Leul Kahsay
Tigray Bureau of Water Resource Development (TBoWRD), Mekelle, Ethiopia
E-mail: leul_kahsay@yahoo.com
Challenges in promoting irrigation in the Tekeze Basin
Details of the research issues
Determining reservoir sedimentation rate and its mitigation measures
Mitigating sedimentation through integrated watershed-management technologies
Effect of excessive dewatering of shallow wells on productivity of grassland
Future strategy in the Tekeze Basin
Abstract
The existing perennial streams in the Tekeze Basin are not dependable sources of irrigation water due to their limited number and flow rate. Thus, harvesting of the seasonal surface runoff is the strategic option to promote irrigation in the Tekeze Basin. In the absence of locally developed design guidelines and code of practice, the use of empirical equations, developed for other parts of the world, has been very helpful. Nevertheless, there has been a case of under- and over-estimation of design parameters as a result of which the performance of the irrigation schemes was adversely affected. This paper presents background information, challenges encountered and research needs in the design of community-based irrigation schemes.
Water has been recognised as the most important factor for the transformation of the agrarian system in Ethiopia (UNDP/ECA/FAO 1994). This statement has been reinforced by the FDRE policy and strategy on rural development issued in 2002 (MoI 2002). It is indicated that reliable water supply and efficient water management are prerequisites for fast and sustainable agricultural development in Ethiopia.
The existing perennial streams in the Tekeze Basin are not dependable sources of irrigation water due to their limited number and flow rate. Thus, harvesting seasonal surface runoff is the strategic option to promote irrigation in the Tekeze Basin (UNDP/ECA/FAO 1994). In accordance with this, Commission for Sustainable Agriculture and Environmental Rehabilitation in Tigray (CoSAERT) and Commission for Sustainable Agriculture and Environmental Rehabilitation in Amhara (CoSAERAR), in Tigray and Amhara, respectively, were established in 1996 with the mandate to undertake study, design and construction of irrigation schemes. The design procedure followed by these institutions was based on standard guidelines and empirical equations commonly found in textbooks. Moreover, multidisciplinary professionals are involved in the study and design of the irrigation schemes.
In the absence of locally developed guidelines and code of practice, the use of empirical equations, developed for other parts of the world, has been very helpful. Nevertheless, there has been a case of under- and over-estimation of design parameters as a result of which the performance of the irrigation schemes was adversely affected. Though limitations in technical capacity had a significant share in design problems, issues that need research are listed below:
This paper intends to present background information, challenges encountered and research needs in the design of community-based irrigation schemes. Future strategy in the Tekeze Basin is also discussed briefly.
Tekeze River Basin has a catchment area of about 68,751 km2 within Ethiopia (MoWR/NEDECO 1996). Out of this, 35,288 km2 or 51% lies in Tigray and 27,617 km2 or 40% in Amhara (Figure 1). The lowland part of the Tekeze River Basin consists of about 1500 km2 area, which is almost flat land (WoWR/NEDECO 1996). Observation of topographic map indicates that the above area has high irrigation potential in view of the water availability from Tekeze and its tributaries.
Source: MoWR/NEDECO (1997).
Figure 1. Tekeze River Basin.
About 70% of the basin lies in the highlands at an altitude of over 1500 metres above sea level (masl). The area of land above 2000 masl covers about 40% of the total basin. The arable land in the highland part of the basin is found in the plateaus and valleys. In those areas there is a significant elevation difference between the major riverbed level and the agricultural fields. Figures 2 and 3 show two cross sections of Suluh Valley to represent the relative position of the riverbed and plain land in the upstream part of the basin. The cross sections are typical of all rivers in the highland where the riverbed is very deep compared to the nearby agricultural fields. At some potential dam sites, the riverbed is 80 m deeper than the nearby field (CoSAERT/WAPCOS 2001).
