AMBO UNIVERSITY SCHOOL OF GRADUATE STUDIES, DEPARTMENT OF BIOLOGY, ENVIROMENTAL SCIENCE PROGRAM CHARACTERISTICS AND ONSITE COSTS OF THE SEDIMENT LOST BY RUNOFF FROM DAPO AND CHEKORSA WATERSHEDS, DIGGA DISTRICT By: Alemayehu Wudneh Advisor: Teklu Erkosa (PHD) Co-advisor: Prhaba Devi (PHD) A Thesis is submitted to the School of Graduate Studies of Ambo University in partial Fulfilment of the Requirements for the Degree of Master of Science, in Environmental Science October, 2012 Ambo, Ethiopia APPROVAL SHEET Submitted by: ______________________ _______________ ___________ PG Candidate Signature Date Approved by 1. Advisor _________________________ _____________ ______________ Name Signature Date 2. Co-advisor __________________________ _______________ ________________ Name Signature Date 3. Co-advisor __________________________ _______________ ________________ Name Signature Date 4. College/Institute Dean __________________________ _______________ ________________ Name Signature Date 5. Director, School of Graduate studies _____________________________ _________________ ______________ Name Signature Date ii Statement of the author I declare that this thesis is my work and all sources of materials used for this thesis have been duly acknowledged. This thesis submitted in partial fulfillment of the requirements for MSc degree in environmental science to Ambo University. I solemnly declare that this thesis is not submitted to any other institution anywhere for the award of any academic degree, diploma or certificate. In all other instances requests for permission for reproduction of this thesis in whole or in part, may be grant obtained from the permission of author and IWMI. Name…………………………………………………………. Signature……………… Ambo University, Ambo. Date of submission……………………… i Acknowledgements First of all, I would like to thank almighty God for helping me to start and successfully complete this work. I convey my deepest thanks to my major advisor Dr. Teklu Erkosa, irrigation engineer, IWMI, Addis Ababa, for giving me constructive advice and guidance in preparing the proposal, research guidance and finalizing the thesis. Without his encouragement, suggestion and support the completion of this research work would not have been possible. I am also thankful to my Co-advisor Dr. Prhaba Devi, Ambo University for her comments, suggestion and guidance from the very beginning of my research wok. The authors gratefully acknowledge the International Water Management Institute (IWMI) for funding this study, particularly; I would like to thank Dr. Charlotte MacAlister, Project Leader, IWMI East Africa and Nile Basin, from the deep of my heart for her outstanding join supervision and constructive comments on a number of seminars prepared by IWMI, Addis Ababa. Dr. Brihanu Zemedin, IWMI for his help and advice in executing runoff discharge estimation. The help rendered by all the laboratory technicians of Nekemte Soil Research Institute are highly appreciated for their collaboration. ii Table of contents Acknowledgements. ......................................................................................................................... i Lists of Tables ................................................................................................................................ vi Lists of Figures ............................................................................................................................. vii Lists of Annexes .......................................................................................................................... viii Acronyms ....................................................................................................................................... ix Abstract…………………………………………………………………………………………...ix 1. Introduction ................................................................................................................................ 1 1. 1. Background and Justification ........................................................................................... 1 1.2. Statement of the problem ................................................................................................. 2 1.3. Significance of the study .................................................................................................. 3 1.4. Objectives of the study ..................................................................................................... 3 1.5. Hypothesis ....................................................................................................................... 4 1.6. Scope of the study ............................................................................................................ 4 2. Literature Review ....................................................................................................................... 5 2. 1. Concept of Soil erosion .................................................................................................... 5 2.2. General overview of soil loss extent in Ethiopia………………………………………...6 2.3. Suspended sediment...……………………………………...………...……………..…...7 2.4. Nutrient depletion . ………………………………………………………………………8 2.5. Transportation of nutrients to water body ....................................................................... .9 2.6. On site impact of soil erosion…………………………………………………………..10 2.6.1 Economic impacts of soil erosion………………………………………………......11 2.7. Valuing soil nutrient loss………………………………………………………..….......13 2.8. Sediment fingerprinting…………………………………………………...………..…..15 iii 3.Materials and Methods…………………………………………………………………….....17 3.1.Description of the study area……………………………………………………………....17 3.1.1. Diga area…………………………………………………………………………..17 3.1.2. Characteristics of Dapo and Chekorsa watershed …………………………………..18 3.2. Methods………………………………………………………………………………....20 3.2.1. Data gathering…………………………………………………………………….20 3.2.2. Selection of runoff sampling………………………………………………………..20 3.2.3.The study period…………………………………………………….....................20 3.2.4. Runoff sampling………………………………………………………………….20 3.2.5.Water sampling and storage…...................................................................................22 3.2.6. Estimation of sediment load…….............................................................................22 3.2.7. Chemical analysis.....................................................................................................23 3.2.8. Suspended sediment fingerprinting……………………………………………...24 3.2.9. Effects of nutrient losses………………………………………………………...26 3.3. Data analysis………………………………………………………………………….....27 4. Results and Discussion……………………………………………………………………...28 4.1. Discharge …………………..……………………………………………………………28 4.1.1. Suspended sediment and its interaction with discharge…………………………...28 4.2. Temporal variability of suspended sediment with discharge…………………………....31 4.3. Plant nutrient loss from the watersheds by runoff…….……………………………...34 4.3.1. Plant nutrient enrichment ratio………………………..…………………..……......34 4.3.2. The severity of nutrient loss.........................................................................................36 4.4.Sediment fingerprinting……….......................................................................................37 4.4.1.Decade to decade variation of sediment source types……......................................41 4.5. Costs of nutrient loss…………………………………………………………………...42 5. Conclusion and Recommendations……………………………………………………….....46 6. References . …………………………………………………………………………………...49 iv 7. Annexure……………………………………………………………………………………...56 v Lists of Tables Table 1: Classes of nutrient loss rates (Kg/ha/year)………………………………………………8 Table 2: Global estimated impact of soil erosion on crop production .......................................... 