Abstract This study was conducted in Mizewa watershed which is located in Blue Nile Basin (BNB) to quantify and characterize the suspended sediment (SSL) and to estimate onsite �nancial cost of erosion in terms of yield reduction taking maize as representative crop. For this purpose, discharge measurement and runo� sampling was made during the rainy season of 2011 at the outlet of three sub watersheds (lower Mizewa (MZ 0), Upper Mizewa (MZ 1) and Gindenewur (GN 0)) . The samples were �ltered to separate the sediment which was sub sampled for determination of suspended sediment concentration(SSC), total Kjeldal nitrogen (TN),organic carbon (OC),NO 3

-,NH 4+and

available phosphorous (P) content. The onsite cost of erosion was estimated based on productivity change approach (PCA) focusing on available NP losses. The result revealed that the SSC and its NP content varied in space and time, in which lower SSC occurred towards the end of the rainy season. The mean seasonal discharge was found to be 2.12±0.75,1.49±0.52 and 0.57±0.20 m3/sec at MZ 0, MZ 1 and GN 0 stations in that order while the corresponding sediment concentration was 510±370 mg/l, 230±190 mg/l and 370±220 mg/l. This lead to the total suspended sediment loss (SSL) of 4 ton/ha/year,2 ton/ha/year and 3 ton/ha/year from the respective sub watersheds. The on-site �nancial cost due to N and P lost associated with SSL was estimated to be 200$/ha, 186$/ha and 227$/ha from MZ 0, MZ 1 and GN 0 watersheds respectively. The study revealed that the economic impacts of soil erosion which is variable based on the characteristics of the land resources and management practices deserve due attention. The result may help in sensitizing both farmers and decision makers about the risk of soil erosion and in targeting management practices to overcome the challenges. Keywords : Blue Nile basin, Soil Erosion, Runo�, Sediment concentration, Nutrient loss

Introduction Blue Nile Basin is heavily affected by land degradation problems. Overpopulation, poor cultivation and land use practices are the major cause, which resulting in significant loss of soil fertility, rapid degradation of the natural eco-systems, significant sediment and nutrient depositions in lakes and reservoirs (Tamene et. al,2006).To supplement rainfed agriculture with irrigation, a massive surface water harvesting effort has been undertaken in the dry lands of Ethiopia in the last few years. However, most of the water harvesting schemes is under serious siltation and dry up (Amanul, 2009) due to upstream land degradation mainly soil erosion. Circumstances in turn lead to a reduction of productivity because of risky land use exercise. This study was conducted in Mizewa catchment to determine loss of SSL with associated nutrients in runoff, input to the NBDC Program on Water and Food being implemented in the BNB. The result was helpful for estimating productivity loss and corresponding economic cost which provide crucial evidence to inform the land users and policy makers to take actions. Taking this into considerations, it is essential to design and implement suitable land management practices to curtail optimum utilizations of resources. Therefore, this study was conducted to quantify suspended sediment as well as nutrient loss and to estimate impact of nutrient lost on crop yield along with its financial cost in Mizewa catchments.

