Bioresource Technology 98 (2007) 912–917

Grass mulching effect on infiltration, surface runoffand soil loss of three agricultural soils in Nigeria

K.O. Adekalu a,*, I.A. Olorunfemi b, J.A. Osunbitan a

a Department of Agricultural Engineering, Obafemi Awolowo University, Ile-Ife, Nigeriab Department of Agricultural Engineering, Lagos State Polytechnic, Lagos, Nigeria

Received 20 October 2004; received in revised form 24 February 2006; accepted 25 February 2006Available online 5 May 2006

Abstract

Mulching the soil surface with a layer of plant residue is an effective method of conserving water and soil because it reduces surfacerunoff, increases infiltration of water into the soil and retard soil erosion. The effectiveness of using elephant grass (Pennisetum purpur-

eum) as mulching material was evaluated in the laboratory using a rainfall simulator set at rainfall intensities typical of the tropics. Sixsoil samples, two from each of the three major soil series representing the main agricultural soils in South Western Nigeria were collected,placed on three different slopes, and mulched with different rates of the grass. The surface runoff, soil loss, and apparent cumulative infil-tration were then measured under each condition.

The results with elephant grass compared favorably with results from previous experiments using rice straw. Runoff and soil lossdecreased with the amount of mulch used and increased with slope. Surface runoff, infiltration and soil loss had high correlations(R = 0.90, 0.89, and 0.86, respectively) with slope and mulch cover using surface response analysis. The mean surface runoff was corre-lated negatively with sand content, while mean soil loss was correlated positively with colloidal content (clay and organic matter) of thesoil. Infiltration was increased and soil loss was reduced greatly with the highest cover. Mulching the soils with elephant grass residuemay benefit late cropping (second cropping) by increasing stored soil water for use during dry weather and help to reduce erosion onsloping land.� 2006 Elsevier Ltd. All rights reserved.

Keywords: Surface runoff; Mulching; Rainfall simulator; Infiltration

1. Introduction

The rate of infiltration of water into the soil depends onthe intensity and duration of the rainfall, slope of the field,nature of the soil surfaces and physical characteristics ofthe soil. Mulching or covering the soil surface with a layerof plant residue is an effective method of conserving water,because it reduces surface runoff and increases infiltrationof water into the soil (Ghawi and Battikhi, 1986). Mulchalso reduces the depletion of water within the root zonebecause it suppresses evaporation. In addition, mulchingdecreases crusting of the soil due to rainfall impact, which

0960-8524/$ - see front matter � 2006 Elsevier Ltd. All rights reserved.

doi:10.1016/j.biortech.2006.02.044

* Corresponding author. Tel.: +234 8037250823.E-mail address: koadekalu2002@yahoo.co.uk (K.O. Adekalu).

reduces erosion by absorbing the kinetic energy of the rain-drops (Schwab et al., 1993; Lal, 1979).

Adams (1966) found that mulching with rice straw sig-nificantly increased the infiltration of clay pan soils onsloping land. Bernett et al. (1967) observed a runoff of17% of rainfall and 3.4 tons/ha of soil loss for rice strawmulched plots compared to 38% and 20.2 tons/ha for anunmulched plot. Lattanzi et al. (1974) showed that interillerosion was reduced by approximately 40% when wheatstraw mulch was applied at a rate of 6 tons/ha and by anestimated 80% at a rate of 9.2 tons/ha. McCalla et al.(1963) found rice straw mulch cover to be more effectiveat increasing infiltration than incorporation of organicmatter. Mannering and Meyer (1961) found that soil lossfrom plots covered with shredded corn stalks was about

Fig. 1. Map of Nigeria showing the study area.

Table 2Some selected properties of the elephant grass mulch cover material (drymatter basis)

Parameters Values

Total C (g kg�1) 420Total N (g kg�1) 16.5C/N ratio 25Average length (cm) 75Average width (cm) 4

K.O. Adekalu et al. / Bioresource Technology 98 (2007) 912–917 913

one-half of unmulched plots. Meyer et al. (1970) reportedthat 0.5 tons/ha of rice straw mulch can reduce soil loss byone-third of that with no mulch cover, and at 5 tons/hacould reduce soil loss by 95%. Khan et al. (1988) foundmulching with rice straw to be more effective than usingcover crop. Lal (1979) found mulching tilled soil with 4–6 tons/ha of rice straw to be effective in reducing soil lossand runoff on slopes ranging from 1% to 15% and that theeffectiveness of no-tillage in preventing runoff and erosionwas comparable to applying 4–6 tons/ha of rice straw mulch.