11 Table 3: Methods and procedures used for the chemical analysis of sediment and water ...........23 Table 4 : Comparison of Dapo and Chekorsa River with Statistical analysis ............................ 31 Table 5: Comparison of mean nutrient surface soil and sediment, and enrichment ratio ............ 35 Table 6: FAO (1999) severity classes of the lost available nutrients ........................................... 37 Table 7: The results of Kruskal–Wallis H test ........................................................................ …..39 Table 8: The result of step wise DFA ……………………………………………...…………....39 Table 9: Amount of available nutrient of N and P loss in each decade from DR and CR.............43 Table 10: Estimated monitory value of available nutrient loss by the two Rivers ....................... 45 vi Lists of Figures Figure 1: On site effects of soil erosion ........................................................................................ 10 Figure 2: A conceptual model of suspended sediment finger printing…………………………...16 Figure 3: Map of Diga District...................................................................................................... 17 Figure 4: (a) Dapo watershed outlet (b) Land use and land cover near DR outlet…..…………..18 Figure 5: Land use land cover of Dapo watershed ............................................................ ...........19 Figure 6: Cross-sectional areas of Dapo River shown schematically…………………..………..21 Figure 7: Current Meter model used for measuring river flow and staff gauge (E) at DR monitoring station.…………………………………………………………………………….....22 Figure 8: Stations/Sites where the surface soil samples were taken in the DRC .......................... 24 Figure 9: Fingerprinting procedure of Walling and Collins (2002) ………………………….....25 Figure 10: Schematic illustration of overall methodology followed….……………………..….27 Figure 11: Discharge rating curve for Dapo and Chekorsa River ................................................ 28 Figure 12: Trends of total suspended sediment change with discharge of DR ............................ 29 Figure 13: Changes of SSC Vs decade average Q of Dapo (a) and Chekorsa Rivers (b)………..29 Figure 14: Relationship of SSC with Q (a) and Change in SSC with decade (b) of DR .............. 32 Figure 15: Photos showing accumulated sediment at the edge of cultivated field with teff (Eragrostis tef) and vegetation along the Dapo Stream serving as riparian zone.……………..33 Figure 16: Change in SSC with decade (B) and relationship of SSC with Q (A) of CR .............. 34 Figure 17: TKN before (a) and after pooling (b) the upper two transect line respectively……...38 Figure 18: Before (a) and after pooled (b) C: N ratio of surface soil and the sediment…………38 Figure 19: Sediment contribution from the two sub areas for each subsequent decade…………42 Figure 20: Grain yield response curve of maize-nitrogen rates around BRC…………………..44 Figure 21: Grain yield response curve of maize to phosphorus rate around BARC.................44 vii Lists of Annexure Annex 1: Crop type, planting and harvesting date in the upper and lower stream………………56 Annex 2: Natural forest coverage in the upper part of DRC ........................................................ 56 Annex 3: Cultivation of deep slope on the upper stream and corallo practice as manuring......... 57 Annex 4: Depth and cross sectional area of Dapo River ............................................................ 58 Annex 5: Dapo hydrological data of average velocity and discharge .......................................... 59 Annex 6: Chekorsa’s collected hydrological data ...................................................................... 60 Annex 7: SSL and Q for dapo and Chekorsa Rivers .................................................................... 61 Annex 8: Readymade Water and sediment sample for chemical analysis.................................... 61 Annex 9: Total essential nutrient loss with sediment and discharge ............................................ 62 Annex10: Tracers properties and the sediment sediment for CR (A) and DR (B).........................63 Annex 11: R2 regration analysis of nutrient with texture for CR (A) and DR (B)...……………..64 Annex 12: Relationship between soil loss and yields. Source (FAO, 1999)…………………….64 viii Acronyms AVP: Available phosphorus BARC: Bako Agricultural Research Center C/N: Carbon/Nitrogen CR: Chekorsa River CRC: Checkorsa River Catchment d; decade DDS: Dapo down Stream DFA: Discriminate Function Analysis DR: Dapo River DRC: Dapo River Catchment DUS: Dapo Upperstream FAO Food and Agricultural Organization GDP: Gross Domestic Product gm/L: gm per liter INM: Integrated soil nutrient management IWMI: International Water Management Institute kg/ ha/yr: Kg per hectare per year L: liter LD: Lelisa Dimtu m.a.s.l: Meter above sea level Mg: Mega gram Mha: Million hectare Mm3/d: Mega metric cube NBDC: Nile Basin Development Challenge ix NH4-N: Ammonium nitrate nitrogen NO3-N: Nitrate nitrogen NPK: Nitrogen, phosphorus and potassium OC: Organic carbon PO4-3: Phosphate ion Ppm: Parts per million Q: discharge SPSS: Statistical Package for Social Science SSC: Suspended Sediment Concentration SSL: Suspended Sediment Load SWC: Soil and water conservation t/ha/d: tonnes per hectare per decade TKN: Total Kjeldal Nitrogen tons /ha /yr: tons/hectare/year USLE /RUSLE: Revised/Universal Soil Loss Equation WAO: Woreda Agricultural Office x Abstract This study conducted in two sub catchments of the Abay Basin identified the quantity and quality of sediment loss and its origin though most studies conducted in Ethiopia focus on quantification of soil loss. Also, the onsite economic cost in terms of yield reduction was estimated taking maize (Zea mays) as representative crop. For this purpose, two monitoring stations were selected at the outlet of the two watersheds. Depth integrated runoff samples were collected during the rainy season in 2011 while discharge of the Rivers was estimated from staff gauge-discharge relationship. Daily runoff samples were bulked for ten consecutive days and filtered to separate the sediment from the water. The water and sediment were subsampled for oven dry to determine sediment concentration and for chemical analysis to determine the Nitrogen and Phosphorus content at Ambo University laboratory. The difference in sediment concentration between the two Rivers was statistically significant. Regression analysis between that suspended sediment concentration is related to discharge for Dapo River (R2=0.7) but this relation was very weak for Chekorsa River (R2=0.286). The concentration of the plant nutrients considered was greater in the sediment delivered to the outlet than that of the original surface soil. The concentration of available P in the sediment was 2.7 to 9 times its concentration in surface soil from Dapo river catchment and Chekorsa river catchment, respectively. The soil nutrients in the sediment and surface soil of the lower and upper catchment were used to identify sediment source areas using a quantitative composite sediment fingerprinting method with 87% of source type correctly classified. The contribution of the upper stream part to the sediment load of River Dapo was greater than its downstream part, with values ranging from 37% to 67% using Total Kjeldal Nitrogen, and 44% and 56% using organic carbon to nitrogen ratio but in average 56% to 63%. Mean lost of available nitrogen and phosphorus was 1.6+0.14 and 0.4+0.06, and 1.5+0.17 and1.1+0.13 in Kg per decade from Chekorsa and Dapo River, respectively. As a result, the estimated onsite cost to farmers due to the total loss of nitrogen and phosphorus throughout the study period was about 3321 and 4975 Birr ha-1 for Dapo, and 3545 and 2324 Birr ha-1 for Chekorsa catchments in that order. The study therefore helps to understand the processes and cause of nutrient loss at a micro watershed level and to implement targeted management interventions. Keywords: Soil loss, sediment fingerprinting, soil nutrient depletion, Blue Nile Basin xi 1. Introduction 1. 