertilizer5 on their land.
Table 2
Sample summary statistics for the Gedeb watershed.
Variable Description Farmers without soil conservation
measures
Farmers with soil conservation measures
All farmers M-W Test
N¼250 N¼500 N¼750
Mean Std. dev. Mean Std. dev. Mean Std. dev. Z-value
Average farm income Birr/year 10,985.6 13,772.4 8900.7 8584.2 9595.7 10,635.7 0.7
Average cropland size ha 0.85 0.50 0.99 0.50 0.94 0.51 3.9***
Average crop yield kg/year 7443.7 3957.8 10,974.7 5074.9 9797.7 5014.0 9.2***
Average fertilizer use kg/ha/year 66.0 16.4 66.7 19.9 66.5 18.8 0.0
Average age Years 42 14.1 46 14.3 45 14.4 3.3***
Average available labour Persons 7.0 3.9 9.1 4.1 8.4 4.2 6.7***
Average family size Persons 5 1.8 5 1.8 5 1.8 0.8
Average off farm income Birr/year 408.2 761.4 241.1 564.7 296.8 641.4 2.8**
Average livestock Tropical Livestock
Unitsa
4.8 3.4 4.3 3.1 4.5 3.2 1.4
Share male % 94 92 93
Share literate % 49 58 55
Share from Gozamn % 25 34 31
Share from Senan % 45 31 36
Share from Machakel % 30 35 33
Share with soil bunds % 71
Share with stone bunds % 12
Share with fanya juu bunds
% 17
Share with fertile land % 83 68 73
***
Significant at 1%.
**Significant at 5%. a
1 TLU¼250 kg life weight.
Table 3
Major crops and land allocation in the study area. Major crops Percentage of households
cultivating
Average cropland in hectare per household N¼750 Wheat 100 0.31 Teff 95.5 0.26 Barley 65 0.21 Engedo 52 0.09 Other 0.07 Total 0.94a a
The average land holding for the whole sample (with and without soil con-servation measures) is 0.94 ha.
5
Farmers mainly use inorganic fertilizer such as Urea and DAP in the study area. The main reason is the fact that manure is used as fuel for cooking, lighting A. Tesfaye et al. / International Soil and Water Conservation Research 4 (2016) 20–29
Average yields are almost 50 percent higher for the group taking soil conservation measures. Although farm income is, on average, almost 20 percent lower in the group taking soil conservation measures, this difference is not statistically significant at the 10 percent level. Off-farm household income is significantly higher for the group not taking soil conservation measures. No significant difference is found between the two groups in terms of livestock holding.
Major crops grown in the area include wheat, tef (Eragrostis tef), barley, potato and Engedo (Avena sativa). All the farmers reported that they grow wheat. When we look at the proportion of cropland covered by major crops, wheat covers 0.31 ha of their cropland while teff and barely takes 0.26 and 0.21 ha respectively.Table 3
reports the detail.
Finally, the 500 households who have implemented soil con-servation measures are more or less equally distributed across the three districts. Thirty-four percent were found in Gozamn, 31 percent in Senan and 35 percent in Machakel. Soil bunds are the most widely implemented conservation measures, followed by fanya juu and stone bunds.
6. Estimated production function
The CD production function was estimated in STATA version 11. All the 750 sampled farmers were willing to respond to the survey questionnaire resulting in a 100 percent response rate. The re-gression results are presented inTable 4. The model is highly sig-nificant: the null hypothesis of non-significance of the entire model is convincingly rejected at the one percent level by the chi-square test. The explanatory (independent) factors included in the model help to explain 46 percent of the variation found in the dependent variable (individual farmer crop yield per hectare per year based on the self-reported 5-year average). The baseline category for the regression model consists of farmers who take no soil conservation measures and live in the midland district Gozamn.