Mulching is highly beneficial in the tropics with high rain-fall intensity especially on sloping ground because it reducesrunoff and soil loss. Mulching materials in the tropics includeleafy crop residues such as sugarcane trash (Sacchariumesculenta) or banana (Musa parasidica) leaves and easilygrown bulky fodder crops such as elephant or guinea grass.

In Southern Nigeria, elephant grass (Pennisetum purpur-

eum) is a potential mulching material because it is easilyobtained and grows on fallowed land and has similar mor-phology with rice straw. The objective of this research wasto quantify the effect of mulch cover of elephant grass onthe surface runoff, infiltration, and soil loss of three majoragricultural soils in Southern Nigeria under three fieldslopes common in the area.

2. Methods

Two samples from each of the three most common agri-cultural soils in Nigeria were collected from the upper15 cm of uncultivated land from three locations in theSouth Western Nigeria (Fig. 1). The two samples for eachsoil type were bulked to give a composite. All soils hadbeen under bush fallow for five to ten years prior to beingsampled. The physico-chemical properties of the soils aresummarized in Table 1. The iron oxide in the soil wasremoved by centrifuging with citrate-bicarbonate, per-oxide, sodium dithionite and saturated sodium chlorideuntil a clear centrifugate was obtained prior to analysis(Gee and Bauder, 1986). Some selected properties of thegrass used for the mulching are shown in Table 2.

The soils were sun-dried to moisture content of 7% (drybasis), and large clods were reduced to smaller fractions.The soils were then passed through a No. 10 sieve, withan apparent opening size of 2 mm. A portion of the soilswas further dried for mechanical analysis, while theremaining portions were kept in sealed polythene bags.

A rainfall simulator similar to the one described byMeyer and McCune (1958) was used in the laboratory to

Table 1Particle size distribution and initial organic matter content of the Nigerian so

Soil series Location Plastic limit Liquid limit Organic mattercontent (%)

Par

San

Apomu Ilorin 14.2 20.5 2.16 67.6Jago Epe 13.8 19.8 1.61 73.7Iwo Ile-Ife 9.5 15.8 1.56 80.4

simulate 100 mm/h rainfall intensity, which is typical ofthe tropics (Lal, 1984). The experiment used a 3 · 3 · 3 ·3 · 4 factorial design with four independent variables(three of three levels each, and the other of four levels).The independent variables and their levels were:

slope; 6%, 9% and 12%,rainfall duration; 30, 60 and 90 min,soil types; Iwo, Apomu and Jago,percent ground cover by mulch; 0%, 30%, 60% and 90%.

The zero percent mulch cover served as the controlexperiment. This gave 108 experimental runs and therewere four replicates of each run. The dependent variableswere surface runoff, soil loss, and apparent cumulativeinfiltration. The amount of grass mulch needed for a givenmulch cover was estimated using the following equationderived by Gregory (1982):

ils

ticle size distribution Textural classification Taxonomy

d (%) Silt (%) Clay (%)

15.8 17.1 Sandy loam Inceptisol4.2 22.1 Sandy clay loam Entisols4.7 14.9 Loamy sand Alfisols

Table 3Analysis of variance for the runoff and soil loss data generated in therainfall simulator

Sources DF F-value

Soil loss Runoff

Soil 2 11.2* 11.3*

Mulch cover 3 131.3** 130.6**

Slope 2 59.7** 58.5**

Duration 2 42.8** 46.7**

Soil * slope 4 10.6* 11.8*

Mulch * slope 6 12.4* 10.5*

Soil * mulch 6 11.8* 10.3*

Soil * duration 4 2.5ns 1.8ns

Duration * mulch 6 1.9ns 1.7ns

Duration * slope 4 2.3ns 2.1ns

Soil * slope * mulch 12 10.5* 11.2*

Soil * mulch * duration 12 1.6ns 2.2ns

Soil * slope * duration 8 2.1ns 2.5ns

Slope * mulch * duration 8 2.7ns 3.4ns

Slope * mulch * duration * soil 24 0.8ns 1.2ns

ns—not significant.* Significant at 5%.