1. Background and Justification Environmental problems have become major global concerns. Soil erosion by water is among the severe environmental and agricultural production problem across the world. Erosion causes significant loss of soil fertility and productivity (Mequanint Tenaw and Seleshi Bekele, 2009). Soil erosion is among the common threats to agricultural production in Ethiopia (Lakew Desta, 2000, Sileshi Bekele and Holden, 1998). In the Ethiopian highlands, soil loss due to water erosion is about 1493 Mt-1yr (Hurni, 1993). Nearly half of this is estimated to come from cultivated fields, which account for only about 13% of the country’s total area. These losses will inevitably cause decrease in yield unless appropriate measures are taken. In the Abay basin (the Ethiopian part of the Blue Nile Basin) soil erosion by water is a major cause of soil fertility and productivity loss (Mequanint Tenaw and Seleshi Bekele, 2009). Understanding soil loss and its process is crucial in order to select and implement integrated nutrient management options to attain sustainable agricultural production. The provision of reliable information on the provenance of suspended sediment transported by rivers is important from a number of perspectives such as to establish catchment sediment budget, validation based on physically distributed soil erosion and sediment yield model. The targeting of sediment management strategies is a key requirement in developing countries because of the limited resources available (Collins and Walling, 2002). However, most studies conducted in Africa including Ethiopia focus on quantification of soil loss using the University soil Loss Equation (USLE) or its revised version (RUSLE), erosion pins, runoff plots or remote sensing technologies. The efforts have given little attention to the original provenance or sources of sediment. Estimation of sediment yield has important economic consequences (Gruhn et al., 2000). Most current evaluations of the costs of land degradation have focused on the loss of soil from farm plots and the loss of nutrients resulting in decreased productivity or the need for increased inputs 1 to maintain productivity (Berry et al., 2003). However, the cost of nutrient loss through rivers and streams from small catchments has not been well researched. Soil degradation in the form of nutrient depletion, is an important factor for the declining agricultural production in Ethiopia (Sileshi Bekele and Holden, 1998). According to Getnet Dubale et al. (2009) soil erosion induced productivity losses are distinct in the Upper Blue Nile Basin. The cost of soil erosion to farmers is two-fold; loss of productivity due to loss of plant nutrients and economic cost of fertilizer in order to compensate the lost nutrients (Gruhn et al., 2000). The physical, chemical and biological effects of soil degradation on the ecosystems and human populations have been researched to some degree, but little research has been done about the economic costs of soil degradation (Görlach et al., 2004), in particular sediment and nutrient loss by rivers from small catchment. Diga District where the two study catchments are found is located in the western Oromia, Ethiopia were low soil fertility is one of the major factors limiting maize production and productivity (Wakene Negassa, 2005). Dapo and Chekorsa streams are tributaries of Didesa River, the largest tributary of the Blue Nile River in terms of volume of water, contributing roughly a quarter of the total flow as measured at the Sudanese border (MWRE, 2010). Conducting such studies in the Blue Nile Basin benefits not only the upper stream community but helps to plan interventions that minimize the offsite costs such as siltation of dams and reduction of water quality for domestic uses. This study was made mainly to understand the processes and cause of soil and its onsite costs to the farmers’ interms of yield lost at micro levels. 1.2. Statement of the problem Soil nutrient depletion has become a major agricultural problem in central highlands of Ethiopia due to improper land management practices. It is understood that it is impossible to achieve food security in the region without overcoming the problem of nutrient depletion (Belayneh Ayele, and Hager, 2010). In the study area, local communities are cultivating the top and bottom of the slopes, aggravating the problems of soil erosion and loss of soil fertility, which are the major challenges of the watershed (Brihanu Zemedin et al., 2010). In some parts of the watersheds, all 2 the top soil has been lost exposing the bed rocks and tree roots. In response to the productivity declines, farmers open a new agricultural land which increases deforestation. The quantity and quality of soil lost by water erosion and the sources of the sediment was not determined. Determining the concentration of major nutrients lost is very helpful for estimating productivity loss and corresponding economic cost. As the on- site economic impacts of sediment loss on the livelihood of the local people have not been estimated, this study attempted to generate such crucial evidence, which can be used to inform the local community and policy makers so that appropriate actions will be taken on the ground. 1.3. Significance of the study The study is an input to the Nile Basin Development Challenge Program of the Challenge Program on Water and Food being implemented in the Blue Nile Basin. Meanwhile, the Ethiopian government has launched the building of the Millennium Dam which is located at the outlet of the Abay River. This study being conducted in one of the major tributaries of Abay River is essential to design and implement suitable management practices to curtail the siltation and eutrophication risks that may affect the Dam. The finding of the study helps the local farmers to recognize the cost of sediment lost and it may assist policymakers to know the “concealed” costs of soil-nutrient losses so as to highlight the potential impacts and benefit of soil-conservation investments on the environment and economy of the local communities. 