Figure 2. Section at 1/3 of the Suluh reach from U/S
Figure 3. Section at ½ of the Suluh reach from U/S
The importance of irrigation in the Tekeze Basin has been recognised many generations back according to members of the indigenous irrigation schemes. Field observations made in most parts of the basin show that almost the entire available perennial surface water source is used for irrigation, except those found in deep gorges, as of many decades back. In Tigray alone, the total area irrigated by 2002 was 4773 ha or 0.44% of the total arable land (BoANR 2003). In some unpublished reports, the total irrigated area is stated as 6500 ha. The fluctuation in size of irrigable area from one year to the other could be due to the drying of the water source following drought or shortage of rainfall.
Over the last decade, government agencies and non-governmental organisations (NGOs) have intervened to develop new irrigation schemes and improve the indigenous irrigation schemes by constructing more stable hydraulic structures. As yet, a total of 56 irrigation schemes with a potential irrigable area of 3845 ha have been designed and constructed by CoSAERT alone. But, very little has been done to improve the agronomic and water management practices. Thus, more work is still required to maximise benefit per unit volume of water.
Although not without merit, efforts focusing on improving the productivity of existing water source might not be adequate in ensuring food security in the Tekeze Basin. Irrigation could play a significant role in enhancing agricultural production in the Tekeze Basin only if water harvesting is considered as an integral part of the exercise.
The average yield of wheat and maize in the 'high rainfall' areas is in the order of 20–25 and 50–55 qt/ha, respectively. In the 'low rainfall' areas, the average wheat and maize yield is 7–11 and 8–11 qt/ha, respectively (BoANR 1992–1998). This information could give a clue on the effect of moisture deficiency in the rain-fed agriculture.
The mean annual rainfall in the Tekeze Basin ranges from 300 in the north-eastern end to over 1200 mm in the south-western (Figure 4). The cumulative amount of rainfall could have been sufficient for crop production even in the 'low' rainfall area. The problem is that there is large spatial and temporal rainfall variability to the extent of causing low agricultural production.
Figure 4. Mean annual rainfall in Tigray, mm.
The variability of annual rainfall ranges from 20% in the highlands of Western Tigray to 40% in Eastern Tigray (UNDP/ECA/FAO 1994). As shown in Figure 5, 70–96% of the rainfall during the growing season occurs in July and August. This fact explains that rain-fed agriculture suffers from moisture deficiency especially during the last two months (September and October) of the growing season. Therefore, supplementary irrigation would be of paramount importance to obtain at least one reliable harvest. The extent of the moisture deficiency during the growing season is depicted in Figure 6, taking the 75% dependable rainfall and maize into consideration. For Quiha station, for instance, the moisture deficiency in September and October is 146 and 81 mm, respectively.
Source: BoPED (1998).
Figure 5. Seasonal rainfall and proportion falling in July and August.
Figure 6. Moisture deficiency at three selected stations in the Tekeze Basin.
This paper is prepared in light of the efforts made to promote irrigated agriculture in Tigray during 1996–2003 as 51% of the basin falls in Tigray. The facts indicated in the paper are based on field observation as well as review of study and design document.
Tekeze River is perennial and near an international border with a minimum flow of 2 m3/s (MoWR/NEDECO 1996). In the highland part of the basin only large rivers are perennial with some flows. The minimum dry season discharge of those streams ranges from 0.5 to over 70 l/s (CoSAERT 2002) (Figures 7 and 8). Most of these streams and some springs are already used for irrigation. A reconnaissance survey conducted in 2000 shows a total of 3492 ha has been irrigated by 466 indigenous schemes. With some investment in the 466 schemes, it was indicated that the irrigable area could be expanded to 5272 ha. However, as per an unpublished document from the TBoWRD, the above-mentioned size of potential irrigable area is found to be highly exaggerated following a ground verification exercise. In most of the schemes, deficit irrigation is practiced due to the limitations in the stream flow rate and seepage along the canal..
Figure 7. Suluh River monthly discharge in 2002 compared to average previous years.
Source: MoWR (1999).
Figure 8. llalla River monthly discharge in 2002 compared to average of previous years.