All the relevant input factors land, labour,6 fertilizer, and the
different soil conservation measures (soil bunds, stone bunds, fanya juu bunds) are significant at the one percent level. Due to the lack of available monitoring data, land quality conditions are measured based on farmers’ perceived land fertility (included here as a
dummy variable whether the land is perceived as fertile or not). Hence, the model results show that whilst controlling for other influencing factors, the implementation of soil, stone and fanya juu bunds on agricultural land significantly contributes to increased crop productivity. The positive coefficients of the three soil con-servation measures reflect the relative increase in crop productivity due to their implementation on one hectare of land. Fanya juu bunds have the highest impact on crop productivity (almost 24 percent), followed by soil bunds (17 percent) and then stone bunds (10 percent). However, the observed differences between fanya juu and soil bunds and soil and stone bunds is not statistically sig-nificant at the 5 percent level based on the Wald test,7 only the
difference between fanya juu and stone bunds. A significant positive relationship exists as expected between productivity and land size, labour, and fertilizer input. Hence, as the amount of productive land, labour and fertilizer input increases, so does productivity. Some studies have found an inverse relationship between productivity and land size (e.g. Ansoms, Verdoodt, & Van Ranst, 2008; Carter, 1984). However, in rural Ethiopian areas land is a major constraint in production and most farm households typically have relatively little land (no more than one hectare), meaning that an increase in land results in an almost proportionate increase in productivity.
Table 4
Estimated Cobb–Douglas production function for the Gedeb watershed.
Variable Description Coefficient estimate Standard error
Constant 2.695***
0.136
Land size Area size in ha 1.038***
0.075
Labour input Family and hired labour in man-days per ha 0.385***
0.028
Fertilizer application kg of fertilizer applied per ha 0.294***
0.056
Condition of land Dummy (1¼fertile) 0.063*** 0.016
Soil conservation measure
Soil bunds Dummy (1¼soil bunds implemented) 0.175***
0.027
Stone bunds Dummy (1¼stone bunds implemented) 0.100***
0.029
Fanya juu bunds Dummy (1¼fanya juu implemented) 0.237***
0.041 District
Senan Dummy (1¼farmer lives in Senan) 0.167***
0.026
Machakel Dummy (1¼farmer lives in Machakel) 0.045 0.027
Fanya juu bunds Senan Dummy (1¼farmer in Senan implementing fanya juu) 0.174*** 0.062
Adjusted R Squared 0.461
Chi-squared F-test 36.62***
Number of observations 750
***Significant at 1%.
Table 5
Summary results of the cost–benefit analysis of soil conservation measures in the Gedeb watershed.
Soil bunds Stone bunds Fanya juu bunds
Costsa 18,514,703 14,269,761 24,326,056 Benefitsa 27,270,757 15,981,797 36,990,595 NPV 2,333,965 760,031 452,496 B–C ratio 1.24 0.90 1.03 IRR (%) 17.6 4.3 8.6
Note: The discount rate used in the analysis is 8 percent and the life time of the conservation measures is 27 years.
NPV: Net Present Value. B–C ratio: discounted Benefit–Cost ratio. IRR: Internal Rate of Return.
a
Undiscounted costs and benefits in 2010 US dollars.
(footnote continued)
and heating houses as is the case in most rural parts of the country (Mekonnen & Köhlin, 2008;Tucho & Nonhebel, 2015).
6
The estimated shadow wage rate of the farm household is US$0.92.
7
Wald test is a way of testing the significance of particular explanatory vari-ables in a statistical model. It is one of a number of ways of testing whether the parameters associated with a group of explanatory variables are zero. If for a par-ticular explanatory variable, or group of explanatory variables, the Wald test is significant, then we would conclude that the parameters associated with these variables are not zero, so that the variables should be included in the model. If the Wald test is not significant then these explanatory variables can be omitted from the model (Agresti, 1990).
Similarly, the perceived fertility conditions of the land contribute significantly positive to its productivity. As expected, more fertile land results in a higher productivity.