1.4. Objective of the study The overall objective is to analyze the quantity and characteristics of soil lost by runoff and identify sediment contributory areas. The specific objectives of the studies are: • To estimate the sediment concentration at the outlet of Dapo and Checkorsa watersheds • To analyze the major plant nutrients lost with the sediment • To identify the potential subarea contributors of the sediment to Dapo River • To estimate the crop productivity loss due to soil erosion 3 1.5. Hypothesis Water erosion in the study area is taking the top fertile soil and thereby delivering the major nutrients to the outlet which significantly influences the agricultural productivity of the watershed. 1.6. Scope of the study The study was based on three months of water sampling during the period characterized by high rainfall and sediment concentration in the runoff at monitoring stations. The discharges of the rivers carrying sediment from the watersheds were quantified and chemical properties of the sediment were analyzed in order to estimate the amount of nutrient lost from the catchments. The economic cost of nutrient losses from the watershed was also included in the study to create an easily understandable result for the local farmers and policy makers. The study also included information which is difficult to obtain using manual and digital monitoring techniques in combination i.e. the sources of the suspended sediment transported by rivers whether the dominant source is from the upper or the lower areas of the Dapo watershed. . 4 2. Literature Review 2. 1. Concept of soil erosion Soil erosion caused by water and wind is a widespread problem in both rural and urban areas of the world. Soil erosion is normally a natural process occurring over geological timescales; but where (and when) the natural rate has been significantly increased by anthropogenic activity accelerated soil erosion becomes a process of degradation and thus an identifiable threat to soil (Le Bas, and Kozak, 2007). About 80% of the world's agricultural land suffers moderate to severe erosion, and 10% suffers slight to moderate erosion. Croplands are the most susceptible to erosion because their soil is repeatedly tilled and left without a protective cover of vegetation (Pimentel, 1995). Most studies showed soil erosion is severe in the Ethiopian Highland. FAO (1999) indicated that Ethiopia is among the countries with high degrees of erosion with highest nutrient depletion rates. a. What is soil erosion Christine and Josef (2007) defined soil erosion as the wearing away of the land surface by physical forces such as rainfall, flowing water, wind, ice, temperature change, gravity or other natural or anthropogenic agents that abrade, detach and remove soil or geological material from one point on the earth's surface to be deposited elsewhere’. Soil erosion is normally a natural process occurring over geological timescales; but where (and when) the natural rate has been significantly increased by anthropogenic activity accelerated soil erosion become a process of degradation and thus an identifiable threat to soil. Erosion occurs when soil is left exposed to rain or wind energy. Water is the main cause of erosion in the highlands of Ethiopia particularly during the concentrated rain in three to four months of summer season (Paulos Dubale, 2001). Relevance to this work as it affects the two study watersheds is soil erosion by water known as water erosion. Water erosion depends on four factors: rainfall, soil type, slope gradient, and soil use/vegetation cover (Ballayan, 2000). Raindrops hit exposed soil with great energy and easily dislodge the soil particles from the surface in the form of runoff (Pimentel, 2006). Soil erosion by water is a process in which the detachment of individual soil particles from the soil mass cause a breakdown of the soil aggregates. The detached soil particles would be 5 transported by the water known as surface runoff. Runoff mostly formed when the rainfall intensity is higher than the infiltration rate (Helmecke, 2009). 2.2. General overview of soil loss extent in Ethiopia The excessive dependence of the Ethiopian rural population on natural resources, particularly land, as a means of livelihood is underlying cause for degradation of land and other natural resources (Drechsel et al., 2004). Soil erosion by water represents among the major threats to the long-term productivity of agriculture particularly in the Ethiopian highlands. As a result, productivity is rapidly declining (Tegenu Ashagrie, 2009 and Tilaye Teklewold, 2007). All physical and economic evidence shows that loss of land resource productivity is an important problem in Ethiopia and with continued population growth the problem is likely to be even more important in the future (Berry et al., 2003). There are several studies that deal with the severity of land degradation at the national level in Ethiopia. For example, Shibru Tefera (2010) remarked the relative probability of greater impact of nutrient depletion in Ethiopia, where it is more severe than the other SSA countries. Water erosion was the most important process and that in mid 1980’s 27 million ha or almost 50% of the highland area was significantly eroded, 14 million ha seriously eroded and over 2 million ha beyond reclamation (Berry et al., 2003). The total soil eroded within the landscape in the Abay Basin is estimated to be 302.8 million tons per annum out of whichb101.8 tons per annum was estimated to be from cultivated land (Fistum Hagos, 2009). Berry et al. (2003) estimated the rate as less as 130 t-1ha-1yr for cropland and 35 t- 1ha-1 yr averages for all land in the highlands, but even at the time these were regarded as high estimates. According to Getnet Dubale et al. (2009) soil loss in the Blue Nile Basin is above 2.00- 4.00 t /km2 /yr. The same author estimated that about 24 Million ton per year sediment is deposited in river channels within the Upper Blue Nile. Another study by Biniam Biruk (2009 ) estimated loss of 16-50 t-1ha-1yr from the Ethiopia highlands. According to Hurni (1993) soil losses in the Ethiopian highlands may reach as high as 200-300 t-1ha-1yr. According to Getnet Dubale et al. (2009), the amount of sediment yield delivered at Ethiopia Sudanese boarder from the upper Blue Nile is estimated to be 62 Million ton per year. 6 The loss of nutrient-rich top soil by water leads to loss of soil quality and hence reduced crop yield. Soil erosion by water and its associated effects are therefore recognized to be severe threats to the national economy of Ethiopia. In Ethiopia, particularly on the Gumera watershed, the study by Mequanint Tenaw and Seleshi Bekele (2009) showed that about 72% of erosion potential area with an average annual sediment load ranging from 11 to 22 t/ha/yr exceeding tolerable soil loss rates of Ethiopia. The same author remarked that sheet and rill erosion are by far the most widespread kinds of accelerated water erosion and principal cause of land degradation in the country and their combined effect significantly affect agricultural production and productivity. Berry et al. (2003) estimated a loss of $106 million a year or about three percent of agricultural GDP from a combination of soil and nutrient loss. Most of the sediment in the Nile flows from the Ethiopian Highlands through the Blue Nile and Atbara River. Nearly all of the sediment (~ 90%) enter into Sudan from the Blue Nile during the flood season (July - October) (Abdalla Abdelsalam, 2008). 