Characterization of Erodibility Using Soil Strength and Stress-Strain Indices for Soils in Some Selected Sites in Imo State

Research Journal of Environmental and Earth Sciences 4(7): 688-696, 212 ISSN: 241 492 Maxwell Scientific Organization, 212 Submitted: April 29, 212 Accepted: June 8, 212 Published: July, 212 Characterization of Erodibility Using Soil Strength and Stress-Strain Indices for Soils in Some Selected Sites in Imo State C.C. Egwuonwu and N.A.A. Okereke Department of Agricultural Engineering, Federal University of Technology, Owerri, Nigeria Abstract: In this study, initial soil strength indices (q u ) and stress-strain characteristics namely failure strain (ε f ); area under the stress-strain curve up to failure (I s ) and stress-strain modulus between no load and failure (E s ) were investigated as potential indicators for characterizing the erosion resistance of two compacted soils namely Sandy Clay Loam (SCL) and Clay Loam (CL) in some selected sites in Imo State, Nigeria. The unconfined compressive strength (used in obtaining strength indices) and stress-strain measurements were obtained as a function of moisture content in percentage (mc %) and dry density (γ d ). Test were conducted over a range of 8 to 3% moisture content and 1. to 2. g/cm 3 dry density at applied loads of 2,, 8, 16 and 32 kpa. Based on the results, it was found out that initial soil strength alone was not a good indicator of erosion resistance. For instance in the comparison of exponents of mc % and γ d for jet index or erosion resistance index (J i ) and the strength measurements, q u and E s (Table 1) agree in signs for mc %, but are opposite in signs for γ d. Therefore there is an inconsistency in exponents making it difficult to develop a relationship between the strength parameters and J i for this data set. In contrast, the exponents of mc % and γ d for J i and ε f and I s are opposite in signs (Table 1), there is potential for an inverse relationship. The measured stress-strain characteristics however, appeared to have potential in providing useful information on erosion resistance. The models developed for the prediction of the extent or the susceptibility of soils to erosion and subjected to sensitivity test on some selected sites as shown in Table 2 achieved over 9% efficiency in their functions. Keywords: Characterization, erodibility, Imo state, indices, Nigeria, selected sites, soil strength, soils, stress-strain INTRODUCTION Soil erosion is one of the most important physical and socio-economic problems affecting our development in this part of the world today. Apart from the fact that it constitutes a menace to the environment and despite its destruction of our infrastructure, highways etc and soil erosion creates a major problem on our agricultural land thereby interfering seriously with mass food production hence the need for predicting the susceptibility of soils to erosion. An accurate estimate of soil erodibility (the susceptibility or vulnerability of soil to erosion) is important to engineers involved in the design of water management projects. Defining soil erodibility, however, is a difficult task. This is because soil detachment is a complex function of both soil and eroding fluid properties (Egwuonwu and Uzoije, 29; Akintola, 21). Predicting the susceptibility of a soil to erosion prior to a concentrated flow event is an important problem in many engineering projects such as irrigation, channels, levees, highways, railways, spillways, construction sites, mining area reclamation and land management:. In many of these engineering projects not only is the erodibility of the soil at the surface of interest, but also the erodibility with depth (Hanson, 1996). Earth spillways are an example in which erosion exposes different materials with depth and the erodibility of these materials is important in the performance of the spillways under concentrated flow conditions (Hanson, 1996). Measuring erodibility directly is often difficult. Therefore development of measurable soil parameters that indicate erodibility is attractive (Elliot et al., 199). Soil strength indices have been a common soil parameter investigated for characterization of soil erodibility. Strength indices are commonly used because they can be easily and rapidly obtained, measured incrementally with depth and related to other soil parameters that affect soil erodibility such as dry density (γ d ) and water content (mc %). Soil strength indices are also appealing because soil erosion is a function of the forces applied by the flow and resistance to erosion is offered by bonding forces between soil particles and other materials in the soil matrix (Hanson, 1996). Since the state of bonding is manifested in the soil strength, resistance to erosion should be Corresponding Author: C.C. Egwuonwu, Department of Agricultural Engineering, Federal University of Technology, Owerri, Nigeria 688