Results & Discussion The average discharge was 2.12±0.75, 1.49±0.52 and 0.57±0.20 m3/sec with total �ow volume of 18.34(106m3), 12.87(106m3), and 4.92(106m3) per season at MZ 0, MZ 1 and GN 0 rivers respectively. Peak daily discharge was observed around August 17- August 26 in the three stations, may be due to excess in saturation and full vegetation land coverage. Mean �ow rate was statistically signi�cant between sites (t=2.68 and P≤0.025 between MZ 0 and MZ 0, t = 6.58 and P≤0.000 between MZ 0 and GN 0 and t=5.54 and P≤0.000 between MZ 1 and GN 0 rivers. Mean SSC during the rainy season was 510±370 mg/l, 230±190 mg/l and 370±220 mg/l from MZ 0, MZ 1 and GN 0 stations respectively and signi�cant variations was observed between time and space. In MZ 0 mean SSC varied from 67 mg/l to 900 mg/l per decade. SSC coe�cient of variation between periods was 73%, 82 % and 61% for MZ 0, MZ 1 and GN 0 stations, respectively implying that SSC in MZ 1 was more variable over time than MZ 0 and GN 0 rivers. Statistical test for temporal and spatial variability of mean SSC (mg/l) over decades has shown that there is a signi�cant variation between periods (F=4.51 and p≤0.0032) and sites (F=5.61 and p≤0.013) between the three sites. From the general trend of hysteresis loop, it is possible to conclude for MZ 0 and MZ 1 stations for a given runo� discharge, lower SSC values occur towards the end of the rainy season than at the beginning. This may be due to an increase in vegetation cover which decreased sediment sources; though, discharge has positive correlation to SSC (Amanuel, 2009). Walling (1977) indicated that scatter SSC –Q relationship is typical of ‘supply-limited’ or sediment sources conditions in its upper catchments which can be explained by clockwise hysteresis e�ects of sediment transport systems. This is mostly attributed to sediment depletion in upper slopes of a basin, sometimes even before the runo� has peaked sediment is derived from the bed and banks of the channel or areas adjacent to the channel (Ongley,1996). GN 0 station display counterclockwise hysteresis loop early in runo� (from D1 to D4) reversing to clockwise afterward. This may be caused by a variety of factors related to sediment sources and discharge conditions; initial sediment contribution from the streambed and its banks, a delayed contribution of sediment from sub catchment and occurrence of dry periods (Seeger et al., 2004). Mean clay, silt and �ne sand content of sediment was 42±10, 39±5, 20±15% for MZ 0; 37±13, 36±4, 27±17% for MZ 1 and 40±1138±3, 22±15 % for GN 0 station respectively. In this study concentration of silt and clay have a decreasing trained with time; while, the proportion of �ne sand in the sediment increases towards end of runo� sediment monitoring seasons (Amanul,2009). Particle size was highly related to the sediment loss, this is because the active fraction of sediment is usually cited as that portion which is smaller than 63µm (silt + clay) (Lal , 998). Statistically textural content of suspended sediment was not signi�cant (P<=0.05) between the stations. Signi�cant correlation of sediment texture was observed with all NC and SSC in all stations. This implies that �ne soil particles play great role in the process of erosion in the watershed (Lal , 1998); re�ects the rate and severity of erosion in the study watershed and it was a challenge for the livelihood of the poor farmers. This is because nutrients are strongly adsorbed to the �ner soil fractions, which are preferentially transported by the sedimentation processes because of their high speci�c surface areas (Haregeweyn et al.2008). Mean soil loss from MZ 0 catchment using RUSEL model was estimated to be 13.21 ton/ha/year, nearly 80 % of soil loss was from 15 - 50 % slope class dominated by �nger millet, nigger seed and te� cultivation, and this is above tolerable soil los limit in Ethiopia. Cultivated land use system contributing more than 60% of total soil los (mean=20.4 ton/ha/year) which was about 1.5 times mean annual soil loss rate in Mizewa. Seasonal OC concentration was 23.8±10.1 g/kg(CV= 42%) in MZ 0 (Fig .2.a), 19.76±9.44 g/kg(CV= 48%) in MZ 1 (Fig .2.b) and 10.37±6.12 g/kg (CV= 59%) in GN 0 (Fig .2.c) monitoring stations. This OC lost revealed that area speci�c organic carbon loss of 97 kg/ha, 35 kg/ha and 30.4 kg/ha from the respective watersheds. This may result in a serious detrimental e�ect land productivity in both short and long terms in which threatening the food security of the local people, this is because in the process of erosion, loss of OC leads to depletion of soil and other nutrients associated with the organic fraction (Lal, 1998). In addition to OC loss, TN concentrations in suspended sediment was vary from 0.43 g/kg in MZ 1 (Fig .2b) to 4.46 g/kg at MZ 0 (Fig .2a) with mean value of 2.05±0.87 in MZ 0 (CV= 42%), 1.68±0.85 in MZ 1 (CV= 50%) and 0.88±0.47 in GN 0 (CV= 54%) stations. Area speci�c TN lost through runo� sediment in each monitoring station was 8.4 kg/ha at MZ 0, 3.1 kg/ha at MZ 1 and 2.5 kg/ha at GN 0 only in monitoring period. Signi�cant plant available nutrients were lost in associated with runo� and sediment. Mean area speci�c plant available NP lost was (2.3 kg N/ha, 4.0 kg P2O5/ha), (1.6 kg N/ha, 4.1 kgP2O5/ha) and (2.3 kg N/ha, 4.8 kgP2O5/ha) from MZ 0, MZ 1 and GN 0 catchments respectively. Statically, there was a clear di�erence in the concentration of sediment nitrate (F=6.23, p=0.006) and NH 4