2.3. Suspended sediment River suspended-sediment concentrations provide insights to the erosion and transport of materials from a landscape, and changes in concentrations with time may result from landscape processes or human disturbance. The behavior of suspended sediment in watercourses is often a function of energy conditions, i.e. sediment is stored at low flow and transported under high discharge conditions. However sediment transport rates are also a function of sediment availability (Baca, 2002). Traditionally, these dynamics are characterized by empirical relationships between suspended sediment concentration and discharge. These relationships are normally not homogenous in time, neither within nor between events (Baca, 2002). Experimental data has shown that there are three common shapes of the hysteresis loops encompassing (i) clockwise, (ii) counter clockwise and (iii) , though it is possible to obtain loops which are (iv) single valued or (v) single valued plus a loop (Sander et al., 2011). 7 2.4. Nutrient depletion Soil nutrient availability changes over time. Soil fertility is one of the key factors in determining agricultural output, and soil fertility depletion is seen as the most important process in the land degradation equation and a primary constraint to improving food security in developing countries (Drechsel et al., 2004). Of the global cultivated area for the crops in the year 2000, 56% was affected by N deficit at an average rate of 17.4 kg-1 ha-1yr, 80% by P deficit at that of 5.0 kg-1ha-1yr and 56% by K deficit at that of 38.7 kg-1ha-1yr(Tan et al., 2005). The same author also remarked that at the global scale, a shortage of N, P, and K was observed in developing and least developed countries. Developed countries were still deficit in N and P in an area of 108 Mha (52%) for N and 151 Mha (73%) for P despite being less serious than in other countries. Table 1: Global nutrient loss rate classes (kg-1ha-1year) Class N P2O5 K2O Low <10 <4 <10 Moderate 10-20 4-7 10-20 High 21-40 8-15 21-40 Very high >40 >15 >40 Source FAO (1999) The above nutrient deficits were due to the considerable nutrient depletion from cultivated land. About 86 percent of the countries in Africa lose more than 30 kg-1ha-1yr of NPK (Henao and Baanante, 1999). Likewise, Gruhn, et al. (2000) indicated that in Sub- Saharan Africa net annual nutrient depletion was estimated at 22 kg-1ha of nitrogen, 2.5 kg-1ha of phosphorus, and 15 kg-1ha of potassium during 1982-84. And according to the report of World Bank (1999) the estimate is much greater in Sub-Saharan Africa reaching a net loss of about 700 kg-1ha of nitrogen, 100 kg- 1ha of phosphorus, and 450 kg-1ha of potassium in about 100 Mha of cultivated land over the last 30 years. In addition, Henao and Baanante (1999) suggest that nutrient mining may be accelerating. It is well researched that erosion plays a major role in nutrient removal from cultivated land. 8 2.5. Transportation of nutrient to water body Runoff carries some inorganic nitrogen, primarily as nitrate and ammonium, at concentrations that are commonly 3 ppm or less (Castro, 2004, Wortmann 2006). The same authors indicated Nitrate-N is generally leached into the soil and ammonium nitrogen becomes attached to soil particles with precipitation that occurs before runoff begins In addition to creating water deficiencies, runoff and soil erosion cause shortages of basic plant nutrients, such as nitrogen, phosphorus, potassium, and calcium, which are essential for crop production. Pimental et al., (1995) showed a ton of fertile agricultural topsoil typically contains 1 to 6 kg of nitrogen, 1 to 3 kg of phosphorus, and 2 to 30 kg of potassium, whereas a severely eroded soil may have nitrogen levels of only 0.1 to 0.5 kg per ton. They also suggested that wind and water erosion selectively remove the fine organic particles, leaving behind large particles and stones. Eroded soil typically contains about three times more nutrients than the soil left behind. Similarly, Jun et al. (2005) indicated that the entire nutrient in surface soil had lower values than that in sediment. There are abundant examples which demonstrate how sediment quality has been affected in response to human activities. A well-known example is the widespread particulate phosphorus increase in many agricultural river basins in the world Philip et al. (2010). They also remarked that fertilizer use and accelerated soil erosion on agricultural river basin have resulted in elevated sediment inputs and phosphorus concentrations in stream and lake beds. According to Sharpley et al. (2000) soil P levels are higher in the top 5 cm of the surface soil. Soil detachment and transport in surface runoff preferentially erode finer particles. This results in eroded material with higher total phosphorus (>0.45) content in the runoff compared to the soil in the source area. In addition, overland flow is efficiently removing high concentration of P, because of the largest concentration of P in the surface layers, and the greatest concentrated hydrologic energy on the soil surface than the subsurface (Zaimes, and Schultz, 2002 et al., 1998 work). 9 The removal of soil particles, from the topsoil can have a devastating impact on overall soil organic matter levels because organic materials are concentrated in the surface layer of the soil (Van-Camp, 2004). Nitrogen is lost to surface waters and ground waters through overland flow and leaching and below-ground movement of nitrate (Wortmann, 2006). The amount of nitrogen delivered depends on the volume of drainage water and nitrate concentration in the soil solution (Wortmann, 2006). For example, it has been estimated that in Albania, water erosion washes away 60 million tons of course materials every year. These comprise 1.2 million tons of organic matter, 100,000 tons of nitrates, 60,000 tons of phosphates, and 16,000 tons of potassium (Van- Camp, 2004). 2.6. On site impact of soil erosion The impacts of soil erosion can be on-site and off-site (Figure 1). The farmer will probably be more concerned about the former, which occur on the eroded land itself. They describe the decline in crops productivity, the reduction of the soil’s water holding capacity, its nutrients and organic matter, which often revealed as a decline of productivity (Helmecke 2009). On-site economic cost Figure 1: On site effects of soil erosion Source (Helmecke, 2009). 10 2.6.1. Economic impacts of soil erosion Plants need relatively large amounts of nitrogen, phosphorus, and potassium. These nutrients are referred to us macro nutrients, and they are most frequently supplied to plants as fertilizers. When insufficient, these primary nutrients are most often responsible for limiting crop growth. Their balance in soil depends on the rate which they naturally regenerate, applied in the form of fertilizers and the rate at which they are removed from the soil system by plants and soil erosion. The cumulative effect of yearly negative nutrient balances on crop yields is often seen through the impact of soil erosion on productivity (Gruhn et al., 2000). Table 2: Global estimated impact of soil erosion on crop production Net production (Mg Estimated production loss Estimated production Commodity 106) (%) if there were no erosion (106Mg) Cereals 1896 10 2086 Soybeans 126 5 132 Pulses 56 5 59 Roots and Tubers 609 12 682 Total 2687 32 2959 Source (Helmecke 2009) In 1995, a total production loss of 32 per cent was estimated to have resulted from soil erosion (Table 2). Some deficiency caused by erosion can be temporarily compensated by