The relative contribution of the other production factors is high, but less than proportionate and substantially less than that of land in this study. Also the district dummy for Senan is significant, and the interaction term between Senan and fanya juu bunds. No other significant interactions could be detected. Also other farm household characteristics such as age or amount of livestock did not appear to have a significant impact on crop productivity. The coefficient estimate for the fanya juu bunds in Senan result in significantly lower yield efficiency than fanya juu bunds in the other districts. This is probably due to the fact that Senan has the highest slope gradient among the three districts and fanya juu bunds are not the type of measure recommended for such areas
(Hurni, 1986). Compared to farmers not taking soil conservation
measures in the midland district Gozamn, farmers taking soil conservation measures in Senan benefit through an increase in crop productivity of 16 percent. The dummy variable for Machakel is not significant at the conventional 5 percent level. Hence, no significant difference can be detected in yield efficiency between farmers living in the lowland Machakel and midland Gozamn.
Finally, the increasing return to scale in our estimation of the Cobb–Douglas production function is consistent with a couple of other studies conducted in Ethiopia and neighbouring country. For example,Fekadu and Bezabih (2009)who studied technical ef fi-ciency of wheat production in Machakel woreda estimated the return to scale from wheat production to be 1.08 which indicates increasing return to scale. Similarly, a study by Ali, Imad, and
Yousif (2012)in Northern Sudan found the sums of elasticities of
wheat production to be 2.7 and 2.4 in two different localities showing increasing return to scale.
7. Cost–benefit analysis
Costs and benefits were scaled up by multiplying the indicator values presented inTable 1with the area size in need of additional soil conservation measures in the Gedeb watershed. Under the assumption that soil, stone and fanya juu bunds have an effective life time of 12, 8 and 27 years respectively and future costs and benefits are discounted at an 8 percent discount rate, and con-sidering consecutive years of up to 27 years in calculating the NPV, the net benefit is positive for soil and fanya juu bunds, but not for stone bunds (Table 5). The benefit–cost ratio is highest for soil bunds, while the ratio is just above the break-even point for fanya juu bunds and is less than one for stone bunds.
Scaling up the present values of the costs and benefits of the soil bunds for the whole Gedeb watershed over the next 27 years across 104.9 thousand hectares of agricultural land in need of protection results in a NPV of 2.3 million US$. The implementation of fanya juu bunds is more costly, but still results in a NPV of 0.45 million US$ for rural farm households in the Gedeb watershed. The implementation of stone bunds is not profitable. The internal rate of return is as high as 17.6 percent for soil bunds, meaning that the discount rate can increase up to 17.6 percent before costs and benefits of soil bunds implementation reach their break-even point. This internal rate of return is 8.6 percent for fanya juu bunds. Hence in the latter case, the discount rate cannot be more than 8.6 percent in order for the discounted benefits to be higher than the discounted costs (financial investments in soil conservation measures become unprofitable when they dive under the break-even benefit-cost ratio of 1).
The most important reason for the higher profitability of soil bunds compared to fanya juu and stone bunds is the fact that the investment cost of soil bunds is less than the other two (Herweg &
Ludi, 1999), which also leads soil bunds to be the most commonly
implemented soil conservation measure throughout the wa-tershed (Tesfaye & Brouwer, 2012). Fanya juu bunds appear to be less profitable in this analysis compared to soil bunds since the implementation of fanya juu bunds is more labour intensive than soil and stone bunds, but at the same time it is also the most ef-fective conservation measure in reducing soil erosion (Herweg & Ludi, 1999;Tenge, De Graaff, & Hella, 2004).
A number of tests were carried out to see how sensitive the outcomes of the CBA are to the assumptions made. This includes varying the one-off investment and annual maintenance costs, possible changes in the market price of grains in Ethiopia and their impacts on the estimated benefits of soil conservation measures, the effect of the project lifetime and higher discount rate on the investment decision. If the costs of the soil conservation measures increase only a little bit, the benefit–cost ratio for fanya juu bunds
Fig. 2. Sensitivity of the benefit–cost ratios for different soil conservation measures to varying cost.
Fig. 3. Sensitivity of the benefit–cost ratios for different soil conservation measures to varying grain price.