Res. J. Environ. Earth Sci., 4(7): 688-696, 212 characterized by the measurement of soil strength. Although soil strength indices have been used to characterize erodibility of soils, their success have been limited (Elliot et al., 199; Parker et al., 1995) hence the need for the use of more quantifiable parameters of soil such as strength and stress-strain characteristics. Unconfined compressive strength, triaxial shear test, vane shear strength, pocket penetrometer as well as other tests have been used to characterize soil strength (Nearing and West, 1988). The objectives of the study are: To measure and observe the relationship between soil strength indices, stress-strain characteristics and changes in moisture content and dry density for the selected sites and to develop models for predicting the susceptibility or extent of soils to erosion in the selected sites. MATERIALS AND METHODS Study area: Characterization of erodibility using soil strength and stress-strain indices was studied for some selected towns in Imo State Nigeria namely: Ideato, Mbaise and Okigwe. These towns lie between latitude 5.3 to 6.5 o N and longitude 6.8 to 8.1 o E and are within the tropical rainforest zone. The study was carried out in February 211. Soil materials: The soils used in this study were a Sandy Clay Loam (SCL) and a Clay Loam (CL) obtained from three (3) sites in Imo State, Nigeria namely Imo: Okigwe; Imo Mbaise; Imo Ideato The influence of compaction (mc % and γ d ) on resistance to erosion of these soils was reported in Hanson and Robinson (1993) and Hanson (1992, 1993). Soil testing: Test was conducted on the compacted samples. Strength and stress-strain values were measured with the unconfined compression test. The unconfined compressive tests were conducted in accordance with ASTM D2166. In addition to the unconfined compressive strength (q u ), stress-strain characteristics namely failure strain (ε f ), area under the stress-strain curve up to failure (I s ) and stress-strain modulus between no load and failure (E s ) were determined for the unconfined compressive test. Unconfined compressive tests were continued until the load values decreased with increasing strain or until 2% strain was reached. The samples of soil (Sandy Clay Loam (SCL) and Clay Loam (CL)) were tested for resistance to erodibility using a submerged jet testing device. Water was fed under a constant head of.91 m, through a circular nozzle, 13 mm in diameter, at a set height of.22 m above a level bed of prepared soil. The rate of scour was monitored with time. The erosion resistance of the samples was based on a jet index (J i ). The lower the J i values the less erosion occurred and the more resistant the material (Hanson, 1991). Data analysis: A comparison of nonlinear power curve regression fits or power equation or models for jet index (J i ), failure strain (ε f ), area under the stress-strain curve up to failure (I s ), unconfined compressive strength (q u ) and stress-strain modulus between no load and failure (E s ) versus the variables mc % and γ d was developed. The non linear power curve function served as potential for functional relationships. Multiple Linear Regressions in conjunction with Naïve-Gauss Elimination Method and Gauss-Jordan Matrix Inverses were used to obtain the models. The coefficient of determination (r 2 ) was computed for each developed model to ascertain the level of accuracy or exactness of each model. Soil preparation: Soil samples were prepared by static load compaction. Thirty test samples of each soil were compacted using a device similar to a fixed ring consolidometer. Static pressure was applied to the soil by pneumatic bellows. Loads of 2,, 8, 16 and 32 kpa were applied to the samples. The size of prepared soil sample was 445 mm in diameter and the final height varied from to 2 mm depending on the load applied and the mc %. Samples were compacted at various mc % ranging from 8 to 3% for each load. The soils were wet by spraying to various mc % mixed thoroughly and stored in polyethylene bags for 24 h to allow the water to permeate through the soil. 689 RESULTS The strength test and stress-strain results as well as the jet index result for SCL and CL soils are shown in Table 3-8 for the sites selected in Imo State. The comparison of nonlinear power curve regression fits for J i, ε f, I s, q u and E s versus the variables mc % and γ d is shown in Table 1. The prediction of the extent or the susceptibility of soils to erosion on some selected sites using the J i models developed is illustrated in Table 2. The predicted J i versus q u and J i versus ε f for SCL and CL soils (Imo) in Fig. 1 to 4 and Fig. 5 to 8, respectively.

