GTZ im TZ-Verlag

Nill, D.; Schwertmann, U.; S bel-Koschella, U.; Bernhard, M .; Breuer, J.

Soil Erosion by Water in Africa

Principles, Prediction and Protection

Deutsche Gesellschaft fur Technische Zusiimmenarbeit (GI )H

Die Deutsche Hibliothek - CIP-Einheitsaufn;ll~r~~e

NU, D.; Schwertmann, U.; Sabel-Koschella, U.; Bernhard,M.; Breuer, J.: Soil Erosion by water in Africa / Nill, I).; Schwert~nann, [I.; Sahel-Koschella, 17.; Bernhard,M.; Hreuer, J . [Deutsche Gesellschaft fiir Technische Zusarnmenarlxit (GTZ) G1nhH1.- Kossdorf: TZ-Verl.-Ges. 1996

ISBN 3-88085-5 14-5 (GTZ

NE: Nill, I).; Schwertrnann, LT:

I~ l l i s l i ec l 1 : l>eutsche Gescllschaft fiir Technische Zusammen:ir-lxit (GTZ) G 1lll3H 1'ostk~ch 5 180 D-05726 Esch1x)rn

Contents

Contents

List of tables List of figures Preliminary note 1 Causes for soil erosion 2 Damages caused by erosion

2.1 Damages in agriculture (on-site) 2.2 Off-site damages

3 The erosion process 4 Soil loss determining factors

4.1 Rainfall 4.2 Soil properties 4.3 Topography 4.4 Cover, tillage and protection techniques

4.4.1 Reduced tillage and notillage 4.4.2 Contouring 4.4.3 Soil cover by organic mulch 4.4.4 Inorganic mulch 4.4.5 Surface forming practices (contour ridging,

heaping, tied ridging, bedding etc.) 4.4.6 Bufferstrips 4.4.7 Contour bunds (stone bunds, earthen bunds,

diguettes) 4.4.8 Protective ditches 4.4.9 Terraces

5 Indicators for soil erosion 6 Assessment of soil erosion

6.1 Rainfall simulator studies 6.1.1 Laboratory studies with simulated rainfall 6.1.2 Field studies with simulated rainpal

6.2 Runoff plots 6.3 Erosion measurement within existing fields

6.3.1 Erosion nails 6.3.2 Sediment traps 6.3.3 Diverse techniques

6.4 Sediment yield from river basins 7 Soil loss prediction with the Universal Soil Loss Equation

7.1 The erosivity of rain (R Factor)

Contents

7.2 Soil erodibility (K factor) 99 7.3 The topographic factor (LS factor) 1 1 I 7.4 The cover and management (C) factor 117 7.5 The effect of protective methods -

Support practice factor (P) 145 7.5.1 Contouring, contour-ridging, tied-ridging 145 7.5.2 Bufferstrips 153 7.5.3 Contour bunds and heaps 156 7.5.4 Ditches and terraces 158

7.6 Soil loss tolerance limrts 160 Annex 1 Rainfall and erosivity 163

Annex 1 . 1 Erosivity for single sites 1 64 Annex 1.2 Erosivity regressions 168 Annex 1.3 National iso-erodent maps 171 Annex 1.4 Regional iso-erodent maps 177 Annex 1.5 National rainfall distrib~~tion maps 179 Annex 1.6 Rain volun~e and distribution for \~ngle I94 Annex 1.7 Estimation of the erocivity of the 10 year stonn 2 10

Annex 2 Slope length and gradient 2 12 Annex 2.1 Device for meawing \lope length and gradient 2 13 Annex 2.2 Conversion o f slope gradient in degrees to percent 2 14

Annex 3 Cover and management factor 215 Annex 3.1 Nurnber of day in the year- and co~responding datc 2 16 Annex 3.2 Field methods for the meacuremcnt o f mulch

cover and canopy cover 217 Anncx 3.3 Growth curve\ for mono- and mixcd crop4 226 Annex 3.4 Detailed C factors 230