Fig. 4. Sensitivity of the benefit–cost ratios for different soil conservation measures to varying lifetimes.
A. Tesfaye et al. / International Soil and Water Conservation Research 4 (2016) 20–29 26
rapidly dives under the threshold line of 1 (all other things re-maining equal ). The benefit–cost ratio for stone bunds was al-ready less than one and will therefore only become lower. Re-markably, the implementation of soil bunds remains beneficial even if costs increase by as much as 45 percent. Hence, the positive results for soil bunds are fairly robust when changing the cost assumptions. The benefit–cost ratio for soil bunds dives under the threshold line of 1 if the cost increases by 50 percent (Fig. 2).
Similarly, the market price for grain crops has to fall by almost 20 percent for the investment in soil bunds to become unprofitable (all other things remaining equal). Again, the decision to invest in fanya juu bunds is much more sensitive to the overall grain price. If the price goes down by even less than 10 percent, the benefit–cost ratio drops below one. Stone bunds becomes profitable if the grain price increases from its current level by more than 20 percent in the future (Fig. 3).
The results from the CBA also appear to be sensitive to the length of time over which soil conservation measures remain ef-fective. This is shown inFig. 4. Soil bunds become profitable after 7 years. This is 21 years for fanya juu bunds while for stone bunds to become profitable their lifetime has to be extended to almost 48 years.
Finally, the sensitivity of the CBA results is also tested by in-creasing the discount rate by as much as 20 percent. The benefit– cost ratio for soil bunds drops under the threshold when the discount rate increases by 18 percent while the ratio for stone and fanya juu bunds indicate a loss at 5 and 9 percent discount rates respectively. These results are illustrated graphically (Fig. 5).
8. Discussion
Land use policy in Ethiopia has long failed to provide the right incentives to farmers to invest in sustainable land use management practices and hence food security. There is growing consensus that many of the soil conservation programmes in the past were dis-appointing and ineffective for various reasons. They used aflawed ‘environmental narrative’ to promote large scale top-down inter-ventions, gave inadequate consideration to farmers’ perspectives, constraints, and local conditions, provided limited options to farmers, and in some cases even promoted unprofitable alter-natives. Studies (e.g.Anley, Bogale, & Haile-Gabriel, 2007;Bekele & Drake, 2003; Shiferaw & Holden, 1998)find that the diffusion of information about available technological options or rather the lack thereof has a significant effect on soil conservation investment decisions.
We show in our study that whilst controlling for other pro-duction factors, taking soil conservation measures significantly
increases crop productivity. Implementing fanya juu bunds in-creases– all other things being equal – crop productivity by almost 24 percent, followed by soil bunds, which increases crop pro-ductivity by 17 percent. Stone bunds are least effective as expected in this high rainfall case study area. Stone bunds have been shown to become especially interesting in dry areas where they also play an important role in water harvesting (Gebremedhin et al., 1999;
Kassie et al., 2008). Moreover, unlike soil and fanya juu bunds, the implementation of stone bunds depends crucially on the avail-ability of stones which in our case study area are in short supply. In the first year of construction of soil conservation measures, farmers may perceive yield loss. However, the perceived increase in yield within a few years may encourage them to continue to adopt these measures. Though we did not research about farmers perception about the yield loss in thefirst year in our study, the result of Wolka, Moges, and Yimer (2013) indicate that farmers assume that in the absence of bunds, the entire farmland area would suffer a significant reduction in yield, and that crop pro-duction improves as result of their construction.
Accounting for this increase in crop productivity in the CBA and the opportunity costs of constructing the soil conservation mea-sures, including the marginal value of land and the shadow price of family labour, the implementation of especially soil bunds ap-pears to result in a substantialfinancial welfare improvement for farm households under current conditions and is hence highly recommendable. The internal rate of return of the investment in soil bunds is as high as 17 percent. So, even if the external social costs and benefits of soil erosion and conservation respectively (which are expected to be large but unknown) are not accounted for, we show that solely based on the private benefits of soil conservation measures in terms of increased crop revenues, they clearly outweigh their implementation and maintenance costs. The study furthermore indicates that a longer term perspective is needed for these private welfare gains to materialize. The benefits of soil bunds are also fairly robust to changes in the costs of im-plementation or changes in crop price levels.