There is no detailed hydro-geological study made in the Tekeze Basin. An important document that would give a clue about the ground water potential in the region is a geological map prepared in 1996 by the Ethiopian Institute of Geological Survey.
About 25% of Tigray is covered with sedimentary rock, which is believed to be relatively better water-bearing formation. However, the ground water recharge rate in the indicated area is adversely affected by low and/or intensive rainfall condition (300–550 mm), ruggedness of the topography and absence of surface vegetative cover.
Igneous rock covers about 54% of Tigray. This rock could be a source of ground water if it is weathered or have fractures in it. However, the igneous rock covering the area west of the Tekeze River, which is the major part, has little or no fractures and weathering.
The third major rock category is basement and it covers about 21% of Tigray. This type of rock has no potential ground water source.
Thus, about 75% of the region, covered by igneous and basement, has little or no ground water potential.
In areas where ground water is mined for domestic water supply, a continuous decline of the water table is being observed. According to unpublished document from the TBoWRD, a total of 350 (16.7%) hand pumps, 6 (4%) deep wells and 108 (21%) springs and very high number of shallow hand dug wells have been reported to be dry in connection with the 2002 drought. Monitoring of the Aynalem well field, (which is the source of water for Mekelle town) shows that the water table has been declining at a rate of 2–2.4 m per year during 1998–2003 (Figure 9). Therefore, the possibility of using ground water for irrigation is very remote.
Figure 9. Trend of falling ground water level at Aynalem well field.
In some valley plains, composed of either heavy clay soil or loose sandy/gravely soil, water table is found at a depth of 1–9 metre. Such source of water is being used for domestic water supply and irrigation.
For instance, by 2003 the number of hand dug shallow wells and the cumulative irrigated area in Tahitay Koraro woreda was 700 and 20 ha, respectively. The water lifting mechanism being used includes buckets, shadoof, and treadle pump. In some cases where motorised pumps were introduced, the yield of the wells decreases mainly after January or about 5 month of the end of the rain season.
As of January 2003, small ponds with a capacity of 150–185 m3 are being excavated by mobilising people. Each pond is expected to provide water for a supplementary irrigation of a plot size of 500–1000 m2. The positive part of these ponds is that they are owned by a household, which makes the operation and maintenance simple and smooth. The major challenge in this case is the possibility of water loss due to seepage. The other point is that due to the smallness of the plot size, the increment in production might not satisfy the annual food requirement of a family. Assuming an average wheat yield of 20–30 qt/ha, the total production from a 0.1 ha plot would be 2–3 qt. A family of 5 may need over 10 qt of food per year. Therefore, either the pond size has to be increased or the family has to look for other source of income.
In the period between 1996 and 2001, 46 reservoir dams with a cumulative storage capacity of 49.91 million m3 and irrigable area of 3115 ha have been constructed (CoSAERT 2001).
However, the performance of the reservoirs was diminished due to seepage and failure to harvest the designed runoff. As shown in Figure 10, the proportion of runoff harvested by some of the reservoirs was 10–96% of the design capacity in 2000 through 2002. The pronounced harvest failure in 2002 is attributed mainly to drought. .
Source: CoSAERT (2001).
Figure 10. Percentage of harvested water by reservoirs as compared to their design capacity.
An unpublished study document made in 1998 by CoSAERT shows that, even under normal rainfall conditions, the failure to harvest the designed runoff was attributed, among other things, to the following:
According to a survey made by CoSAERT in six irrigation schemes in 1999 and nine schemes in 2001, the dominant irrigated crop is maize. As shown in Table 1, the proportion of area covered with maize had increased from 49.1 to 89.5%. With random interviews, the farmers' reason for growing more of maize is that firstly it can be consumed directly and the stalk can be source of feed and fuel. Secondly, it is less laborious and not perishable. Thirdly, the seed and pesticides required for growing horticulture are expensive and not easily available.