Res. J. Environ. Earth Sci., 4(7): 688-696, 212 Fig. 1: Predicted Ji Vs ε F for scl soil, Imo (1) SCL SOIL IMO (1)..16.14.12.1.8.6.4.2.2.4.6.8.1.12.14.16. ε f.14 CL SOIL IMO (2).12.1.8.6.4.2.2.4.6.8.1.12 ε f Fig. 2: VS ε F for cl soil, Imo (2). SCL SOIL IMO (3).15.1.5.1.2.3.4.5.6.7 ε f Fig. 3: Vs ε F for scl soil, Imo (3) 69

Res. J. Environ. Earth Sci., 4(7): 688-696, 212. SCL SOIL IMO (1) CL SOIL IMO (2) SCL SOIL IMO (3).15.1.5.2.4.6.8.1.12 ε f Fig. 4: Combined data for Scl and Cl soils for predicted J i Vs ε F, Imo state (3) 45 SCL SOIL IMO (1) 35 3 2 15 1 5.2.4.6.8.1.12.14.16. q (KPa) u Fig. 5: VS q u for scl soil, Imo (1) 12 CL SOIL IMO (2) 8 6 2.2.4.6.8.1.12 q (KPa) u Fig. 6: Vs Q u for cl soil, Imo (2) 691

Res. J. Environ. Earth Sci., 4(7): 688-696, 212 6 CL SOIL IMO (3) 5 3 2 1 Fig. 7: Vs Q u for scl soil, Imo (3).1.2.3.4.5.6 q (KPa) u.7 12 SCL SOIL IMO (1) CL SOIL IMO (2) SCL SOIL IMO (3) 8 6 2.2.4.6 q (KPa) u.8.1.12 Fig. 8: Combined data for scl and cl soil for predicted J i Vs Q u, Imo state Table 1: Power curve functions of parameters J i, q u, ε f, I s and E s for SCL and CL soils Imo state Imo (1) SCL soil J i = 3.6* 1 3 (mc) -4.3 (γ d ) -3.7 r 2 =.98 ε f = 9.68* 1-6 (mc) 2.67 (γ d ) 3.1 r 2 =.95 I s = 2. * 1-2 (mc).28 (γ d ) 8.23 r 2 =.93 q u = 27. (mc) -1.96 (γ d ) 8.58 r 2 =.94 E s = 1.26 * 1 5 (mc) -2.44 (γ d ) 2.55 r 2 =.88 Imo (2) CL soil J i = 39.24 (cm) -2.48 (γ d ) -4.11 r 2 =.96 ε f = 1.21 * 1-4 (mc) 1.5 (γ d ) 4.23 r 2 =.95 I s = 8.83 * 1-3 (mc).55 (γ d ) 9.27 r 2 =.91 q u = 5.66* 1 2 (mc) -1.94 (γ d ) 7.63 r 2 =.92 E s = 4.6 * 1 5 (mc) -2.76 (γ d ) -4.99 r 2 =.85 Imo (3) SCL soil J i = 22.94 (mc) -2.57 (γ d ) -1.85 r 2 =.93 ε f = 5.36 * 1-6 (mc) 2.95 (γ d ) 2.95 r 2 =.96 I s = 3.35 * 1-4 (mc) 1.69 (γ d ) 8.53 r 2 =.93 q u = 13.26 (mc) -1.8 (γ d ) 7.87 r 2 =.96 E s = 1.84 * 1 5 (mc) -2.8 (γ d ) 3.65 r 2 =.81 Table 2: Prediction of the extent or the susceptibility of soils to erosion on some selected sites using the J i model developed for Imo state Location Average mc % Average (γ d ) J i model Imo (1) A Imo (1) B 13.9 1. 1.87.22.4 Imo (2) A 14. 1.3. Imo (2) B Imo (3) A Imo (3) B. 11.1 14.5 1.73 1.6 1.51 DISCUSSION.3.11 The strength indices (q u ) for the SCL and CL soils for the sites selected followed similar trends (Table 3 to 5). Strength indices increased as the γ d increased for a given mc and as the mc decreased for a given γ d (the asterixed values * and ** in Table 3 to 5). It was observed that q u, ε f, I s and E s increased as γ d increased. 692

Res. J. Environ. Earth Sci., 4(7): 688-696, 212 The values of q u and E s decreased as mc % increased for a constant γ d, whereas ε f and I s tended to increase (the asterixed values * and ** in Table 3 to 5).. At low mc %, the soil tended to fail by brittle fracture. At high mc %, the soil failed plastically. The relationship of ε f, I s and E s to γ d and mc % for the same set of samples shown in the tables listed for strength indices are shown in Table 3 to 5. The E s values were observed to have similar relationship as the strength indices (q u ) had to γ d (Table 3 to 5). Soil strength indices decreased with increases in