Previous Page Table of Contents Next Page


Economic evaluation of improved Vertisol drainage for food crop production in the Ethiopian highlands

Getachew Asamenew, S.C. Jutzi, Abate Tedla and J. McIntire

International Livestock Centre for Africa (ILCA)
PO Box 5689, Addis Ababa, Ethiopia


Abstract
Introduction
Improved drainage technology and evaluation technique
Results
Discussion and conclusions
Acknowledgements
References


Abstract

The Ethiopian highland Vertisols above 1500 m altitude cover about 7.6 million ha, of which only 1.93 million ha are currently cropped. Because of waterlogging during the growing season, food crop productivity of these soils is low. Surface drainage techniques are traditionally used only in limited areas despite evidence for their having a large impact on crop yield levels and stability.

Land cultivation in the Ethiopian highlands traditionally relies an animal power. However, there is no indigenous animal-drawn surface-drainage implement available. Such an implement (a broadbed maker, BBM) has been developed at ILCA and tested on-station and on-farm. The BBM, a low-cost device based on the Ethiopian arc, establishes 120 cm wide broadbeds-and-furrows (BBF) for effective surface drainage.

This paper reports on an economic evaluation of on-farm trials with the BBM in four highland locations. This evaluation, carried out in 1986, involved 66 individual farmers and four producers' cooperatives on an area of 81 ha. The economic performance of the technology was evaluated using partial budgeting methods.

On the Inewari plateau (2600 m altitude), where farmers traditionally make BBF by hand, the BBM considerably reduced human labour input in BBF construction. As a consequence, return to labour for faba bean and wheat increased by 43% and 140%, respectively.

On the Wereilu plateau (2600 m altitude), where no effective surface drainage is traditionally practised, faba bean and wheat yielded 330% and 131% more grain, respectively, when grown on BBF, than when grown on the traditional flat seedbed. Similar increments in crop residues, a very important animal feed in these areas, were recorded.

At Debre Zeit (1850 m altitude), wheat grain yields were 25% higher on BBF than on flat seedbeds despite 30% less than average (1977-85) rainfall in July and August 1986. This resulted in 40% higher return to labour per hectare.

Similar trends were recorded in the Fogera plains (1850 m altitude), although because of extreme meteorological conditions, with very late onset of heavy rains, differences were not significant.

Introduction

Vertisols (deep black clay soils) cover about 13 million ha in Ethiopia. The highland Vertisols above 1500 m altitude cover 7.6 million ha, of which 1.93 million ha are currently cropped (about 23% of all Ethiopian crop land). Average crop yields are very low on these soils (Berhanu Debele, 1985), mainly due to waterlogging in the growing period, caused by high rainfall and by the high content of swelling clays in these soils.

Traditional methods of surface drainage, such as ridges and furrows or drainage furrows at various spacings, are in general use, but they cannot effectively overcome the waterlogging problem. Very effective surface drainage is, however, practiced on the high elevation Inewari plateau at 2600 m altitude in central Ethiopia. There, raised beds about 80 cm wide, with 40 cm-wide furrows in between, are established each year on about 35 000 ha. This work is done by hand, without the help of any tool, which results in considerable human drudgery and low economic returns to labour.

Despite the long tradition of animal traction in the Ethiopian highlands, and despite the general awareness among farmers of the waterlogging constraint, there is no traditional animal-powered surface-drainage implement available. In many highland Vertisol areas farmers follow a strategy of avoiding much of the waterlogging effect by planting crops late in the season, towards the end of the rains. The crops thus rely on residual soil moisture. This practice implies incomplete utilisation of the growing period, low crop yields and considerable soil losses at the start of the rains due to soil erosion. Large areas of highland Vertisols are presently not cropped because of the waterlogging problem.

Research at the International Crops Research Institute for the Semi-Arid Tropics (ICRISAT) (Kanwar et al, 1982) showed that the key to improved Vertisol utilisation for human food production is effective surface drainage. This was also demonstrated in Ethiopia by ILCA (Jutzi et al, 1987) and the Ethiopian Institute of Agricultural Research (JAR, unpublished data). The potential for large increases in food output from these soils using improved management techniques led to the establishment of a collaborative Vertisol Management Project involving ILCA, ICRISAT and the government of Ethiopia.