A third of the sampled farmers reported that they do not im-plement these measures. Lack of incentives and information seem to be the main factors to hold back some of the farmers from in-vesting in soil conservation measures. Our study shows that awareness raising among farmer communities with respect to the benefits of sustainable land use management seems crucial. In another study,Tesfaye, Negatu, Brouwer, and Van der Zaag (2014)
report that among the main driving forces behind farmers’ deci-sion to implement soil conservation measures are access to credit to pay for the initial investment costs, educational background of the household head, and farmers’ perception of their land’s ferti-lity. This result is in agreement with Tenge, Okoba, and Sterk (2007)who reported that availability of credit facilities is an im-portant incentive for farmers to invest on soil and water con-servation measures. Similarly, bigger plots contribute significantly to the farmer’s investment decision. Moreover, farmers with better access to information would be more willing to invest in soil conservation measures. Labour is also one of the crucial inputs for the implementation of soil conservation measures (Tesfaye et al., 2014). These findings are in line with the results presented in
Shiferaw and Holden (1999),Enki, Belay, and Dadi (2001),Asrat,
Belay, and Hamito (2004),Tadesse and Belay (2004),Anley et al.
(2007), De Graaff et al. (2008) and Tiwari, Sitaula, Nyborg, and
Paudel (2008).
9. Conclusions
Unsustainable land use management and the resulting soil erosion are among the most pervasive problems in rural Ethiopia,
Fig. 5. Sensitivity of the benefit–cost ratios for different soil conservation measures to varying discount rates.
where most of the country’s people live, jeopardizing food secur-ity. This paper examines the costs and benefits of three soil con-servation measures applied in the country in three different rural districts facing different degrees of soil erosion problems using survey data collected from 750 farm households. A production function is estimated to quantify the costs and benefits of more sustainable land use management practices. We show that soil conservation measures significantly increase productivity and hence food security.
Although we did not account for this in the CBA presented here, advising farmers to plant grass for fodder or trees for wood fuel or fruit on top of the soil and fanya juu bunds and on their field border could partly offset the yield loss due to the implementation of the measures.
Finally, two important caveats of this study relate to our re-liance on self-reported costs and benefits of soil conservation measures. First, our study is not directly comparable with existing, more detailed agronomicfield studies at agricultural plot level. We derived the impacts from soil conservation measures through a more coarse statistical analysis based on afive year cross-section comparison of farmers who reported that they take soil con-servation measures and farmers who reported that they do not. It is important to compare and where possible combine both sources of information over a longer period of time to identify the pro-ductivity effect of soil conservation measures and their net bene-fits. However, taking into account our bigger sample size (750 respondents) and consistency of our result with similar studies in Ethiopia e.g. Bekele (2005) and Kassie et al. (2008), it is fair to assume that self-reported data could help inform the reader in situation where agronomic experiment is not feasible.
A second caveat relates to the limited availability of informa-tion on slope of individual farm land and soil type which restricted our information on the severity of erosion problem and the ferti-lity condition of cropland. This too is a clear future research need.
Acknowledgements
The Netherlands Organization for Scientific Research (NWO-WOTRO) is gratefully acknowledged for funding this project. We are also grateful for the co-funding received from the Institute for Environmental Studies, VU University Amsterdam. A word of thanks goes to Ted Veldkamp and Jurre Tanja from the VU Uni-versity Amsterdam for their help with the on-site pre-testing of the survey and focus group discussions, and Dr. Melesse Temesgen and Mr. Sebsib Belay from Addis Ababa University for their assis-tance in organizing the pre-test and data collection. Finally, we are grateful to the anonymous reviewers for their valuable comments on previous versions of this paper. As always, the authors remain sole responsible for the content of the paper.