+ (F=3.85, p=0.034) across stations during the study period. There was no temporal variation in available plant nutrient concentration regardless of the stations except only for soluble phosphate (F=10.47, p≤0.000). This available nutrient species composition and magnitude varied widely within the watershed which could be caused by several factors that needs further research and detail data to come up the with control variables for these di�erences among stations. In this study, plant available N and P lost through runo� suspended sediment was responsible for signi�cant economic onsite costs, and this was re�ected in maize grain yield reduction during monitoring season. Regression equationsd between maize grain yield and additional N and P2O5 application based on Tilahun Tadesse et.al (2007) data source was used as a bridge to link soil nutrient lost with grain yield loss of maize crop. R 2 (in the equations shows a wide variation of yield response to the almost equivalent amount of fertilizer level. The show that mean grain yield with no P and N fertilizers were 2691kg/ha and 2537kg/ha, respectively. Correspondingly, the lost net maize grain yield due to the loss of available N and P were (134 kg/ha, 320 kg/ha, total 453 kg/ha) from MZ 0, (93 kg/ha, 328 kg/ha, total 421 kg/ha) from MZ 1 and (134 kg/ha, 382 kg/ha, total 453 kg/ha) from GN 0 watersheds. Taking 20 quintal/ha average maize grain yield productivity in the study area (according to CSA, 2011), the lost maize yield due to available N and P2O5 account 23%, 21% and 26% productivity reduction from MZ 0, MZ 1 and GN 0 watersheds correspondingly. This e�ect of soil erosion on grain yield is above the estimates of (Helmecke, 2009) cereal (10%), pulse (5%) and root (12%) crops production loss estimated at global scale. As a result a farm enterprise having a hectare of land with maize cultivation in the study area has a pro�t loss of about 200$/ha from MZ 0, 186$/ha from MZ 1 and 227$/ha from GN 0 watershed in consequence of plant available N and P lost through runo� soil erosion process only in one particular rainy season.

CHARACTERISTICS AND ESTIMATED ONSITE COSTS OF SEDIMENT LOST BY RUNOFF

FROM MIZEWA CATCHMENTS, BLUE NILE BASIN

CHARACTERISTICS AND ESTIMATED ONSITE COSTS OF SEDIMENT LOST BY RUNOFF

FROM MIZEWA CATCHMENTS, BLUE NILE BASIN Getnet Taye 1, Enyew Adgo 1, Teklu Erkossa 2

1 Bahir Dar University, collage of Agriculture and Environmental science, P.O.Box 78, 2 International Water Management Institute, Adiss Ababa, Ethiopia, P.O.Box 5689

Getnet Taye 1, Enyew Adgo 1, Teklu Erkossa 2 1 Bahir Dar University, collage of Agriculture and Environmental science, P.O.Box 78,