An effective, low-cost, animal-drawn surface-drainage implement, based on the Ethiopian arc, was developed within the project in 1985, and has been tested on-station and on-farm. The implement is called the broadbed maker (BBM).

This paper provides an economic evaluation of the improved surface-drainage technology using the BBM on-farm in four Ethiopian highland Vertisol areas. Partial budgeting methods have been used in the absence of whole-farm data.

Improved drainage technology and evaluation technique

The surface drainage implement

A detailed description of the BBM and some on-station performance results have been reported by Jutzi et al (1986). This implement establishes raised beds about 80 cm wide between two furrows of about 40 em width and 15 em depth. The BBM is cheap and can be made locally without much external input. Its cost structure is given in Table 1. The BBM weighs about 30 kg, compared with 20 kg for the conventional Ethiopian plough. Power requirements of the BBM on an adequately ploughed plot are about 50% higher than for the conventional plough, but are still below the potential continuous power deployment of a pair of highland Zebu oxen.

Table 1. Cost structure of the broadbed-maker (BBM), 1986.

Part

Cost (EB)a

2 tines

20

6 bolts (12 x 200 mm)

21

3 bolts (10 x 90 mm)

3

2 metal hooks and 2 metal pins

7

4 wooden wings

8

Other wooden parts

7

Nylon rope

4

Metal chain

10

Total

80

a. US$1 = 2.07 EB.

On-farm implement evaluation

The promising on-station results prompted the joint Vertisol Management Project to test the BBM in an extensive on-farm verification programme during 1986. While the main focus of this programme was the effect of improved Vertisol drainage on yields and economic returns, fertilizer treatments were overlaid on land preparation treatments to establish differentials in response to plant nutrient supply. In one location improved crop cultivars were used. In all other locations seeds of landraces were used.

The verification programme was carried out on two high elevation Vertisol plateaux at Inewari, northern Shewa, and Wereilu, southern Wello, both at 2600 m altitude, and on two lower lying locations at Debre Zeit, central Shewa, and Fogera plain, southern Gonder, both at 1850 m altitude. Inewari and Wereilu receive about 1000 mm annual rainfall, while Debre Zeit and Fogera plain receive 800 and 1200 mm, respectively. The programme involved 66 farmers and four producers' cooperatives on 81 ha (Table 2). The crops monitored were faba bean, wheat, teff (Eragrostis tef), finger millet, noug (Guizotia abyssinica) and sorghum.

The participating farmers were selected on the basis of their willingness to cooperate in the programme. Female-headed farms were also included in line with their numeric importance in the locality.

All farmers and producers' cooperatives participating in the programme were asked to split their fields into two equal parts, and to prepare one part with traditional land preparation and seeding methods and the other with broadbeds-and-furrows made with the BBM. All other inputs and the planting time were the same for the two treatments. On all plots primary land preparation was done using the conventional Ethiopian plough.

All inputs in all experimental plots were monitored continuously by enumerators residing in the respective research locations. The inputs measured were human and animal labour by origin and operation, seed and fertilizer. Yields of grain and crop residue were estimated by sampling. about 1% of the total area under treatment.

Table 2. Crops, number of plots and area covered for the on-farm drainage technology assessment.

Location

Crop

Number of plotsa

Total area (ha)

Inewari

Faba bean

4

8.3


Wheat

2

4.0

Wereilu

Faba bean

40

11.6


Wheat

34

13.5

Debre Zeit

Teff

62

14.0


Wheat

68

17.0

Fogera plain

Finger millet

18

4.5


Noug

26

6.6


Sorghum

6

1.5


Total

260

81.0

a. A plot is the part of a field which receives uniform treatment (traditional or improved drainage).