increased application of fertilizer and irrigation (Pimentel et al., 1995) but to completely restore the original soil productivity it often needs long physical and biological rehabilitation periods. However, farmers often aim for short-term results and might therefore tend to increase fertilizer input as much as they can afford. Although this might help to cope with the temporary productivity loss it leads to other long term damages (Helmecke 2009). Van-Camp, (2004) suggests that more fertilizer and organic manure are needed on agricultural land on which intensive erosion occurs to counteract the losses caused by soil erosion, compared to the requirements in non-eroded areas. Soils in all major maize growing regions of the country are 11 depleted of nutrients, thus demanding high soil amendments with nitrogen and phosphorous (Kebede Mulatu et al., 1993). Decline in soil fertility due to depletion of macro nutrients in the country is therefore eradicating production including maize production. Loss of soil productivity leads to reduced farm income and food insecurity, particularly among the rural poor and thus continuing or worsening poverty (Shibru Tefera, 2010). In least developed countries, productivity reductions were equivalent to 27% of the average crop yield in the year 2000. And the average yield reduction from N, P, and K deficits was 35% in least developed countries, 27% in developing countries, and 11%. Erosion can decrease rooting depth, and plant-available water reserves (Lal, 1987). Thus, the exposed soil remaining will be less productive in a physical sense. These effects may be cumulative and may not be revealed in the short term. Erosion may also affect yields by influencing the micro-climate (Eaton, 1996). Soils that suffer severe erosion may produce 15 to 30% lower corn yields than uneroded soils, and with fertilization, the yield reductions range from 13 to 19%. Similarly, once the organic matter layer is depleted, soil productivity and crop yields decline because of the degraded soil structure and depletion of nutrients. For example, the reduction of soil organic matter from 4.3 to 1.7% lowered the yield potential for corn by 25%in Michigan (Pimental et al., 1995). Therefore, crop yields on severely eroded soil are lower than those on protected soils because erosion reduces soil fertility and water availability. There is strong evidence that yield decline with erosion follows a curvilinear, negatively exponential form. In other words, there is a sharp initial decline from a status of high productivity, followed by successive stages of decreasing impact. In Ethiopia, soil erosion in 1990 was estimated to have cost (based on1985 prices) nearly 40 million Birr (ETB) in lost agricultural production. Thus in 1990 approximately 17% of the potential agricultural GDP was lost because of soil degradation. The permanent loss in value of the country's soil resources caused by soil erosion in 1990 was estimated at ETB 59 million. This is the amount by which the country's soil stock should be depreciated in the national accounts or which should be deducted from the country's Net National Income (Fistum Hagos, 2009). 12 Investment in measures to reduce degradation is expensive both in terms of improving soil (fertilizers, manure, crop residues) and structures such as terraces, grass lines and hedges that all require investments in labor. Decisions to invest therefore have to be made relative to the benefits that are both on-farm and off-farm, while the investment costs are usually borne on-farm (Berry et al., 2003). 2.7. Valuing soil nutrient loss In order to plan a better environmental decision-making policy, the economic valuation of environmental problems is important. For this reason, soil erosion by water which is considered as a major environmental threat to the sustainability and productivity of agriculture (Pimentel et al., 1995) is the main focus of many countries. Soil deterioration makes itself felt in different ways, and there are different methods of classifying the economic impacts of soil degradation. Different impacts can be classified spatially into on-site and off-site effects, distinguished according to the economic values that are affected (Görlach et al., 2004). Likewise he added those impacts may also be grouped according to causality as direct and indirect impacts. The costs of loss of natural capital are borne at the level of individuals, communities and by the broader economy. But this loss of natural capital also results in changes in economic, human, social, and land capital, the value of investment in land management ( Berry et al. ,2003). Thus, the majority of empirical estimates have centered on the impact that soil degradation has on agriculture and forestry (Görlach et al., 2004), and also here the study concerns the direct, on-site economic effect. FAO (1999) remarked that the estimates of cost could be based primarily on the measurement of two variables: production loss or replacement cost. The basic premise of the replacement-cost approach is that the costs incurred in replacing productive assets damaged by an environmental impact can be measured. These costs can be interpreted as an estimate of the benefits presumed to flow from measures taken to prevent those damages from occurring. The replacement cost is a 13 popular method of assessing the value of soil erosion. To value nutrients via fertilizer prices requires either a translation of the lost nutrients into marketed fertilizer types or an expression of fertilizers in nutrient units (Gruhn et al., 2000). In addition, a number of studies have considered the cost of replacing lost nutrients. Replacement cost is the cost of additional inputs (basically fertilizers) used by farmers in order to maintain production levels on the degraded soils (Görlach et al., 2004). To assess by how much erosion has being causing on site economic impact, it is necessary to consider the multiple factors that influence erosion rates as well as soil component and other agro-ecological conditions prevailed in the specific area that affect productivity (Pimental et al., 1995). A partly, the approach of replacement cost cannot consider this concept whereas estimating the approximate production loss is better. As a result, to estimate onsite economic cost of soil loss by runoff, production loss instead of replacement cost was the concern of this study. Crop yields on severely eroded soil are lower than those on protected soils because erosion reduces soil fertility and water availability. For example in some parts of India corn yield on some severely eroded soils have been reduced by up to 24% and 65% in the Southern Piadmont of Georgia (Pimental et al., 1995). Production loss is the reduced productivity of the soil as a consequence of degradation, which could be expressed as a percentage of production from the undegraded soil (FAO, 1999). Soil erosion can reduce crop production up to 30% (Louis, 2011). Many of these studies are agronomic, focusing on agricultural yield losses associated with soil degradation. FAO (1999) reported that for erosion and soil fertility decline, the assumptions are: a 5-10% production loss for a "light" degree of degradation, 20% for "moderate" and 75% for "strong" degradation. When erosion by water and wind occurs at a rate of 17 tons -1ha-1 year, about 75 mm of water and 462 kg of nutrients are lost per hectare. As a result, an additional $100 /ha would be required for fertilizers to replace the lost nutrients. In some part of the world, where irrigation is not 14 possible or fertilizers are too costly, the price of erosion is paid in reduced food production (Pimental et al., 1995). However, previous research has put much emphasis on the importance of N and P lost by the Rivers for plant nutrition. For examples, Kogbe and Adediran (2003) and Alley (2009) remarked that N is without doubt the most significant nutrient for high maize yields and its deficiency limits production more than any other nutrients and P deficiency also has drastic effects on the maize yields. 