2 International Water Management Institute, Adiss Ababa, Ethiopia, P.O.Box 5689

Methods The study was conducted at Mizewa catchment; Fogera, Northwest of Ethiopia, drained by Mizewa River a tributary of the Rib River that feeds to Lake Tana; covering a total area of 2664 ha, located between latitude 11.88°–11.94°N and longitude 37.78°-37.86°E . Chromic-Luvisols,Chromic Vertisols and Leptosols are the most common soil types with basaltic rock formations(Birhanu et.al.,2012). Mixed crop-livestock farming system with maize, rice, �nger millet, te�, groundnut and barley are the principal crops grown in the area (Birhanu et.al.,2012). Urea and DAP are the commonly used chemical fertilizers. Mizewa lie between 1852m and 2360m, dominated by hill to rolling undulating plain land forms and characterized by unimodal rain fall (mean,1204mm) pattern, peaks around August 20th.Mean annual temperature ranges from 16.730C to 19.320C. Flow height (h), surface �ow velocity (Vs) measurement and suspended sediment sampling were conducted at the three monitoring stations three times a day from July8, 2011 through October 16, 2011 to calculate discharge (Q), suspended sediment concentration (SSC), suspended sediment load (SSL), nutrient concentration (NC) and nutrient loss (NL) . Fertilizer maize grain yield (GY) response data were obtained from research results under similar agro-ecological conditions. Surface �ow velocity was measured using a plastic bottle and converted to the average velocity (V) using Gra� methoda (1996). Record of h was converted into �ow cross-sectional area (A) using an empirical relationshipb between h and A . The volume of water (Q) passing a cross section per unit of time was calculated using the area-velocity methodc (Hudson, 1993). Sample runo� sediment was bulked in to one as decade (D) according to the date of sampling starting from July8, 2011 through October 16, 2011 in order to have enough sediment for laboratory analysis and to reduce cost of laboratory. A composite sub-sample of one litter was taken from bulked samples for analysis. In the laboratory, the decade runo� sediment sample was �ltered using Whatman (0.42mm) �lter paper to have SSC . The �ltered water was analyzed for dissolved nitrate and dissolved phosphate. The sediment left on the �lter paper was air dried and weighted to analyze texture, OC, TN, NO -

3, NH +4, and available P concentration. In laboratory, OC was assessed using

Walkley and Black (1934), texture of sediment was determined using hydrometer method following Sahlemedhin and Taye (2000), Jackson (1958) method was used for TN, while Gregorich and Ellert(1993) method was applied for NO -

3 and NH +4 analysis and Olsen et.al.

(1954) procedure was applied for available P. Dissolved nitrate and dissolved Phosphate were determined using spectrophotometer (Bache and Williams, 1971). Load of sediment (SSL), load of dissolved NO 3

-and PO43- load was product of Q(m3/D) and their concentration in mg/l. Sediment

bounded loads of OC, TN, NH 4+, NO 3

- and available P was the product of mg/kg of each species with SSL mass in Kg/D . Seasonal load was the sum of 10 decades of each species. RUSLE model established by Wischmeier and Smith (1978) was applied to identify hot spot areas for sediment sources. Hysteresis loop was developing to assess the temporal e�ect on sediment and discharge interaction over time. Productivity change approach (PCA) technique Bojo (1995) was used to estimate on-site cost of soil erosion. Available NP lost with runo� suspended sediment, NP fertilizers response of maize GY and market price of GY were the basic data sources for evaluating the cost of nutrient loss through runo� soil erosion. E�ect of soil loss on crop yield and its onsite �nancial cost estimation was calculated taking the loss of available NP nutrients and put a value on it using the equivalent estimated net maize grain yield loss. E�ect on crop yield was simply calculated as the net grain yield lost between potential grain yield due to lost available NP on �tting curve and mean grain yield with no NP fertilizers. Statistical comparisons were performed using SPSS 16. Analyses were performed to make comparison with in groups of runo� sediment and nutrient loss between sites and decades. Signi�cance di�erences in sediment load, rate of discharge, and nutrient loss between sites was determined by t-test at 95% con�dence limit. Pearson correlation analysis was done for e�ect analysis among sediment parameters.