Economic evaluation method

In the absence of whole-farm monitoring, partial budgeting was used. Revenue, return to labour and return to land were computed. The estimation of these economic parameters used market prices at harvest for the value of grain and straw; local wage rate for labour value; prevailing hiring prices of oxen for oxen labour value; and official 1986 price for fertilizer. It was assumed that one BBM is shared by two small farmers, that it can cover 4 ha year-1 and that the life of the metal parts is 10 years. The total cost of the implement is Ethiopian Birr (EB) 80 (Table 1). It was further assumed that the conventional plough is used to cultivate 2 ha year-1 that the life of the wooden parts is 5 years, that the life of the tines and the metal hooks is 10 years and that the cost of a complete conventional plough is EB 32.

Results

Labour use

The number of primary cultivation passes for seedbed preparation depends on the type of crop to be sown. In general, the larger the seed, the fewer the cultivation passes required. Labour investment in primary cultivation varies considerably between locations: at Wereilu for example, 34 hours (CV 19%) were spent by an oxen pair per ha and cultivation pass, while at Debre Zeit only 26 hours (CV 17%) were used.

Table 3 shows the oxen and human labour inputs in traditional seed covering and BBF construction by the BBM. Differences between locations were considerable. Oxen time input for BBF construction with the BBM was higher than for traditional seed covering at Inewari and Wereilu, but lower at Debre Zeit, where the traditional seed covering pass with the conventional plough is carefully done, and at Forage plain. The BBM was operated in the field verification by two handlers, so that the human labour inputs shown in Table 3 are inflated. If only one handler is used, which is technically feasible, introduction of the BBM will provide a very remarkable labour saving in the Inewari area where BBF are traditionally made by hand. Oxen are used at Inewari in the seed covering exercise in order to make furrows about 120 em apart: women and children then scoop up soil from these to establish the BBF.

Regional on-farm verification results

Inewari

The two crops monitored on the Inewari plateau, wheat and faba bean, yielded 54% and 14%, respectively, more grain on the BBF made with the animal-drawn implement than on the traditionally made BBF. This is most certainly due to a greater uniformity of the drainage structure established.

Table 3. Human and oxen labour input in traditional seed covering and BBF construction with the BBM, 1986

Location

Seed covering

Human labour hr ha-1

Oxen pair labour hr ha-1 CV %

Inewari

BBF/BBM

30a

15b

25


Traditional

60

9

21

Wereilu

BBF/BBM

36a

18a

25


Traditional

14

14

19

Debre Zeit

BBF/BBM

35a

17.6a

25


Traditional

24

24

29

Fogera

BBF/BBM

30b

15b

24

plain

Traditional

35

35

87

a. Difference between BBF/BBM and traditional methods significant at P<0.001 level.
b. Difference between BBF/BBM and traditional methods significant at P<0.05 level.

Table 4 shows a summary of the physical and economic evaluation of the BBM on the Inewari plateau. Higher total costs were incurred on plots with hand-made BBF, essentially due to higher labour input. This, in conjunction with the yield impact of the BBM, resulted in considerably higher return to labour per time unit spent on the plots managed with the BBM. Local wage rate is 0.40 EB/hour. Labour investment in crop production on the plots treated with the BBM is therefore comparatively a very attractive option.

The use of fertilizer (all-ammonium-phosphate (DAP) at 100 kg ha-1) on the two crops did not result in a significant change in gross revenue in either system (Appendix). This unsatisfactory input efficiency is explained by the feet that unresponsive landraces were used.

Wereilu

Wereilu recorded very heavy rainfall in 1986 with 30 rainy days in August. This resulted in extreme waterlogging on plots under the traditional ridge-and-furrow land management system. The ridges, established with the conventional plough when covering the seed, are about 30-50 em in width. Early heavy rains tend to level out the fields, so that the initial drainage effect of the system is quickly lost.

On the other hand, the BBF established with the BBM provided sufficient surface drainage and seedbed stability to withstand the impact of the rain. Grain yield of faba bean and wheat on BBF made with BBM were of 330% and 131% higher, respectively, than yields with the traditional planting system (Table 5). Similar yield increases were recorded for straw, which is the main animal feed for much of the year.