mc% (the asterixed values * and ** in Table 3 to 5), whereas the resistance to erosion increased with increases in mc % within the same range ((the asterixed values ** and *** in Table 6 to 8). The conclusion is that the soil strength indices alone, although affected by mc % and γ d, would make poor erosion characterization indicators. The stress-strain characteristics ε f and I s, tended to increase with increases in γ d and mc % (Table 3 to 5) and these concur with that of the erosion resistance index (J i ) (Table 6 to 8). Whereas E s tended to increase with γ d and decrease with mc %. Based on these test results, E s has the same potential problems as strength indices, q u. However, ε f and I s have potential for indicating erosion resistance of a soil. Table 4: Strength and stress-strain indices for CL soil as a function of mc % and γ d for Imo state (2) mc(%) γ d (g/cm 3 ) q u (kpa) ε f I s (kpa) E s (kpa) 14.5 15.6 17.5** 2. 2.9 21.6 1.8 13.5 15.4 17.6** 2.7 21.4 9.5 11.8 14.2 16.3 17.5** 2. 21.9 9. 1.8 11.8 16.5 19.4 2.4 11.8 15.2 17.6**..5 1.9 1.1 1.14 1.26 1.28 1.3 1.1 1. 1.2 1.24 1.37 1.53 1.17 1.21 1.26 1.33 1. 1.55 1.6* 1.33 1.38 1.5 1.6* 1.65 1.47 1.61* 1.64 1.67 1.72 1 9 1 12 1 7 1 15 15 35 15 2 38 45 3 3 7 7 8 7 11 12.15..38..41.6.8.6..7.32.8.1..26.34.8.12.14.38.6.8..1.1..6..8..32 1. 4..14.28.38.8 4. 6..28.32. 1.2 4. 8..7 3. 4. 6. 8. 6 3 75 7 8 75 13 13 1 8 3 19 2 22 3 3 22 2 Table 3: Strength and stress-strain indices for SCL soil as a function of mc % and γ d for Imo state (1) mc (%) γ d (g/cm 3 ) q u (kpa) ε f I s (kpa) E s (kpa) 8.** 12.6 13.7 15.4 15.7.2 19.2 9.1 11.7 14.9 17. 17.9.7 8.1** 11.6 13.8 15. 15.6 17. 8.1** 9.5 1.7 12.3 14.2 8.1** 1. 11.2 12.2 13.7 1.17 1.26 1.27 1.3 1.39 1.58* 1.67 1.29 1.31 1.49 1.58* 1.67 1.31 1.6 1.69 1.78 1.82 1.41 1.44 1.58* 1.62 1.83 1.58* 1.64 1.81 2 7 5 7 12 8 1 1 2 38 6 7 55 48 8 6 7 8.7..12.8...7.16..15.5...12.8.8.13....8.8.22.3. 2.5 4...8 1.8 3. 4. 6.. 1.8 3. 4. 6..35..8 1.5 5.5.5 1. 2. 3. 6. 2 38 3 6 8 8 35 35 12 12 8 17 29 26 23 12 693 Table 5: Strength and stress-strain indices for SCL soil as a function of mc % and γ d for Imo state (3) mc(%) γ d (g/cm 3 ) q u (kpa) ε f I s (kpa) E s (kpa) 9. 12.** 14. 17..5 2. 8. 12.** 13.5 15.6 16.5. 21. 8.4 1. 12.1** 14.2 15. 17. 9. 1. 11.6 15. 17.5 8.1 1. 12.** 13.5 15. 1.15 1.2 1.28 1.6* 1.65 1.3 1.5 1.6* 1.67 1.7 1.75 1.6* 1.7 1.78 1.83 1. 1.45 1.6* 1.72 1.83 1.5 1.65 1.7 1.8 3 6 1 12 2 22 12 13 2 28 33 38 32 17 2 48 6 6 22 28 45 45 6 7 7 78 9.13.16.28.53.15..33.8.13.15.7.28.76.11.94.12.15.35.7.12.1.17..37.55 2.8 4.5.31.75 2.2 3.2 4.8 6..24.34 1.2 2.96 4.8 6..3.39 1. 3. 6.. 1. 2.2 4. 6. 3 51 4 8 7 1 7-48 3 1 8 3 1 16 8 3 23 26 2 145 8

Table 6: Values of J i for SCL soil as a function of mc % and γ d for Imo state (1) mc (%) γ d (g/cm 3 ) J i 13. 12.6** 12.1 11.1 1. 13.3 1.9 14.1 13.7 12.6** 1.9 12.5** 15.9 13.9 12.6** 12. 16.3 16.2 13. 16.2 15. 13.7 14.6 17.4 14.3 1.29*** 1. 1.48 1.69 1.36 1.86 1.29*** 1. 1.61 1.84 1.28*** 1. 1.6 1.8 1.9 1.43 1.55 1.65 1.9 1.73 1.8 1.87 1.83 1.86 1.9 Res. J. Environ. Earth Sci., 4(7): 688-696, 212..17.16.16.16.16.15.8.8.8.6.6.6.5.4.4 Table 7: Values for J i for CL soil as a function of mc % and γ d for Imo state (2) mc (%) γ d (g/cm 3 ) J i 17.3 15. 