Thus, the irrigation schemes are not yet developed up to their capacity where benefit is the highest per unit of water. Under such conditions, it would be difficult for the irrigators to cover the operation and maintenance costs let alone to recover the investment fully or partially.
| Table 1. Irrigation cropping pattern as collected in 1999 and 2001. | ||
Crop type |
Year data collected | |
1999 average of six dams (%) |
2001 average of nine dams (%) | |
Maize |
49.2 |
89.5 |
| Source: COSAERT (2001). | ||
No data has been collected in the area of water use efficiency. However, signs of excessive seepage are observed along canal routes of many irrigation schemes. In some indigenous irrigation schemes, where water is delivered on rotation, moisture deficiency is noticeable. That means no attempt was made to match the stream size and the size of the irrigable area.
In the process of designing water harvesting structures, the value of the runoff coefficient C, which is developed as a function of catchment characteristics, is obtained from textbooks. Although not without merit, the attempt to directly adopt design parameters developed for other geographical location involves serious errors.
An overestimation of the runoff coefficient results in expending unnecessary cost. Similarly, underestimation of the coefficient results in reduced irrigable size (Figure 11).
Figure 11. Area under supplementary irrigation with runoff harvested from 400 mm rain.
Table 2 shows reservoir dams whose dead storage level has been filled with sediment deposition earlier than anticipated.
| Table 2. Actual and designed sedimentation rate in some reservoir dams. | ||||||||
Name of dam |
Location |
Year |
Catchments area (km2) |
Total capacity |
Dead storage volume |
Designed sedimentation rate |
Actual sedimentation rate | |
Constructed |
When dead storage filled up |
|||||||
Gereb Mihize Adi Kenafiz Mai Gassa Gra Shetu |
13º14'16'' |
1998
1998
1997 1998 |
2002
2001
2001 2001 |
17.2
13
9 2.88 |
1.2
0.67
1.3 1.42 |
0.325
0.067
0.153 0.028 |
950
1200
970 1000 |
4724
1719
4270 3291 |
| Source: TBoWRD (unpublished). | ||||||||
The design sedimentation rate indicated in Table 2 is obtained from literature, which was based on a sediment sampling conducted on streams around Mekelle in 1974–1976 (CoSAERT 1997/1998). The limitation in such approach is that, firstly, the data need to be updated in view of the changes in land use. Secondly, the data could not be taken as representative to all areas of interest for there is substantial temporal and spatial variability in factors causing sedimentation.
An underestimated sedimentation rate gives wrong conclusion with regard to the feasibility of the reservoir. In addition, it could mislead the pace of the implementation of soil and water conservation on the catchment area.
It has been long said that the introduction of integrated watershed management practices in the catchment areas of reservoir dams would retard rate of sedimentation. Often, the proposed Soil and Water Conservation (SWC) measures are physical structures, biological measures and closure of a portion of the catchment from human and livestock interference. These activities are implemented with the objective of improving the livelihood of the upland and down stream people.
In some instances, the SWC effort has yielded some benefit in rehabilitating the physical environment. This can be explained in terms of re-growth of indigenous shrubs and grass, reappearance of springs after being dry for long time, stabilisation of gullies and the like.
In most cases, however, the technologies introduced as part of the SWC measure were not observed in bringing a significant economic benefit to the people whose livelihood is connected directly with the catchment areas of reservoir dams.
The expected yield from biological SWC measures, as outlined in the Awash Basin Master Plan, is as follows:
a. Fuel wood
b. Fodder production
The problem with the expected benefit from the existing biological SWC measures, as outlined above, is that 1) the farmers have to wait for a long period before the first harvest and 2) the yield is too low.
Under such conditions, the upland farmers would consider the issue of SWC as the interest of outsiders who think that the resultant benefit would be obtained over the long term. Therefore, it is high time to revise the prevailing watershed management technologies such that the community would get direct benefit in the short term as well. In light of the above-mentioned problem, the following research questions are posed for consideration.
As shown in Figure 6, rain-fed agriculture in most parts of the Tekeze Basin is subjected to moisture deficiency. In addition to the dry spells in July and August, the amount of rainfall in September and October is insignificant under normal conditions. The effect of such moisture deficiency on crop yield has to be researched.