References
Agresti, A. (1990). Categorical Data Analysis. New York: John Wiley and Sons. Ahmed, A. (2003). Sediment transport and watershed management Blue Nile system.
Friend/Nile Project report, Sudan.
Ali, A. A., Imad, E. E., & Yousif, A. K. (2012). Economic efficiency of wheat and faba bean production for small scale farmers in northern state Sudan. Journal of Animal Plant Sciences, 22, 215–223.
Amsalu, A., & de Graaff, J. (2007). Determinants of adoption and continued use of stone terraces for soil and water conservation in an Ethiopian highland wa-tershed. Ecological Economics, 61, 294–302.
Anley, Y., Bogale, A., & Haile-Gabriel, A. (2007). Adoption decision and use intensity of soil and water conservation measures by smallholder subsistence farmers in Dedo district, Western Ethiopia. Land Degradation and Development, 18, 289–302.
Ansoms, A., Verdoodt, A., & Van Ranst, E. (2008). The inverse relationship between farm size and productivity in rural Rwanda. Discussion paper/2008.09. Institute
of Development Policy and Management, University of Antwerp.
Asrat, P., Belay, K., & Hamito, D. (2004). Determinants of farmers’ willingness to pay for soil conservation practices in the southeastern highlands of Ethiopia. Land Degradation and Development, 15, 423–438.
Ayele, G. (1999). Economic analysis of innovation and adoption of vertisol technology: A study of smallholders’ mixed farming in the highlands of Ethiopia. Stuttgart, Germany: University of Hohenheim.
Barrow, C. J. (1991). Land degradation. Cambridge: Cambridge University Press.
Bekele, W., & Drake, L. (2003). Soil and water conservation decision behavior of subsistence farmers in the Eastern highlands of Ethiopia: A case study of Hunde-Lafto area. Ecological Economics, 46, 437–451.
Bekele, W. (2005). Stochastic dominance analysis of soil and water conservation in subsistence crop production in the eastern Ethiopian highlands: The case of the Hunde-Lafto area. Environmental and Resource Economics, 32, 533–550. Carter, M. (1984). Identification of the inverse relationship between farm size and
productivity: An empirical analysis of peasant agricultural production. Oxford Economic Papers, 36, 131–145.
Coelli, T., Rao, D. S. P., & Battese, G. E. (1999). An introduction to efficiency and pro-ductivity analysis. Dordrecht: Kluwer Academic Publishers.
Conway, D. (2000). The climate and hydrology of the Upper Blue Nile River. Geo-graphy Journal, 166, 49–62.
CSA (Central Statistical Authority) (2007). Household income, consumption and ex-penditure survey. Statistical Abstract, Addis Ababa.
De Graaff, J., Amsalu, A., Bodnár, F., Kessler, A., Posthumus, H., & Tenge, A. (2008). Factors influencing adoption and continued use of long-term soil and water conservation measures infive developing countries. Applied Geography, 28, 271–280.
Desta, L., Kassie, M., Benin, S., & Pender, J. (2000). Land degradation and strategies for sustainable development in the Ethiopian highlands: Amhara Region. Socio-eco-nomic and Policy Research Working Paper No. 32. Nairobi: ILRI (International Livestock Research Institute).
Dinwiddy, C., & Teal, F. (1996). Principles of cost–benefit analysis in developing countries. Cambridge: Cambridge University Press.
Emrie, K. (2008). Assessment of land use land cover change and land degradation in Gedeb watershed, choke mountains. Addis Ababa University.
Enki, M., Belay, K., & Dadi, L. (2001). Determinants of adoption of physical soil conservation measures in central highlands of Ethiopia: The case of three districts of northern Shewa. Agrekon, 40, 293–315.
FAO (Food and Agricultural Organization of the United Nations) (2011). Sustainable land management in practice guidelines and best practices for sub-Saharan Africa. Report. Rome, Italy.