** Significant at 1%.

Table 4Surface runoff (%) from three Nigerian soils as affected by elephant mulchcover and slope (average for all duration)

Soil series Slope (%) Mulch cover (%)

0 30 60 90

Iwo 6 61 (4)a 44 (3) 29 (2) 20 (1)9 73 (3) 55 (3) 37 (2) 26 (1)

12 82 (4) 63 (2) 46 (2) 34 (1)

Apomu 6 82 (4) 56 (3) 32 (1) 19 (1)9 84 (4) 60 (3) 44 (2) 26 (1)

12 90 (4) 68 (3) 53 (2) 44 (1)

Jago 6 79 (4) 56 (3) 35 (2) 21 (1)9 82 (4) 61 (3) 44 (1) 25 (1)

12 87 (4) 65 (3) 47 (1) 29 (1)

a Standard deviation values are in brackets.

914 K.O. Adekalu et al. / Bioresource Technology 98 (2007) 912–917

MR ¼ � lnð1�MCÞ=Am ð1Þwhere MR is the mulch rate in tons/ha, MC is the fractionof ground cover by mulch, and Am is the area covered perunit mass of the mulch type. The Am value of 0.38 derivedby Ozara (1992) for grass was used. This gave mulch appli-cation rate of 0.9, 2.4, 6.1 tons/ha for 30%, 60%, and 90%mulch cover, respectively.

Boxes measuring 1.0 m · 0.8 m · 0.3 m (length ·width · height) were used to expose the soils to the simu-lated rain. Holes 2 mm in diameter were drilled at 10 mmapart in the bottom of the box to allow free drainage ofthe water. Each run was conducted using new soil andmulch. The soil was compacted to a density of 1.43 mg/m3, saturated, and left for 1 day for the gravitational waterto drain out. This was done to simulate field conditionsbecause in Nigeria the mulch is applied after ploughing,which is done after the first few rains. After each experi-mental run, the sediment was separated from the runoffwater using a No. 22 sieve, with an apparent opening sizeof 0.2 mm followed by sedimentation–decantation of asample of the filtrate after it had been thoroughly mixed.The volume of water infiltrated was calculated as the differ-ence between the volume of water added and the sum ofrunoff and volume intercepted by the mulch. The amountof intercepted water was estimated as the difference inweight of the mulch before and after each run divided byunit weight of water.

Analysis of variance was performed on the data to deter-mine the mean effect of the treatments and their inter-actions. A response surface of the independent variablesversus the dependent variables was carried out for each soiltype using multiple regression analysis and three-dimen-sional plots of the slope and mulch rate treatmentresponses (averaged over the duration). Linear, quadraticand cubic regressions were performed on the mean datafor the soils to determine the best relationship betweenthe mean value of the dependent variables (soil loss andrunoff) and the soil components (sand, silt, clay andorganic matter).

3. Results and discussion

3.1. Mulch rate and surface runoff

The analysis of variance is shown in Table 3. There wasno significant interaction between duration and all theother variables. Both soil loss and surface runoff increasedalmost linearly with duration for all soil types and mulchcover. There was a three way interaction between soil type,mulch and slope.

The effect of mulch rate on runoff losses for the differentslopes and soils is shown in Table 4. The results were aver-aged over all the durations and presented as relative values(percent of total rainfall). There was a positive correlationbetween runoff amount and slope. Mean runoff losses were44%, 51% and 59% for the 6%, 9% and 12% slopes, respec-

tively. Mulch rate also had a significant effect on runoff.Mean runoff losses were 80%, 59%, 41% and 27% with0%, 30%, 60% and 90% ground cover of mulch, respec-tively. Although the reduction in runoff for Iwo series gen-erally agreed with Lal (1979) who used similar applicationrates of rice straw on the same soil, the reductions in thisstudy were larger in spite of the greater rainfall intensity.This indicates that elephant grass may be a good alterna-tive to rice straw in retaining water. In addition, the easeof obtaining the grass makes it more adoptable by farmersin Nigeria than rice straw because few farmers in this areaof study grow rice.