A preliminary exercise using the Doorenbos and Kassam (1979) model shows that, keeping other agricultural inputs constant, there would be a 41% yield reduction due to moisture deficiency towards the eastern end of the basin. This result more or less coincides with the actual crop yield of rain-fed agriculture.
In light of the above facts, supplementary irrigation is being considered as the key input to obtain one reliable harvest in the basin. Water harvesting is the major means to undertake supplementary irrigation as indicated above. Therefore, many questions need to be answered in connection with water harvesting structures to be used for supplementary irrigation:
The last question can be assessed in terms of the following:
According to unpublished information from the DPPB of Tigray, a large number of population depends on food aid. The number of food insecure people in a given year positively correlates with the rainfall shortages. Stemming from this, every household should be encouraged and assisted to practice supplementary irrigation using a water harvesting structure.
Therefore, it would be of utmost importance to identify the minimum size of a water-harvesting pond that would enable a family produce at least the annual food requirement from a certain plot.
A family of 5 may need about 10 qt food grains per year. Assuming a wheat yield of 30 qt/ha and with manipulation of climatic information, a plot of 0.4 ha can produce 10.1 qt of wheat grain if supported with 4 irrigations beginning September 1st. A 0.5 ha plot may produce 10.4 qt with two supplementary irrigations (Table 3). This has to be verified through detailed fieldwork, the objective being to identify the minimum size of a water-harvesting pond that would enable to produce enough food for a family. Concurrently, appropriate agronomic and water management practices should have to be designed and introduced to the farmer.
Table 3. Expected yield increment attributed to supplementary irrigation using ponds at Adigdom
Source: BoWRD (unpublished data).
As explained above, irrigation water is a critical variable in obtaining at least one reliable harvest in the Tekeze Basin. Only in rare cases, the topography lends itself conveniently for the construction of a reservoir. In such cases, substantial amount of water can be stored per unit area of land or per unit volume of earth fill. However, the ratio of irrigable area to inundated area in most of the reservoir dams, built by CoSAERT, is in the range of 2–3.5. This indicates that it would be difficult to construct a water harvesting structure unless a sizeable area is sacrificed.
Thus, a family or a community has to spare a portion of its land for a reservoir to impound water. If the land to be spared is an agricultural land, then the benefit attributed to the intervention has to overweigh. The analysis to be made in this aspect should include the social and political dimension of the problem by asking the following question: What will happen if food deficiency persists and food aid is stopped?
Under the culture of supplementary irrigation, harvested water is used within two months from the end of the rainy season. Whereas, with double cropping, the water has to be used in about eight months from the end of the rainy season. Figure 12 shows that for a given volume of water harvested from a certain catchment area, more land can be irrigated under supplementary irrigation than double cropping. Such increment in irrigated land is mainly due to the possibility of using the water that could have been lost as evaporation and seepage during November through April. As the catchment area, from which runoff is generated, increases the water use efficiency increases with supplementary irrigation (Figure 13). With small ponds the proportion of evaporation loss is very high. Thus, water productivity would be increased if the harvested water is used only for supplementary irrigation..
Figure 12. Irrigable area under double cropping and supplementary irrigation using water harvested from various catchments.
Figure 13. Proportion of utilized and lost water under double cropping and supplementary irrigation
Most of the shallow wells dug, for domestic water supply and irrigation, are along the periphery of grasslands where thousands of livestock depend for their grazing. As yet, the grass in the indicated area is perennial mainly due to the capillary fringe from the shallow water table. However, this natural irrigation process might be disrupted by the unplanned digging of the shallow wells. Thus, unless a mechanism is introduced to balance the mining and the rate of ground water recharge, the productivity and survival of thousands of livestock is in jeopardy.
Examination of 11 irrigation projects designed by CoSAERT reveals that the average elevation difference between the reservoir outlet and the end of the supply canal is 1.15 m per 100 m. The slope of the field, perpendicular to the supply canal, is often more than 1%. In addition to the indicated elevation difference, the head above the outlet pipe, often up to 10 m, could be sufficient to drive the flow of water through pipes.