Annex 4 Protection and management 257 Anncx 4.1 Detailed support and management ( P ) factor\ 258 Ani~cx 4.2 Some useful $pecic$ for soil and water

conservation 263 Literature 269

List of tables (short titles)

Tahle 21-1:

Table 21-2:

Table 3-1 : Table 42-1 : 'l'able 42-2: Table 449-1:

Tahle 5- 1 :

Table 64-1: Tahle 7 1-3:

Table 72- 1 : 'l'able 72-2: Table 72-3: Table 72-4:

Table 72-5:

Table 72-6:

Table 73- 1 : Table 73-2: 'I'able 73-3:

Table 74-1 : 'l'ahle 74-2: Tahle 74-3:

Table 74-4: Table 74-5:

Table 74-6: Table 74-7:

Sediment enrichment ratio for different grain sizes on an Alfisol 17 Sediment enrichment ratio for organic carbon and major nutrients under different cropping systems 17 Mean runoff coefficients from natural rain 28 Soil loss on different barefallow soils 3 5 Soil propel-ties influencing soil erosion 36 Characteristics and applications for different types of terraces 56 Soil erodibility of soils from different parent materials 66 Annual suspended sediment load of African rivers 83 Erosivity and rain data for sites, countries and regions 93 Soil loss and erosion depth under barekillow 100 Structul-c classes for use in the USLE 102 Permeability classes as used in the USLE 102 Determination of per~neability class by profile information 103 Runoff, soil loss and soil properties for the three erodibility groups 107 Convcrsion of Keqa to Ktrop for soils in group 1 and 2 1 1 0

Slopc lcngth exponent ( m ) for tiifferent gradients 1 12 Soil loss of slope segments 115 Examplc fur the consideralion of ;in irregular slopc with changes in soil erodibility in the USLE I I6 Crop stages as del'ined for the USLE 117 C hictor for a groundnut-maize system I I9 Annual erosivity distribution for Douala/ Camcrow 13 1 Residual effects of savannah and forest fallows 127 Annual C I'uctor calculated with the erosivity clistrihution of sitcs from diffi.1-ent climatic zones 128 Average C k ~ t o r for notill 130 Subfactor c2 for the cl'l'ect of mulch covcr 132

Contents

Table 74-9a: Average C factors for forest, bush and grass vegetation 133 Table 74-9b: Alternative determination of C factor 134 Table 74-10: Average C factors for banana 134 Table 74-11: Average C factors for pineapple 136 Table 74-12: Average C factors for cassava 137 Table 74-13: C factors for ~niscellaneous perennial crops 138 Table 74-14: Average C factors for groundnut 139 Table 74-15: Average C factors for maize 140 Table 74-16: C factors for millet and sorghum 141 Table 74-17: C factors for upland rice 143 Table 74-18: Average C factors for miscellaneous crops 144 Table 751-1: P factor for contouring 146 Table 751-2: Correction of P factors for ridges with side slopes 150 Table 751-3: C x P factor for temporary established ridges 153 Table 752-1: P factors for bufferstrips as calculated by the RUSLE 156 Table 753-1: Correction factor for bunds on different slopes 158 Table 754-1: Sediment delivery ratios for side-slopes 159 Table 76-1 : Rates of soil weathering 161 Table 76-2: Tolerance limits proposed for tropical soils 162 Table 11-lAnnex: Erosivity and rain volurne for single sites 165 Table 12-IAnnex: Regressions for the calculation of erosivity 168 Table 16-1Annex: Annual rain volume and monthly distribution 194 Table 34-1Annex: Detailed C factors for forest, bush and

grass vegetation 230 Table 34-2Annex: Detailed C factors for banana 234 Table 34-3Annex: Detailed C factors for pineapple 235 Table 34-4Annex: Detailed C factors for cassava 238 Table 34-5Annex: Detailed C factors for groundnut 243 Table 34-6Annex: Detailed C factors for maize 245 Table 34-7Annex: Detailed C factors for diverse crops 250 Table 34-8Annex: Detailed support and management fhctors