3.2. Mulch rate and soil loss

The result of the effect of mulch rate and slope on soilloss is shown in Table 5. Like runoff, the soil loss increasedwith increasing slope and decreased with increasing mulch

Table 5Soil loss (kg/ha/mm) from three Nigerian soils as affected by elephantgrass mulch cover and slope (average for all duration)

Soil series Slope (%) Mulch cover (%)

0 30 60 90

Iwo 6 10 (1)a 3 (1) 2 (1) 1 (0)9 16 (2) 9 (1) 4 (1) 1 (0)

12 20 (2) 12 (1) 7 (1) 2 (0)

Apomu 6 62 (4) 20 (2) 6 (1) 5 (1)9 65 (4) 44 (3) 9 (1) 6 (1)

12 69 (4) 51 (3) 11 (1) 9 (2)

Jago 6 20 (2) 13 (1) 3 (1) 2 (1)9 22 (2) 19 (1) 4 (1) 3 (1)

12 23 (2) 20 (1) 4 (1) 4 (1)

a Standard deviation values are in brackets.

K.O. Adekalu et al. / Bioresource Technology 98 (2007) 912–917 915

rate. However, the decrease in soil loss with mulch rate didnot follow the same order as with runoff. Soil loss fromApomu soil was higher despite its appreciable reductionin runoff volume. However, soil loss from Iwo soil waslower than that of Jago probably because of the lowerorganic matter and higher sand content in the Iwo soil.The reduction in soil loss values for Iwo were lower thanthe values obtained by Lal (1979) using rice straw, in spiteof the higher intensity used in this study. For all the soils, asubstantial reduction in soil loss occurred as 60% mulchcover, but the reduction was more pronounced in Apomu.A 60% mulch cover gave a reduction in soil loss over thenon-mulch plot by 70%, 86%, and 81% for Iwo, Apomuand Jago soils respectively, while a 90% mulch cover gavean average reduction of 94%, 91%, and 86%, respectively.

3.3. Mulch rate and infiltration

The influence of the mulching rate and slope on infiltra-tion is shown in Table 6. The result shows that the percent-age of water infiltrated increased with increasing level ofmulch cover and decreased with increasing slope. There

Table 6Cumulative Infiltration (percent of total rainfalla) from three Nigeriansoils as affected by elephant grass mulch cover and slope (average for allduration)

Soil series Slope (%) Mulch cover (%)

0 30 60 90

Iwo 6 39 (3)b 46 (2) 56 (3) 60 (3)9 27 (2) 35 (2) 48 (3) 54 (3)12 18 (2) 27 (2) 39 (2) 46 (3)

Apomu 6 18 (1) 34 (2) 53 (3) 61 (3)9 16 (1) 30 (2) 41 (3) 54 (3)12 10 (1) 22 (2) 32 (2) 36 (2)

Jago 6 21 (1) 34 (2) 50 (3) 59 (3)9 18 (1) 29 (2) 41 (3) 55 (3)12 13 (1) 25 (2) 38 (2) 48 (3)

a 150 mm of rainfall.b Standard deviation values are in brackets.

was no significant difference (P > 0.05) between the 6%and 9% slope from 30% and greater mulch cover forApomu and Iwo soils while the infiltration for 12% slopewas consistently lower than those for other slopes for allmulch cover for the two soils. For the Jago soil there wasno significant difference (P > 0.05) in percent infiltrationamong all the slopes. The mulch increased infiltrationgreatly in Apomu soil at the lower slopes but was less effec-tive at the 12% slope. At this high slope, the delay periodbetween rainfall and runoff on set by the mulch may below thus reducing infiltration. The maximum percentageincrease in infiltration of the mulched plots over the non-mulched plots for the three slopes is shown in Table 7.The mean values of percentage increase of the mulchedplots over the non-mulched plots were 27%, 36% and38% for Iwo, Jago, and Apomu soils, respectively.

3.4. Statistical analysis results

The computed surface response models for surfacerunoff are as follows:

RO ¼ �0:6mþ 5:4s� 2:6� 10�3m2 � 2:0� 10�3m2s

þ 3:5� 10�5m3 � 7:5� 10�3s3 þ 8:1� 10�7m2s3

� 1:0� 10�9m3s2 � 1:5� 10�4ms3

� 7:3� 10�9m3s3 þ 30:1 ðR ¼ 0:98Þ. ð2ÞRO ¼ �0:5mþ 1:5s� 3:7� 10�3m2 � 4:6� 10�3m2s

þ 6:7� 10�5m3 þ 1:1� 10�2s3 þ 1:3� 10�5m2s3

� 2:7� 10�7m3s2 þ 2:4� 10�5ms3

� 1:1� 10�7m3s3 þ 89:0 ðR ¼ 0:86Þ. ð3ÞRO ¼ 0:5mþ 3:7� 10�3s� 2:7� 10�3m2 � 3:8� 10�3m2s