Thus, gravity being the source of energy to drive the flow, there could be a good demand for pipes and drip laterals if the price is attractive to the small-scale subsistence farmer.
Use of short furrow length has significant contribution in minimising the difference in infiltrated volume between the head and tail of the furrow. One disadvantage of short furrow length is that labour requirement is higher compared to longer furrows. However, this is not the case in the Tekeze Basin as in almost all of the irrigation schemes the furrow length being considered is in the range of 5–15 m. Farmers prefer short furrow length because it allows uniform water distribution in the field. On the other hand, short furrow length requires longer distribution system and hence more conveyance losses.
Therefore, the merit and demerit of the irrigation practice with short furrow length has to be examined.
The total annual volume of water draining from the Tekeze Basin is 7.36 million m3 with peak flow rate in July and August (MoWR 2001). However, due to topographic limitations it would be difficult to use a significant portion of the indicated volume of water except in the low land bordering the Sudan.
Likewise, the tributaries of Tekeze River have their peak flow rate in July and August and minimal flow rate during the remaining period of the year. For instance, under average conditions, Suluh River is a perennial river with discharge varying from 0.26 to 4.4 m3/sec in January and August, respectively. In drought years, like that of 2002, the flow rate of Suluh River becomes nil as of the fourth month from the end of the rain season (Figure 7).
On the other hand, observation of the cross section of the catchments of the tributaries of Tekeze River shows that the streambed is 20–100 m below the potential arable land. That is, the water is inaccessible unless expensive water lifting mechanism is introduced. Thus, the possible source of irrigation water for the arable land located in the highland part of the Tekeze Basin is seasonal runoff water to be harvested from the small catchments.
Therefore, with the objective of benefiting more people it is becoming necessary to maximise the productivity of the harvested water.
Ghinassi et al. (2002) reported that seasonal water supplies ranging from 20 to 60% full irrigation yielded from 70 to about 90% marketable yield of tomato.
The Tekeze Basin Master Plan specifies a large-scale irrigation in the lowland part of the basin. Such irrigation scheme would get its water from reservoir dams to be built for hydropower generation.
Beside this, there is a possibility of constructing medium-scale reservoir-based irrigations schemes in the highland (CoSAERT 1996). Small-scale reservoirs and ponds which could be constructed by mobilising people shall be given due consideration for their big role in ensuring food security. Attempts have been made at Suluh Valley to conduct an integrated water resource development study taking the whole 960 km2 wide valley as a unit. The aim was to study the possibility of using both ground water and surface water for irrigation.
Irrigation development in the Tekeze Basin is possible mainly through water harvesting which in turn requires the design and construction of various hydraulic structures. The design is expected to yield stable and less expensive structure that would enable to obtain maximum benefit per unit volume of water. To this effect, design work needs to be based on locally tested procedures wherever applicable.
BoANR (Bureau of Agriculture and Natural Resources). 2001. Five year agronomic plan for 2001–2005. BoANR, Mekelle, Ethiopia. (Unpublished).
BoANR (Bureau of Agriculture and Natural Resources). 1999. Report on the 1998 production season. BoANR, Mekelle, Ethiopia. (Unpublished)
BoANR (Bureau of Agriculture and Natural Resources). 2002. Evaluation of 2002 production season. BoANR, Mekelle, Ethiopia.
BoANR (Bureau of Agriculture and Natural Resources). 2003. Strategic plan of BoANR for 2003–2006. BoANR, Mekelle, Ethiopia. (Unpublished)
BoPED (Bureau of Planning and Economic Development). 1998. Atlas of Tigray. BoPED Physical Planning Department. BoPED, Mekelle, Ethiopia.