13.6 13. 12.4 1.6 17.** 15.9 13.6 13.1 12.4 11. 19.1 15.5 15. 14.3 12.6 2.5 17.1** 19.6 16.6 13.8 2.5.3 16.9 15..4 19.8.7 17.** 14.5 1.21 1.3 1.51*** 1.63 1.75 1.26 1.32 1. 1.5*** 1.64 1.74 1.34 1.5*** 1.6 1.47 1.5*** 1.6 1.73 1.58 1.6 1.6 1.68 1.64 1.7 1.74 1.74.8.8.8.8.8.7.7.6.6.6.55.58.58.5.5... 694 Table 8: Values for J i for SCL soil as a function of mc % and γ d for Imo state (3) mc (%) γ d (g/cm 3 ) J i 12.4 11.6 1.8 12.8 11.5 1.7 13.8 12.5** 11.6 14.1 13. 11. 12.9 16.1 14.2 13. 12.4** 14.7 17.5 15.8 12.4** 15.8 17.5 14.3 16.5 13.8 1.38 1.56*** 1.58 1.81 1.32 1.6 1.75 1.32 1.56*** 1. 1.33 1.56*** 1.76 1.86 1.53 1.46 1.56*** 1.94 1.72 1.68 1.85 1.84 1.94.21.21.21.17.17.17.16.16.16.14.11.9.9.7.6.6.4.4 A comparison of nonlinear power curve regression fits for J i, ε f, I s, q u and E s versus the variables mc% and γ d are shown in Table 1. The nonlinear power curve function served as potential for functional relationships and was consistently a good fitting function in all cases and allowed comparison between parameters. The exponents of mc % and γ d indicate their functional relationship to J i, ε f, I s, q u and E s. Based on the exponents J i decreases as mc % increases holding γ d constant. This comparison indicates that even though the exponents of mc % and γ d for J i and ε f and I s are opposite in signs, there is potential for an inverse relationship. In contrast, the comparison of the exponents for mc % and γ d for J i and the strength measurements, q u and E s agree in signs for mc %, but are opposite in signs for γ d. Therefore, there is an inconsistency in exponents making it difficult to develop a relationship between the strength parameters and J i for this data set. The jet index (J i ) test results and the strength (q u ) and stress-strain (ε f ) test results are independent data sets, a prediction of J i can be made for the strength and stress-stain data using the functional relationships in Table 1 for the SCL and CL soil. As an example of the potential relationships for J i versus ε f and q u a comparison of the predicted J i versus ε f and predicted J i versus q u for Imo State is shown in Fig. 1 to 4 and Fig. 5 to 8, respectively for SCL and CL soil. These plots show that there does appear to be a general trend as q u increases J i decreases (Fig. 5 to 7),

but there is quite a bit of scatter. For a parameter to be effective in predicting erodibility, it must also be consistent between soils. Combing the data in Fig. 1 to 3 for predicted J i versus ε f brings the data for SCL and CL soils together (Fig. 4). However, combining the data in Fig. 5 to 7 for q u versus J i does not bring the data together (Fig. 8). Using the erosion prediction model J i obtained for the three sites in Imo State, two locations in each of the sites were each tested for the mc % and γ d. The two sites included one that had visibly undergone erosion (A) and one that had not visibly undergone or partially undergone erosion (B). The parameters mc % and γ d were obtained at random locations (2) on the site. The average values of the