Table 4. Economic evaluation of improved, animal-powered surface drainage at Inewari, 1986.

Land preparation and crop

Average grain yield
(kg ha-1)

Gross revenuea,c
(EB ha-1)

Total costa,c
(EB ha-1)

Return to laboura,d
(EB hr-1)

Increase in return to labour over traditional land preparation
(%)

Faba bean






BBF/BBM

810

475

224

1.0

+ 43

Traditionale

709

436

264

0.7


Wheat






BBF/BBM

618

535

234

1.2

+ 140

Traditionale

402

339

278

0.5


a. US$1 = 2.07 EB.
b. Includes value of grain and straw.
c. Operational and fixed costs.
d. After deducting all cost except for labour.
e. The traditional land preparation system is hand-made BBF.

Due to waterlogging damage to crops on traditionally managed plots, labour inputs for harvesting and threshing were lower than for BBF plots. This increased total operational cost of BBF plots (Table 5). Gross revenue from faba bean and wheat on the BBF plots was 347% and 131% higher, respectively, than on the traditionally managed plots. A lower coefficient of variation in the yield figures from the BBF plots is an indication of more stable yields.

As shown in Table 5, returns to labour from the BBF technology were much higher than from the traditional system. Results from the traditional system were very poor: return to labour was negative in the ease of faba bean and just equal to the local wage rate in the ease of wheat.

It is important to note that grain yields from the traditional system are not sufficient to meet the grain subsistence requirements of the farm families. The average cropland holding in the project area is 1.4 ha and the average family size is five: the family subsistence requirements could only be met in 1986 at the productivity levels achieved under improved drainage.

Fertilizer inputs (DAP at 100 kg ha-1) only marginally improved crop performance on both BBF and ridge-and-furrow plots, which is attributed to the unsatisfactory response to improved soil fertility of the landraces used (Appendix).

Table 5. Economic evaluation of improved, animal-powered surface drainage at Wereilu, 1986.

Land preparation and crop

Average grain yield
(kg/ha-1)

CV%

Gross revenuea,b
(EB ha-1)

Total costa,c
(EB ha-1)

Return to laboura,d
(EB hr-1)

Net returna
(EB ha-1)

Faba bean







BBF/BBM

736e

59

398e

206

0.62e

192e

Traditionalf

171

72

89

162

-0.07

-73

Wheat







BBF/BBM

689e

46

432e

214

0.61e

218e

Traditionalf

298

71

187

177

0.22

9

a. US$1 = 2.07 EB.
b. Includes value of grain and straw.
c. Operational and fixed costs.
d. After deducting all cost except for labour.
e. Difference between BBF and ridge-furrow significant at P<0.001 level.
f. The traditional land preparation system is ridges and furrows made with the conventional plough.

Debre Zeit

Rainfall during the main rainy season in 1986 in Debre Zeit was evenly distributed but 30% less than average (1977-85) in July and August. Very little run-off was therefore recorded. Despite these rather dry conditions yields and economic returns from wheat on BBF were 25% higher than on the traditional flat seedbeds (Table 6). Return to labour was 40% higher and the net return per ha was 53% higher. The cost of growing wheat did not significantly differ between the two land preparation methods.

Teff, a traditional Vertisol crop with some waterlogging tolerance, showed quite remarkably high grain and straw yields and even higher economic returns than wheat due to higher producer prices. However, grain and straw yields did not differ significantly between the two systems.

Fogera Plain

In 1986 the Fogera plain experienced extreme meteorological conditions with a very late onset of extremely heavy rains. This resulted in a serious delay and in low quality of land preparation, and in very poor condition of the working animals due to shortage of feed from natural range. As a consequence of poor seedbed preparation, resulting in heavy weed infestation, and of very heavy late rains and extreme waterlogging, the crop yields were low from both land management systems (Table 7). Sorghum completely failed on traditionally managed plots while the performance of the other crops was slightly, but not significantly, better on BBF than on traditionally managed plots.