CoSAERT (Commission for Sustainable Agriculture and Environmental Rehabilitation in Tigray). 1996. Feasibility study report of Maigassa Reservoir Dam. CoSAERT, BoANR, Mekelle, Ethiopia. (Unpublished)
CoSAERT (Commission for Sustainable Agriculture and Environmental Rehabilitation in Tigray). 1997. Feasibility study report of Grashitu Reservoir Dam. CoSAERT, BoANR, Mekelle, Ethiopia. (Unpublished)
CoSAERT (Commission for Sustainable Agriculture and Environmental Rehabilitation in Tigray). 1999. Cost recovery on small-scale irritation project in Tigray. Concert Engineering. CoSAERT, BoANR, Mekelle, Ethiopia.
CoSAERT (Commission for Sustainable Agriculture and Environmental Rehabilitation in Tigray). 1997. Feasibility study report of Adikenafiz Reservoir Dam. CoSAERT, BoANR, Mekelle, Ethiopia. (Unpublished)
CoSAERT (Commission for Sustainable Agriculture and Environmental Rehabilitation in Tigray). 1997. Feasibility study report of Gereb Mihiz Reservoir Dam. CoSAERT, BoANR, Mekelle, Ethiopia. (Unpublished)
CoSAERT (Commission for Sustainable Agriculture and Environmental Rehabilitation in Tigray). 1998. Case study on causes of harvest failure of dams bordering the eastern escarpment. CoSAERT, BoANR, Mekelle, Ethiopia. (Unpublished)
CoSAERT (Commission for Sustainable Agriculture and Environmental Rehabilitation in Tigray). 2001. Preliminary assessment on the overall impact of dams. CoSAERT, BoANR, Mekelle, Ethiopia.
CoSAERT/WAPCOS (Commission for Sustainable Agriculture and Environmental Rehabilitation in Tigray/ ). 2001. Suluh Valley integrated rural, agriculture and water resources development study, identification and reconnaissance report. CoSAERT, BoANR, Mekelle, Ethiopia. (Unpublished)
Ghinassi G., Giacomin A. and Neri L. 2002. Irrigation under water scarcity: Field experiences for a methodological approach. Proceedings of 18th congress on irrigation and drainage held in Montreal, Canada, 21–28 July 2002. ICID (International Commission on Irrigation and Drainage), Montreal, Canada.
Leul K. 1997. Practices for maximising benefit per unit volume of harvested water. CoSAERT (Commission for Sustainable Agriculture and Environmental Rehabilitation in Tigray), Mekelle, Ethiopia.
Leul K. 1998. Financial viability of using imported low pressure gated pipe irrigation system in Tigray. CoSAERT (Commission for Sustainable Agriculture and Environmental Rehabilitation in Tigray), Mekelle, Ethiopia.
MoI (Ministry of Information). 2002. FDRE Government rural development policies and strategies. MoI Press and Audiovisual Department, Addis Ababa, Ethiopia.
Halcrow H. 1989. Master plan for the development of surface water resources in the Awash Basin. MoWR (Ministry of Water Resources), Addis Ababa, Ethiopia.
MoWR/NEDECO (Ministry of Water Resources/). 1997. Tekeze Master Plan. MoWR, Addis Ababa, Ethiopia.
MoWR (Ministry of Water Resources). 2001. Water resource development and management in Ethiopia. Proceedings of a panel discussion. MoWR, Addis Ababa, Ethiopia
REST (Relief Society of Tigray). 2000. Study on traditional irrigation system in Tigray. Concert Engineering. REST, Mekelle, Ethiopia.
Rongfang L., Dagen X. and Junzu X. 1996. Final report on medium and large scale water resources and irrigation development in Giba and Rama basins of Tigray. COSAERT (Commission for Sustainable Agriculture and Environmental Rehabilitation in Tigray), Mekelle, Ethiopia. (Unpublished)
UNDP/ECA/FAO (United Nations Development Programme/Economic Commission of Africa/Food and Agriculture Organization of the United Nations). 1994. Sustainable agriculture and environmental rehabilitation in Tigray. SAERT (Sustainable Agriculture and Environmental Rehabilitation in Tigray) working document. ECA, Addis Ababa, Ethiopia.
Yigzaw H. 1998. Case on causes of harvest failure of dams bordering the eastern escarpment. (Unpublished)