parameter were imputed in the J i model to ascertain the extent of or the susceptibility of soils to erosion and to also verify the efficacy or exactness of the model obtained. This is shown in Table 2. Based on the jet index (J i ), a highly erodible soil will have a J i of approximately whilst an erosion resistant soil will have a J i of approximately.2. From Table 2; Imo 1A, Imo 2A and Imo 3A had J i values close to or slightly above it. This explains the adverse extent of erosion on these soil to the extent it was glaringly visible. Imo 1B and Imo 2B had J i values close to.2 with no visible signs of erosion. These soils are still erosion resistant. However, with a J i value of.11 for Imo 3B, is an indication of the susceptibility of the soil to erosion even though it is not yet prominent. However, if nothing is done to curb the erosive tendencies at this stage of Imo 3B will subject the soil to severe erosion. CONCLUSION AND RECOMMENDATIONS The soil strength indices (q u ) for SCL and CL soil in the sites increased with increases in dry density (γ d ) (Table 3 to 5). The resistance to erosion (J i ) of the SCL and CL soils in the region also increased with increases in γ d (Table 6 to 8). q u decreased with increases in moisture content (mc %) (Table 3 to 5), whereas J i increased with increases in mc % within the same range (Table 6 to 8). Soil strength (q u ) and erodibility (J i ) are both affected by mc % and γ d, but not necessarily in the same functional manner (Table 1). The comparison of the exponents for mc % and γ d for J i and the strength measurements; q u and E s agree in signs for mc %, but are opposite in signs for γ d. Therefore, there is an inconsistency in exponents making it difficult to develop a relationship between the strength parameters and J i for the data set. Res. J. Environ. Earth Sci., 4(7): 688-696, 212 695 Furthermore, the predicted J i versus q u for SCL and CL (Fig. 5 to 7) soils does not indicate potential for a relationship between soils. For a parameter to be effective in predicting erodibility it must also be consistent between soils. However, the combined data in Fig. 5 to 7, for J i and q u does not bring the data together (Fig. 8). This does not discard soil strength as a potential indicator, but it does indicate that it should not be solely relied upon for prediction of soil erodibility. The results from compacted samples of SCL and CL soils indicate that stress-strain characteristics namely failure strain (ε f ) and area under the stress strain curve up to failure (I s ) are helpful in predicting erodibility. The stress-strain indices (ε f and I s ) increased with increases in both γ d and mc% (Table 3 to 5). The resistance to erosion (J i ) of the SCL and CL soil also increased with increase in both γ d and mc % within the same range (Table 6 to 8). ε f and I s had similar functional relationships to J i (Table 1). Even though, the exponents of mc % and γ d for J i and ε f and I s are opposite in signs, there is potential for an inverse relationship. The predicted J i versus ε f for SCL and CL soils (Fig. 1 to 3) indicate potential for a relationship between