Fekadu, G., & Bezabih, E. (2009). Analysis of technical efficiency of wheat produc-tion: A study in Machakel Woreda. Ethiopian Journal of Agricultural Economics, 7, 1–33.
Gebremedhin, B., Swinton, S. M., & Tilahun, Y. (1999). Effects of stone terraces on crop yields and farm profitability: Results of on-farm research in Tigray, northern Ethiopia. Journal of Soil and Water Conservation, 54, 568–573.
Gebreselassie, Y., Amdemariam, T., Haile, M., & Yamoah, C. (2009). Lessons from upstream soil conservation measures to mitigate soil erosion and its impact on upstream and downstream users of the Nile River: intermediate results dis-semination workshop, February 5–6, 2009. Addis Ababa, Ethiopia: IWMI Sub regional Office for East Africa and Nile Basin.
Geist, H. J., & Lambin, E. F. (2004). Dynamic causal patterns of desertification. Bioscience, 54, 817–829.
Gittinger, J. P. (1982). Economic analysis of agricultural projects. London: Johns Hopkins University Press.
Gozamn (2009). Senan and Machakel Agricultural and Rural Development District Offices. Ethiopia: Amhara Regional State.
Herweg, K., & Ludi, E. (1999). The performance of selected soil and water con-servation measures: Case studies from Ethiopia and Eritrea. Catena, 36, 99–114.
Hurni, H. (1983). Ethiopian highland reclamation study. Soil formation rates in Ethiopia. Working paper No. 2. Addis Ababa, Ethiopia.
Hurni, H. (1986). Guidelines for development agents on soil conservation in Ethiopia. Addis Ababa: CFSCDD, MOA.
Hurni, H. (1993). Land degradation, famine, and land resource scenarios in Ethiopia In: D. Pimentel (Ed.), World soil erosion and conservation (pp. 27–62). Cam-bridge: Cambridge University Press.
Hurni, H. (1998). Agroecological Belts of Ethiopia: explanatory notes on three maps at a scale of 1:1,000,000. Research Report, Soil Conservation Research Program. Addis Ababa.
Hydrosult Inc; Tecsult; DHV; and their Associates Nile Consult, Comatex Nilotica; and T and A Consulting (2006). Trans-boundary analysis: Abay– Blue Nile Sub-basin. NBI-ENTRO (Nile Basin Initiative-Eastern Nile Technical Regional Organization).
Jacoby, H. G. (1993). Shadow wages and peasant family labor supply: An econo-metric application to the Peruvian Sierra. Review of Economic Studies, 60, 903–921.
Kassie, M., Pender, J., Yesuf, M., Kohlin, G., Bluffstone, R., & Mulugeta, E. (2008). Estimating returns to soil conservation adoption in the northern Ethiopian highlands. Agricultural Economics, 38, 213–232.
Kassie, M., Holden, S., Kohlin, G., & Bluffstone, R. (2008). Economics of soil con-servation adoption in high-rainfall areas of the Ethiopian highlands Environment for Development (EDF). Discussion Paper Series.
Kato, E., Ringler, C., Yesuf, M., & Bryan, E. (2011). Soil and water conservation technologies: a buffer against production risk in the face of climate change? Insights from the Nile basin in Ethiopia. Agricultural Economics, 42, 593–604.
A. Tesfaye et al. / International Soil and Water Conservation Research 4 (2016) 20–29 28
Kauffman, S., Droogers, P., Hunink, J., Mwaniki, B., Muchena, F., Gicheru, P., Bin-draban, P., Onduru, D., Cleveringa, R., & Bouma, J. (2014). Green Water Credits— Exploring its potential to enhance ecosystem services by reducing soil erosion in the Upper Tana basin, Kenya. International Journal of Biodiversity Science, Ecosystem Servics & Management., http://dx.doi.org/10.1080/
21513732.2014.890670.
Ludi, E. (2002). Economic analysis of soil conservation: Case studies from the highlands of Amhara region, Ethiopia. University of Bern.