Conclusions & Recommendations DuringwaterwhichunderstandduringstudydownstreamTheOC/ha,32GNtransport,measuringsedimentThecharacterizedgreatandtheindicatedthethoughnaturalThisandMZtonutrientconsideredTheansamestudybene�tlettingmeasures,Thereforenationaltoimpactproductivityexpediteprevent

Acknowledgements ThisChallenge

References1.

2.

3.

4.

5.

6.

7.

8.

9.

10.

11.

12.

13.

14.

15.

16.

Conclusions & Recommendations During the monitoring period 18.34x106m3, 12.87x106m3 and 4.92x106m3 of water was lost from MZ 0, MZ 1 and GN 0 watersheds in the form of runo� which has a potential to irrigate a signi�cant hectare of land, so that one could understand the valuable bene�ts gained by farmers if this water was used during dry season through water harvesting technologies though it needs detail study for recommendations, as the o�-site costs from sedimentation and other downstream impacts were not investigated in this paper. The mean SSL of 4 ton/ha, 2 ton/ha and 3 ton/ha in association with ((97 kg OC/ha, 8.4 kg TN/ha, 2.3 kg available N/ha, and 4 kg P2O5/ha), (35 kg OC/ha, 3.1 kg TN/ha, 1.6 kg available N/ha, and 4.1 kg P2O5/ha) and (30.4 kg OC/ha, 2.5 kg TN/ha, 2.3 kg available N/ha, and 4.8 kg P2O5/ha)) from MZ 0, MZ 1 and GN 0 watersheds respectively. The sediment lost did not consider bed-load transport, which might be important in the Mizewa catchments. Hence, measuring bed load in future is important in order to obtain more realistic total sediment and nutrient load calculation. The relationship between discharge and suspended sediment concentration is characterized by clockwise hysteresis for MZ 0 and MZ 1 stations, despite the great di�erences between the decades, i.e. di�erences in terms of pre-decade and decade. Estimated sediment loss (RUSEL, 13.2 ton/ha/year) data regarding the loss of SSL and associated plant nutrients during the monitoring period indicated that the ridges of MZ 1 and GN 0 rivers along with the middle part of the watershed and lower part of MZ 0 were the most critical source of sediment; though it needs further investigation as of the complex interaction of multiple natural and anthropogenic factors. This study conclude that, a reduction in maize gain productivity of 23%, 21% and 26% and equivalent �nancial cost of 200$/ha, 186$/ha and 227$/ha from MZ 0, MZ 1 and GN 0 catchments respectively was estimated. However, in order to obtain a better picture of erosion impacts in the area, studies on other nutrient losses like calcium and magnesium, o�-site e�ects need to be considered. The study also recommend a detail study as of runo� water harvesting also is an opportunity for enhancing rural livelihoods and food security and at the same time minimizes the risk of erosion in the Mizewa watersheds and as the study translates the onsite e�ect of soil erosion into economic terms; this will bene�t the understanding of the problem by land users and/or policy makers, letting them see the need to promote and/or implement soil conservation measures, as that is the language that they usually understand best. Therefore, the researcher recommends considering cost of soil erosion in national economy accounting is important to show signi�cance of soil erosion to policy makers and to land users; even if, more economic and environmental impact analyses at the country level are needed to help set priorities for land productivity issues, to assess the costs and bene�ts of policy decisions, and to expedite identi�cation of the type of investments that will be required to prevent land resources degradation and increase production.

Acknowledgements This paper presents �ndings from IWMI Nile4 project of the CGIAR Challenge Program on Water and Food ; Blue Nile Basin, East Africa, Ethiopia.

References 1. Amanuel Z., 2009. Assessment of spatial and temporal variability of river

discharge, sediment yield and sediment-�xed nutrient export in Geba river catchment, northern Ethiopia, PhD thesis, Department of Earth and Environmental Sciences, K.U. Leuven.