Table 6. Economic evaluation of improved, animal-powered surface drainage at Debre Zeit, 1986.

Land preparation and crop

Average grain yield
(kg ha-1)

CV%

Average straw yield
(kg ha-1)

Gross revenuea,b
(EB ha-1)

Total costa,c
(EB ha-1)

Return to laboura,d
(EB ha-1)

Net returna
(EB ha-1)

Wheat








BBF/BBM

1541e

76

5407e

1357e

378

2.1e

979e

Traditionalf

1228

41

4261

1078

437

1.5

641

Teff








BBF/BBM

1654

32

4918

1913

537

2.1

1376

Traditionalf

1558

27

4538

1791

400

2.4

1391

a. US$1 = 2.07 EB.
b. Includes value of grain and straw.
c. Operational and fixed costs.
d. After deducting all cost except for labour.
e. Difference between BBF and flat seedbed significant at P<0.01 level.
f. The traditional land preparation system is flat seedbeds.

Table 7. Economic evaluation of improved, animal-powered surface drainage at Fogera plain, 1986.

Land preparation and crop

Average grain yield
(kg ha-1)

Average straw yield
(kg ha-1)

Gross revenuea,b
(EB ha-1)

Total costa,c
(EB ha-1)

Net returna
(EB ha-1)

Sorghum






BBF/BBM

336

1124

202

188

14

Traditionald

0

0

0

102

-102

Noug






BBF/BBM

195

1733

156

104

52

Traditionald

139

1384

111

79

32

Finger millet






BBF/BBM

522

1242

346

251

95

Traditionald

372

1051

257

223

34

a. US$1 = 2.07 EB.
b. Includes value of grain and straw.
c. Operational and fixed costs.
d. The traditional land preparation system is flat seedbeds.

Applying fertilizer (DAP, 100 kg ha-1) slightly, but not significantly, increased the grain and residue yields of noug and finger millet on both BBF and traditionally managed plots (Appendix). This lack of response was due to the use of unresponsive local crop cultivars.

Discussion and conclusions

The performance of the animal-powered drainage implement (REM) in terms of effects on crop yields and economic returns for the farm differ between locations, basically in line with the prevailing meteorological conditions.

At Inewari, where all main season crops except teff are grown on hand-made BBF, the considerable effects of the BBM on economic returns to labour are further upgraded by the feet that using this implement reduces human drudgery. The BBM is therefore likely to be quickly adopted in this area. General use of the BBM will also increase the value of the available draught cattle in the system.

At Wereilu the use of the BBM dramatically improved returns to labour. The 1986 season was considered, by farmers interviewed in a base-line survey, as a bad agricultural season with exceptionally heavy waterlogging. The BBM allowed participating farmers to overcome this constraint and achieve yield levels that were regarded by the same farmers as equal to those that would be expected in a good year. The crop yields achieved in the traditional system are below subsistence, while the ones achieved on BBF plots are considerably above this threshold. The BBM therefore contributed to providing food security to the participating farmers. This is the primary objective of introducing the technology in this area.

At Debre Zeit high yields were achieved on both land management systems in a meteorologically favourable year. However, there were indications that waterlogging decreased yields even under these low-rainfall conditions, which suggests that waterlogging constrains crop growth in most Ethiopian highland Vertisol areas in any one year. In Debre Zeit the BBM also contributed to higher returns to labour due to time saving in seed covering.

The low yields and economic returns in the Fogera plain are explained by the extreme meteorological conditions, which resulted in poor land preparation, high weed infestation and poor quality of the BBF made.

Fertilizer effects on crops were consistently disappointing in both land preparation systems. It is assumed that this is primarily due to the fact that the landraces which were monitored tend to be not very responsive to improved soil fertility.

The on-farm evaluation of the animal-drawn surface-drainage implement is being continued for another season in order to strengthen the encouraging first year information base, which will serve as a justification for a larger scale extension scheme for the technology to be undertaken by the Extension Services of the Ministry of Agriculture. Other components of the improved Vertisol management technology, such as improved cultivars with higher response potential to improved soil fertility, improved cropping systems, and water harvesting for re-use in irrigation, are being integrated into the programme in order to further strengthen the impact of improved Vertisol surface drainage.