þ 4:8� 10�5m3 þ 5:0� 10�3s3 þ 9:3� 10�6m2s3

� 1:6� 10�7m3s2 � 6:8� 10�8ms3

� 8:5� 10�8m3s3 þ 77:8 ðR ¼ 0:88Þ ð4Þ

for Iwo, Apomu and Jago soils, respectively.Those for soil loss are:

SL ¼ �0:5mþ 3:2sþ 5:1� 10�3m2 þ 1:8� 10�3m2s

� 2:5� 10�5m3 � 6:0� 10�3s3 � 1:5� 10�6m2s3

� 1:4� 10�9m3s2 � 6:3� 10�5ms3

þ 7:5� 10�9m3s3 � 8:1 ðR ¼ 0:97Þ ð5Þ

Table 7Maximum percentage increase in infiltration of mulched plots over thenon-mulched plots for the three Nigerian soils

Slope Soil type

Iwo Jago Apomu

6 22 40 409 28 36 38

12 30 32 36

Sand content (%)

64 66 68 70 72 74 76 78 80 82 84

Surf

ace

runo

ff (

%)

44

46

48

50

52

54

56

58

(a)

Colloidal content (%)

14 16 18 20 22 24 26

Soil

loss

(kg

/ha/

mm

of

rain

fall)

5

10

15

20

25

30

35

(b)

Fig. 2. Mean surface runoff and soil loss from the Nigerian soils as afunction of (a) soil sand content and (b) soil colloidal contents.

916 K.O. Adekalu et al. / Bioresource Technology 98 (2007) 912–917

SL ¼ �1:8mþ 6:0sþ 1:7� 10�2m2 þ 7:0� 10�5m2s

� 3:4� 10�5m3 � 2:1� 10�2s3 � 3:8� 10�5m2s3

� 1:6� 10�7m3s2 þ 1:5� 10�3ms3

þ 2:4� 10�7m3s3 þ 28:9 ðR ¼ 0:91Þ. ð6ÞSL ¼ �3:4� 10�2mþ 2:1s� 1:1� 10�2m2

þ 9:7� 10�4m2sþ 9:0� 10�5m3

� 6:3� 10�3s3 � 8:6� 10�6m2s3

� 2:3� 10�9m3s2 þ 2:4� 10�5ms3

þ 6:1� 10�8m3s3 þ 8:1 ðR ¼ 0:84Þ ð7Þ

and those for infiltration are:

CI ¼ 4:3� 10�3m� 5:4sþ 1:6� 10�2m2 � 1:8� 10�3m2s

� 1:5� 10�4m3 þ 7:3� 10�3s3 � 1:1� 10�5m2s2

þ 1:9� 10�7m3s2 þ 3:4� 10�4ms3

þ 1:1� 10�7m3s3 þ 69:6 ðR ¼ 0:88Þ. ð8ÞCI ¼ 5:9� 10�2m� 1:5sþ 9:2� 10�3m2

þ 4:6� 10�3m2s� 9:8� 10�5m3

� 1:1� 10�2s3 � 1:3� 10�5m2s3

þ 2:3� 10�7m3s2 � 2:4� 10�3ms3

þ 1:1� 10�7m3s3 þ 11:0 ðR ¼ 0:81Þ. ð9ÞCI ¼ 7:0m� 3:1� 10�2sþ 8:2� 10�3m2

þ 3:9� 10�3m2s� 7:9� 10�5m3

� 5:1� 10�3s3 � 9:2� 10�6m2s3

þ 1:9� 10�7m3s2 þ 2:6� 10�6ms3

þ 8:2� 10�8m3s3 þ 22:2 ðR ¼ 0:89Þ. ð10Þ

where, RO is runoff in percentage of total rainfall, SL is soilloss in kg/ha-mm, CI is cumulative infiltration in percentageof total rainfall, m is ground mulch cover in percentage, s isground slope in percentage and R is regression coefficient.

The equations show linear, quadratic and cubic func-tions of the independent variables and combination ofinteraction terms. Some of the multiple power interactionterms were eliminated without significantly altering the fitof the model to the data. The regression coefficients indi-cate that about 0.90, 0.89 and 0.86 of the surface runoff,soil loss and infiltration are accounted for by the indepen-dent variables, respectively.