MoA (Ministry of Agriculture) (2000). Agro-ecological zones of Ethiopia on 1:2,000,000 scale. Addis Ababa: Natural Resource Management and Regulatory Department.
Mekonnen, A., & Köhlin, G. (2008). Determinants of household fuel choice in major cities in Ethiopia. Environment for Development. Discussion Paper Series, EfD DP 08-18.
Pearce, D. W. (1987). Cost–benefit analysis. London: Macmillan.
Pender, J., Gebremehedhin, B., Benin, S., & Ehui, S. (2001). Strategies for sustainable development in Ethiopian highlands. American Journal of Agricultural Econom-ics, 83, 1231–1240.
Pimentel, D. (2006). Soil Erosion: a food and environmental threat. Environment, Development and Sustainability, 8, 119–137.
Shiferaw, B., & Holden, S. T. (1998). Resource degradation and adoption of land conservation technologies by smallholders in the Ethiopia highlands: A study in Andit Tid, Northern Shoa. Agricultural Economics, 18, 233–247.
Shiferaw, B., & Holden, S. T. (1999). Soil erosion and smallholders’ conservation decisions in the highlands of Ethiopia. World Development, 27, 739–752. Shiferaw, B., & Holden, S. T. (2001). Farm-level benefit to investment for mitigating
land degradation: Empirical evidence from Ethiopia. Environment and Devel-opment Economics, 6, 335–358.
Sonneveld, B. (2002). Land under pressure: Impact of water erosion on food produc-tion in Ethiopia. Amsterdam: Shaker publishing.
Sonneveld, B. G. J. S., & Keyzer, M. A. (2002). Land under pressure: Soil conservation concerns and opportunities for Ethiopia. Land Degradation & Development, 14, 5–23.
Taddese, G. (2001). Land degradation: A challenge to Ethiopia. Environmental Management, 27, 815–824.
Tadesse, M., & Belay, K. (2004). Factors influencing adoption of soil conservation measures in southern Ethiopia: The case of Gununo Area. Journal of Agriculture and Rural Development in Tropics, 105, 49–62.
Tefera, B., & Sterk, G. (2010). Land management, erosion problems and soil and water conservation in Fincha’a watershed, western Ethiopia. Land Use Policy, 27, 1027–1037.
Tenge, A. J., De Graaff, J., & Hella, J. P. (2004). Social and economic factors affecting the adoption of soil and water conservation in West Usambara highlands, Tanzania. Land Degradation & Development, 15, 99–114.
Tenge, A. J., De graaff, J., & Hella, J. P. (2005). Financial efficiency of major soil and water conservation measures in West Usambara highlands, Tanzania. Applied Geography, 25, 348–366.
Tenge, A. J., Okoba, B. O., & Sterk, G. (2007). Participatory soil and water con-servation planning using afinancial analysis tool in the West Usambara high-lands of Tanzania. Land Degradation & Development, 18, 321–337.
Tesfaye, A., & Brouwer, R. (2012). Testing participation constraints in contract de-sign for sustainable soil conservation in Ethiopia. Ecological Economics, 73, 168–178.
Tesfaye, A., Negatu, W., Brouwer, R., & Van der Zaag, P. (2014). Understanding soil conservation decision of farmers in the Gedeb watershed, Ethiopia. Land De-gradation & Development, 25, 71–79.
Tiwari, K. R., Sitaula, B. K., Nyborg, I. L. P., & Paudel, G. S. (2008). Determinants of farmers’ adoption of improved soil conservation technology in a middle mountain watershed of central Nepal. Environmental Management, 42, 210–222. Tucho, G., & Nonhebel, S. (2015). Bio-Wastes as an alternative household cooking
energy source in Ethiopia. Energies, 8, 9565–9583.
Wolka, K., Moges, A., & Yimer, F. (2013). Farmers’ perception of the effects of soil and water conservation structures on crop production: The case of Bokole watershed, Southern Ethiopia. African Journal of Environmental Science and Technology, 7, 990–1000.