Appendix. Fertilizer effects on crop yields and economic return on plots with traditional management and with improved drainage at Inewari Wereilu and Fogera plain 1986.

Location land preparation and crop

DAPa
(kg ha-1)

Grain yield
(kg ha-1)

CVb
(%)

Straw yield
(kg ha-1)

Gross revenuec
(EB)

Marginal rate of returnc,d
(EB)

Inewari







BBF/BBM







faba bean

0

810


1110

475


faba bean

100

907


1246

533

0.7

wheat

0

618


1953

535


wheat

100

708


2234

613

1.0

BBF traditional







faba bean

0

709


1171

436


faba bean

100

732


1534

482

0.6

wheat

0

402


1175

339


wheat

100

569


1797

493

1.9

Wereilu







BBF/BBM







faba bean

0

736

59

591

398


faba bean

100

858

64

731

468

0.86

wheat

0

689

46

877

432


wheat

100

848

35

1297

554

1.51

Ridge-furrow







faba bean

0

171

72

187

89


faba bean

100

257

125

243

143

0.67

wheat

0

298

72

379

187


wheat

100

469

67

616

296

1.35

Fogera plain







BBF/BBM







sorghum

0

336

93

1124

202


sorghum

100

258

141

685

144

- 0.7

noug

0

195

32

1733

156


noug

100

251

16

2849

201

0.6

finger millet

0

522

66

1242

346


finger millet

100

921

36

2245

614

2.3

Ridge-furrow







sorghum

0

0

-

0

0


sorghum

100

0

-

0

0

-

noug

0

139

51

1384

111


noug

100

166

35

2080

133

0.3

finger millet

0

372

89

1051

257


finger millet

100

692

29

2125

688

2.9

a. DAP = all-ammonium-phosphate: 18% N 46% P2O5.

b. A meaningful CV for grain yield in Inewari was not computed because only two plots were used for each treatment for wheat and four for faba bean.

c. US$1 = 2.07 EB.

d. Value of the grain yield effect of each additional EB of DAP.

Acknowledgements

The authors are most indebted to Girma Alemayehu, Abebe Misgina, Asfaw Yimegnuhal and Wagnew Ayalneh, Vertisol outreach sub-project leaders in Inewari, Wereilu, Debre Zeit and Fogera plain, respectively, for their dedicated work in the generation of the data reported in this paper.

The Joint Project on Improved Management of Vertisols is grateful to the governments of Switzerland, Norway and Finland and to Oxfam America and Caritas Switzerland for funding its operations.

References

Berhanu Debele. 1985. The Vertisols of Ethiopia: their properties, classification and management. Fifth Meeting of the Eastern African Sub-Committee for Soil Correlation and Land Evaluation. Wad Medani, Sudan, 5-10 December 1983. World Soil Resources Reports No. 56. FAO (Food and Agriculture Organization), Rome. pp. 31-54.

Jutzi S C, Anderson F M and Abiye Astatke. 1986. Low-cost modifications of the traditional Ethiopian tine plough for land shaping and surface drainage of heavy clay soils: preliminary results from on-farm verification trials. ILCA Bulletin 27:28-31.

Jutzi S C, Getachew Asamenew, Haque I, Abate Tedla and Abiye Astatke. 1987. Intermediate technology for increased food and feed production from deep black clay soils in the Ethiopian highlands. Paper presented at the FAO/SIDA Seminar on Increased Food Production through Low-cost Food Crops Technology, Harare, Zimbabwe, 1-17 March 1987.

Kanwar J S. Kampen J and Virmani S M. 1982. Management of Vertisols for maximizing crop production - ICRISAT experience. In: Vertisols and rice soils of the tropics. Twelfth International Congress of Soil Science, New Delhi, India, 8-16 February 1982. Indian Society of Soil Science, New Delhi, India. pp. 94-118.


Previous Page Top of Page Next Page