2.8. Sediment fingerprinting The targeting of sediment management strategies is a key requirement in developing countries because of the limited resources available. Such targeting is, however, hampered by the lack of reliable information on catchment sediment sources. There is an increasing need for reliable information concerning the source of the suspended sediment transported by rivers. Such information is required both to design effective sediment and non-point pollution control strategies and to provide an improved understanding of erosion and suspended sediment transport within a basin which is an essential precursor to establishing sediment budgets, developing distributed sediment yield models, and interpreting sediment yields in terms of landscape evolution (Walling, 1993). Sediment fingerprinting has been developed by researchers over the past three decades for watershed sediment transport research. Sediment fingerprinting is founded on the premise that spatial and temporal variations in sediment properties directly reflect spatial and temporal variations in the relative contributions of sediment from distinguishable sources (Collins et al., 2001). This technique makes use of chemical and physical properties of the sediment to trace its source. It involves, firstly, the selection of a physical or chemical property which clearly differentiates potential source materials, and, secondly, comparison of measurements of the same property obtained from suspended sediment with the equivalent values for the potential sources, to establish the likely source of the sediment (Figure 2) (Walling, 1993 and Collins, 2001). Sediment fingerprinting is a method to identify sediment sources in a watershed and allocate the amount of sediment contributed by each source through the use of natural tracer technology with a combination of field data collection, laboratory analyses of sediments, and statistical modeling techniques. This method utilizes one or more unique physical or biogeochemical properties known as natural tracers (Davis and Fox, 2009). 15 Effective precipitation event Geological subareas Spatial s o u r c e Tributary subareas E r o s i o n o f c a t c h m e n t s e d i m e n t s o u r c e s Surface and subsurface S o u r c e s t y p e Land use type and channel bank Mixing sources during sediment delivery Se diment flux at catchment outlet Comparison of source Sediment source material and suspended ascription sediment samples using finger printing properties Figure 2: A conceptual model of sediment fingerprinting Source: Collins and Walling (2002) 16 3. Materials and Methods 3. 1. Description of the study area 3.1.1. Diga area The study was carried out at Diga district, East Wollega Zone of Oromia Regional State. It is located at about 346 km from Addis Ababa and 15km from Nekemte town to the West (Figure 3). The total area of the District is estimated at 40,788 hectares. Figure 3: Map of Diga District Source WAO, 2011 According to Joshua, et al. (2010), the District is stratified into two agro-climatic regions; the middle altitude to high altitude which ranges in between 2100-2342m.a.s.l and the low land which range in between 1200-2100 m asl. According to the District Agricultural Office report in 2010, middle to high altitude and the low lands covers 42% and 58% of the district, respectively. The report also shows topography of the district where the study area found is characterized as flat, gentle slope, steep slope, very steep slope and hill. The mixed cropping system is common in the district. In the lowlands maize is the dominant field crop followed by sorghum (sorghum bicolor), millet ( Eleusine coracana) and sesame (Sesamum indicum L) while perennial crops such as coffee ( coffea arabica) and mango 17 (Mangifera indica) are also prevalent. In the midland, tef, millet and maize are important in that order. Livestock keeping is common allover (Brihanu Zemedin et al., 2010). And according to Diga District Water Resource Office (2010), the land use of the area is divided into arable land, grazing land, forest land, bushes and shrubs, construction and others which are yet to be classified. The high land areas of Diga District receive rainfall varying from 1376- 2037mm, and the annual mean temperature varies from 14.60 to 30.40 C (Joshua et al., 2010). Regarding water resources, the district has 26 perennial and unprotected rivers and 167 streams out which 75 are annual while 29 are protected for drinking and other uses and 138 are unprotected. There is only one unprotected reservoir (Diga District Water Resource Office, 2010). The watersheds are generally located at the high altitude and receive high rainfall during rainy season, which begin in late April, and ends in early September. 3.1.2. Characteristics of Dapo and Chekorsa watersheds a. Size and location Dapo and Chekorsa rivers are among the 26 perennial rivers found in the District. The catchment area of Dapo and Chekorsa are 16.2 Km2 and 5.60Km2 and their altitude ranges between 1,347 – 2011 and 1266 – 1430m asl, respectively. Sampling locations of the two rivers is for Dapo River at bridge on the Digga to Arjo Gudatu road at 09o03.141’ N, 36o17.650’E (Figure 8) whereas for Chekorsa at bridge on the Lelisa Dimtu old State farm at 09o03.410’N; 36o13.978’E. b. Physiography a b Figure 4: (a) Dapo watershed outlet and (b) land use and land cover condition around the outlet Photo credit: (Brihanu Zemedin, 2011) 18 Both Dapo and Chekorsa rivers are tributaries of Didesa River the largest tributary to the Blue Nile River in terms of water volume (MWRE, 2010). The watersheds are adjacent and these rivers drain separate in the same direction. Both Rivers has numerous first and second order streams flowing directly to the Rivers. Similar to CRC, the physiographic, land use and land cover condition in the downstream of DRC, around the water level gauging site, consists of mango trees and sparsely populated natural vegetation cover, lowland maize fields and, flat grazing areas in the downstream side of the bridge (Figure 4) (Brihanu Zemedin, 2011). Different to Checkorsa River, Dapo River has well established natural riparian zone Figure 15. Figure 5: Land use land cover of Dapo watershed, Source: IWMI (2012) The dominant crop types of DRC were maize, sesame, and finger millet and about one third of the watershed area is covered by forest located at the most upper part of the watershed (Figure 5). But, in the CRC no dence forest is found. All parts of the watersheds have being used for agricultural activities. Soil textural class of DRC is clay loam whereas silt clay loam for CRC (Joshua, 2011). 