The equations gave substantial soil loss (30% reduction)and runoff (27% reduction) responses to slope and mulchrate cover at low level of 30% mulch cover and high levelof 12% slope. This is evident of the mulch ground coverby slope interaction. At high mulch cover (>60%), the soilloss was almost uniform for the different slopes. However,the effect of slope on runoff was still evident at the highmulch cover, though it was significantly reduced. Fig. 2shows that the mean surface runoff was correlated nega-tively with sand content (r = 0.65), and mean soil losswas correlated positively with colloidal content (r = 0.62).

4. Conclusion

Increasing soil cover with a mulch of elephant grassdecreases runoff and soil loss and increases apparent infil-tration at the slopes tested. Mulch increased infiltrationgreatly on the Apomu soil at the lower slopes but it was lesseffective at the highest slope. The influence of the mulch onsoil loss reduction on the Apomu soil was more pro-nounced as in runoff. Increasing mulch rate increased thecumulative infiltration of the Jago soil more than for theIwo soil.

The differences among the soils may be attributed totheir differences in texture and organic matter content. A60% mulch cover for the Apomu soil may adequately pre-vent soil loss and reduce runoff at lower slopes. For the Iwoand Jago soils, up to 90% cover may be necessary especiallyif the organic matter is low and the sand content is high.This study shows that elephant grass, an alternative to ricestraw, can be used effectively to reduce erosion and increaseinfiltration on sloping land in South Western Nigeria. Thiswill be highly beneficial to the farmers especially during thesecond cropping season (July–September) where there isusually a dry spell (August break). This will not onlyreduce erosion but provide enough moisture storage tocarry the crops through the dry spell. In addition, the use

K.O. Adekalu et al. / Bioresource Technology 98 (2007) 912–917 917

of elephant grass will be economical since it grows natu-rally on fallow and uncultivated lands.

References

Adams, J.B., 1966. Influence of mulches on runoff, erosion and soilmoisture depletion. America Society of Soil Science Proceedings 30,110–114.

Bernett, A.P., Disketer, E.G., Richardson, E.C., 1967. Evaluation ofmulching methods for erosion control on newly prepared and seededhighway backslope. Agronomy Journal 59, 83–85.

Gee, G.W., Bauder, J.W., 1986. Particle size analysis. In: Klute, A.,Campbell, G.S., Jackson, R.D., Mortland, M.M., Nielsen, D.R.(Eds.), Methods of Soil Analysis. Soil Science Society of America andAmerican Society of Agronomy, Madison, Wisconsin USA, pp. 400–402.

Ghawi, I., Battikhi, A., 1986. Water melon production under mulch andtrickle irrigation in the Jordan valley. Journal of Agronomy and CropScience 157, 145–155.

Gregory, J.M., 1982. Soil cover prediction with various amounts and typesof crop residue. Transactions of the ASAE 25 (5), 1333–1337.

Khan, M.J., Monke, E.J., Foster, G.R., 1988. Mulch cover and canopyeffect on soil loss. Transactions of the ASAE 131 (3), 706–771.

Lal, R., 1979. Soil erosion on alfisols in Western Nigeria II effect of mulchrates. Geoderma 16, 377–382.

Lal, R., 1984. Soil erosion from tropical arable land and its control.Advances in Agronomy 37, 183–247.

Lattanzi, A.R., Meyer, L.D., Baumgardner, M.F., 1974. Influences ofmulch rate and slope steepness of an interill erosion. Soil ScienceSociety American Proceedings 38 (6), 946–950.

Mannering, J.V., Meyer, L.D., 1961. The effects of different methods ofcorn stalk residue management on runoff and erosion as evaluated bysimulated rainfall. Soil Science Society American Proceedings 25, 506–510.

McCalla, T.M., Army, J.J., Wittfield, C.I., 1963. Stubble mulch farming.Journal of Soil and Water Conservation 17, 204–208.

Meyer, L.D., McCune, D.L., 1958. Rainfall simulation for runoff plots.Agricultural Engineering 39, 644–648.

Meyer, L.D., Wischmeier, W.H., Forster, G.R., 1970. Mulch rate requiredfor erosion control on steep slopes. Soil Science Society AmericanProceedings 34, 928–931.

Ozara, N.A., 1992. The amount of soil cover provided by a unit mass ofguinea grass, banana leaf and palm leaf residues. Soil Use andManagement 8 (2), 103–105.

Schwab, G.O., Frever, R.K., Edminster, T.W., Barnes, K.K., 1993. Soilsand Water Conservation Engineering. John Wiley and Sons, Inc.,New York.