stress-strain and erodibility for a soil and between soils. The combined data in Fig. 1 to 3 for predicted J i versus ε f brings the data for SCL and CL soils together (Fig. 4). These results indicate that stress-strain characteristics should be included alongside strength indices in studies when developing relationships between soil erodibility and other soil parameters. The models developed for the prediction of the extent or the susceptibility of soils to erosion and subjected to sensitivity test on some selected sites as shown in Table 2 achieved over 9% efficiency in their functions. From the discoveries made in this project, it is appropriate to recommend the following: Similar studies should be extended to other areas in the country that are prone to erosion in order to develop models that will ascertain the susceptibility or extent of soils to erosion. Since the control of erosion is capital intensive, early detection will minimize cost and ultimately save the environment from its menace. Studies in soil erosion are usually intensive and expensive in terms of finance. Government through institutions, agencies and parastatals should make available adequate funds to sustain such studies. With proper funding results of such studies will go a long way in totally annihilating the menace of soil erosion in our environment if sincerely implemented?

Res. J. Environ. Earth Sci., 4(7): 688-696, 212 Soil strength should not be solely relied upon for measuring erodibility of soils. However, other parameters i.e., stress-strain characteristics etc should be included alongside strength indices when developing relationships between soil erodibility and other soil parameters. REFERENCES Akintola, J.O., 21. Determination of rainfall erosivity for different agro-ecological zones in Nigeria. Unpublished M.Sc. Thesis, Department of Agric. Engrg, University of Ibadan. Egwuonwu, C.C. and A.P. Uzoije, 29. A Comparative Analysis of Coconut, Palm Frond and Palm Stem Fibres as Erosion Control Materials on Embankments. Asset International Journal, University of Agriculture Abeokuta, Nigeria. Elliot, W.J., L.J. Olivieri, J.M. Laflen and K.D. Kohl, 199. Predicting Soil Erodibility from Strength Measurements. ASAE Paper No. 9-29, St. Joseph, Mich. ASAE. Hanson, G.J., 1991. Development of a jet index to characterize erosion resistance of soils in earthen spillways. Trans. ASAE, 34(5): 215-22. Hanson, G.J., 1992. Erosion resistance of compacted soils. TRB Transport Res., 1369: 26-3. Hanson, G.J., 1993. Effects of Consolidation on Soil Erodibility. ASAE Paper No St. Joseph, Mich., pp: 93-291. Hanson, G.J., 1996. Investigating soil strength and stress-strain indices to characterize erodibility. Trans. ASAE, 39(3): 883-89. Hanson, G.J. and K.R. Robinson, 1993. The influence of soil moisture and compaction on spillway erosion. Trans. ASAE, 36(5): 1349-1352. Nearing, M.A. and L.T. West, 1988. Soil strength indices as indicators of consolidation. Trans. ASAE, 31(2): 471-475. Parker, D.B., T.G. Michel and J.L. Smith, 1995. Compaction and water velocity effects on soil erosion in shallow flow. J. Irrigat. Drainage Engrg. ASCE, 121(2): 17-178. 696