19 3.2. Methods 3.2.1. Data gathering Both primary and secondary data were collected for this study to estimate of several parameters illustrated conceptual framework of Figure 10. Hydrological measurements was conducted at the two monitoring stations to generate the following information: discharge (Q) of the rivers, suspended sediment concentration (SSC), suspended sediment load (SSL), its chemical analysis, and fertilizer yield response data for the study area were obtained from different research results under similar agro-ecological conditions. 3.2.2. Selection of runoff sampling site Expert from the International Water Management Institute (IWMI) have identified the bridge on the main highway that goes from Diga to Ghimbi for DRC and the bridge from Arjo Gudatu to Lalisa Dimtu for CRC are ideal locations for establishing flow monitoring stations. The bridges are wide and all flows were contained inside the culvert of the bridges. 3.2.3. The study period The study was from the onset (July) to the offset rainfall (September) 2011 which makes a three months period. Each month was divided into three decades (d) 10 consecutive days. 3.2.4. Runoff sampling Based on the concept of Gierke (2002) the flow rate of the river at the outlet was determined using current meter (Model 0012B Surface Display Unit and Model 002 Flow Meter (Figure 7: A and B respectively) as well as, measured depth of the rivers using 1.5m wading rod (Figure 7: E). There were 9 points with 0.5m intervals for the Dapo River (DR) (Figure 6) and 5 points with 0.75m intervals for Chekorsa River (CR) across the rivers at which flow rates and depths (h) were measured simultaneously. Using these depth records, cross sectional areas (Figure 6) was calculated. The cross sectional areas were multiplied with the average flow rates at each point (Equation1a) and then the volume (Q) of the runoff passing the outlet of the watersheds were calculated using equation 1b. qi = vi*ai (a) Q  ∑ qi (b)…………………………………………………….Equation 1 20 where: qi= discharge at each cross sectional area (m3sec-1) vi=flow velocity at each cross sectional area (msec-1) ai= cross sectional area at each point (m2) Q= Total discharge (m3sec-1) A1 A2 A3 A4 A5 A6 A7 A8 h1 h2 h3 h4 h5 h6 h7 h8 h9 h: Depth of the river at nine points across the river (m) Figure 6: River cross-section shown and sub-cross sectional areas where flow velocities were measured Q = c (h + a) b …………………………………………………………………………………….Equation 2 Where: Q = discharge (m3sec-1) h = measured water level (m) a = water level (m) corresponding to Q = 0 ci = coefficients derived for the relationship corresponding to the station characteristics b= coefficient deriver for the power relation the station characteristics Finally, discharge rating curve were developed by fitting the relationship of measured gauge to discharge into power curve (Equation 2) for the two Rivers. And having water levels measured throughout the study period by the installed staff gauge (Figure 7: E), the discharge for each was calculated from the equations of the curves. 21 A B E C D Figure 7: Current Meter model used for measuring river flow and staff gauge (E) at DR monitoring station Photo Credit:Brihanu Zemedin (2011) 3.2.5. Water sampling and storage Depth integrated runoff water were collected manually from catchments at the monitoring stations using one liter plastic bottle three times per day to represent the daily runoff. The daily samples were mixed and two litters were subsampled and bulked in a 20 liter Jerry Can for 10 consecutive days. The bulked sample was kept in the nearby soil laboratory. The bulked water sample were labeled properly and kept in the refrigerator at 4OC in order to minimize further chemical and physical changes of both the sediment and the water (Annex 8). 3.2.6. Estimation of sediment load The sediment in the collected water was allowed to settle down before the top 18L were decanted laboratory beakers and the remaining two litters which contain most of the sediment were filtered using watman filter paper (Annex 8). S=Ms/Vw and SSL=SxQ……………………………………………………………. Equation (3) Where; S: suspended sediment per liter (gm/L) 22 Ms: mass of suspended sediment left on Watman filter paper (gm) Vw: volume of water collected per decade (L) SSL: suspended sediment load per decade (Kg/d) Q: mean discharge of the rivers per decade (L/d) The sediment remained on the filter paper were weighted using digital weight balance for each decade separately. Then, the amount of soil loss per decade was calculated from the estimated mean discharge of water passing the gauged sites for each decade using Equation 3. 3.2.7. Chemical analysis Table 3: Methods and procedure used for the chemical analysis sediment and water Sample Parameter Method Reference OM Wet oxidation/ Walkley-Black Jackson, 1967 TKN Modified Kjeldahl digestion Dalal et al. 1984 Soil NO3-N Magnesium Oxide-Devrda’s alloy Maiti, 2004 NH4-N Magnesium Oxide-Devrda’s alloy Maiti, 2004 P2O5 Alkaline Extraction of Olsen Method Olsen and co-worker ( 1954) Texture Hydrometer Bouyoucos 1962 Dissolved Phenate method using Spectrophotometer; ammonia Modele Eleco SL-160 Double beam UV Patnaik (2010) Water Dissolved Spectrophotometer; Modele Eleco SL-160 nitrate & Double beam UV Patnaik (2010) phosphorus After decantation and filtration process, chemical analysis had been conducted both on the soil and the water at Ambo University. On the air dried soil, the concentrations of OC, total nitrogen, available phosphorous, NH4-N, NO3-N were determined using standard procedures (Table3). Water quality analyses were also conducted for the dissolved PO -3 4 , NH4-N and NO3-N (Table3). 23 3.2.8. Suspended sediment fingerprinting Some part of the watershed surface soil was chemically analyzed by Joshua et al. (2010) representing subareas of Dapo Watershed i.e. transect one representing the lower part of near the out let and transect two and three representing the middle to upper part). Transect for DDS is found between 1353 -1499 and US transect located between 1500 and 1645 (Figure 8).Tracers properties of transect two and three were pooled as showed on Figure 17 and 18, and represented as the upper part of the watershed were agricultural activities practiced excluding the uppermost natural forest ( Figure 5 and Annex 2). Then comparison between the sediment and surface soil properties was done following the conceptual model of Collins and Walling et al. (2001) (Figure 9). The relative contribution from the two transects are done using the assumption of Collins and Walling et al. (2001). Since fingerprinting properties of any suspended sediment samples are dependent upon the corresponding properties in the source materials, the relative proportion of the source materials from downstream (DS) and (US) was estimated using the Mixing model (Equation 4). Monitoring station Figure 8: Points where the surface soil samples were taken in the DRC Source: Brihanu Zemadin et al., 2010 24 Selecting the potential finger printing Bulking of water sample for 10 days properties of source soil from Joshua’s work Filtration of sediment from water S tatistical analysis to identify optimum composi te fingerprinting for Analysis for properties comparing discriminating individual sediment sources optimum composite fingerprinting Comparison of sources and sediment sample fingerprint properties using numerical modeling Sediment source apportionment Figure 9: Summary of fingerprinting procedure of Walling and Collins (2002) Ci = Psu.Sug.Zu.Ou+ Psd.Sid.Zsd.Osd ………………………………Equation 4 0