2. Bache,B .W.and Williams, E .G. 1971.Aphosphate sorption index for soils. J. Soil Sci . 22: 289–301.

3. Birhanu Zemadin, Matthew McCartney, and Bharat Sharma, 2012. Establishing Hydrological and Meteorological Monitoring Networks in Jeldu, Diga and Fogera Districts of the Blue Nile Basin, Ethiopia. IWMI .

4. Bojo J., 1995. Land degradation and rehabilitation in Ethiopia, a reassessment. AFTES Working Paper No. 17. The World Bank, Washington DC .

5. CSA, 2011.Central Statistical Agency of Ethiopia Agricultural Sample Survey 2010; Volume I : Report on Area and Production of Crops.

6. Gra� de J. 1996. The Price of Soil Erosion: An Economic Evaluation of Soil Conservation and Watershed Development. PhD Thesis. Agricultural University; Wageningen.

7. Gregorich, E .G. and Ellert , B .H. 1993. Light fraction and macroorganic matter in mineral soils. In M.R . Carter, Ed., Soil Sampling and Methods of Analysis . Canadian Society of Soil Science. Lewis Publishers, Boca Raton, FL, pp. 397–407.

8. Haregeweyn, N., Poesen, J., Deckers,J., Nyssen, J., Haile, M., Govers, G.,Verstraeten, G., Moeyersons, J., 2008a. Sediment-bound nutrient export from micro-dam catchments in Northern Ethiopia. Land degradation and Development, 19: 136-152.

9. Hudson, H. R ., 2003. A case study of approaches for determining di�use suspended sediment sources and processes. In R . F. Hadley, & E . D. Ongley, Sediment and the environment (p.85–94). Wallingford: IAHS (Publication No. 184).

10. LAL, R . 1998. Soil Erosion Research Methods, 2d edition, St. Lucie Press, Delray Beach, FL, 340 pp.

11. Ongley E ., 1996. Water Quality Monitoring - A Practical Guide to the Design and Implementation of Freshwater Quality Studies and Monitoring Programmes. United Nations Environment Program and the World Health Organization.

12. Sahilemedhin Sertsu and Taye Bekele, 2000. Determination of inorganic nitrogen, Procedures for soil and plant analysis. Ethiopian Agricultural Research Organization.Addis Ababa, Ethiopia.

13. Seeger,M., Errea, M.P., Begueria, S.,Marti , C., Garzia-Ruiz, J.M., Arnaez, J., 2004. Catchment soil moisture and rainfall characteristics as determinant factors for discharge/ suspended sediment hysteretic loops in a small headwater catchment in the Spanish Pyrenees. Journal of Hydrology 288, 299–311.

14. Tamene.L .,S .J.Park;R .Dikau;P.L .G.Vlek .2006.Reservoir siltation in Ethiopia: Determinates, source areas, and management options. In UNESCO -Chair in water resource proceeding of international sediment initiatives conference, 12-15 Nov.2006,Khartoum, Sudan.

15. Tilahun Tadesse, Minale Liben, Alemayehu Assefa and Abreham Marie, 2007; Maize fertilizer response at the major maize growing areas of northwest Ethiopia.

16. Walling D.E ., 1977. Limitation of the rating curve technique for estimating suspended sediment loads, with particular reference to British rivers; International Association of Hydrological Sciences Wallingford; 3448.

Nile

a V= 0.6*Vs, b A=5.5h2+3.6h for MZ 0 , A=2.9h2+0.95h+0.05 for MZ 1 and A=9.95h2+7.44h for GN 0, c Q=V*A

d GY =-0.29N2+58.6N+2537.3(R2=0.75) and GY= -0.55(P2

O5

)2+82.25P2

O5

+2690.7 (R2=0.88) regression equations between GY of maize to N and P2

O5

.

This document is licensed for use under a Creative Commons Attribution-Noncommercial-Share Alike 3.0 UnportedLicense July 2013