for notillage 255 Table 41-1Annex: Detailed P f'actors for contouring 258 Table 41-2Annex: Detailed P factors for ridges 259 Table 41-3Annex: Detailed P factors for mounds 260 Table 41-4Annex: Detailed P factors for bunds 260 Table 41-SAnnex: Detailed P factors for bufferstrips 26 1 Table 42-1Annex: Useful species for erosion control, etc. 263 Table 42-2Annex: Useful trees according to rainfall area 268

List of figures (short titles)

Figure 1-1: Causes at the origin of accelerated erosion 12 Figure 1-2: Importance of activities which destroy the soil resource 13 Figure 21-1: Impact of increasing bulk density on productivity 16 Figure 21-2: Influence of soil erosion on potential productivity 18 Figure 21-3: The vicious cycle of soil erosion 19 Figure 3-1: Transpot capacity increases with runoff velocity 23 Figure 3-2: Runoff velocity increases with gradient 24 Figure 3-3: Runoff velocity as related to gradient 24 Figure 3-4: A surface water layer decreases or increases soil loss 25 Figure 3-5: Runoff generation on slopes with sealing and

permeable soils 27 Figure 3-6: Infiltration as influenced by management and vegetation 29 Figure 3-7: Rain volume per unit area and surface storage decrease

with increasing gradient 3 0 Figure 445-1: Ridging, tied-ridging and bedding 44 Figure 445-2: Influence of slope on height of heaps 46 Figure 447-1: Contour bunds form small terraces after some years 49 Figure 447-2: Permeable, semi permeable or impermeable bunds 49 Figure 447-3: Dykes form deposits which are used for cultivation 50 Figure 448-1: Diagram of drainage ditches and Fanya Juu terraces 5 1 Figure 448-2: Hillside ditches f~cilitate access and transport 52 Figure 448-3: Watershed conservation plan with hillside ditches 52 Figure 449-1 : Key parameters for terrace planning 54 Figure 449-2: Different types of bench terraces 54

Figure 449-3: Intermittent terraces can be transformed to bench terraces 55 Figure 449-4: Orchard terraces in combination with individual basins 56 Figure 5-1: Concepts of a peneplain and a pediplain 59 Figure 5-2: Pediplains at different altitudes as formed in Cameroon 60 Figure 5-3: Development of gullies from initiation to maturity 6 1 Figure 5-4: Blockslide and translational slide 6 3 Figure 5-5: Critical storm duration and intensity for landslides 64 Figure 5-6: Sapre growth of trees caused by soil creep 65 Figure 5-7: Runoff and suspended sediment load of watersheds with

different vegetation 67 Figure 611-1: Diagramm of a flume for laboratory rainfall sirnulation 71 Figure 612-1: Set-up for rainfall simulation tests i n the field 73 Figure 612-2: Runoff/soil loss diagramm for a dry- and a wet-run 74

Contents

Figure 62-1: Divider tank systern and Coshocton wheel for the measurement of runoff and \oil loss 76

Figure 631-1: Set-up ofermion nail\ on a slope 78 Figure 631-2: Measurement of nail height with a \lide calliper 79 Figure 632-1: Sediment trap for measurement of runoff and coil 105s 80 Figure 633-1: Calculating the time since exposure of tree root\ 82 Figure 633-2: Meawring the height of loqt soil by expmed tree root$ 83 Figure 64-1: Relationship between watershed drainage area and

sediment delivery ratio X 5 Figure 71-1: Strip chart of' a 20 rnrn \ t o m 90 Figure 71-2: Correction of' ero\ivity for the effect of water mulch 93 Figure 72-1: Change of \oil erodibility with cumulat~ve erosivity 100 Figure 72-2: Compar~son of calculated and mea\iired \oil erodi bll~ty 103 Figure 72-3: Mea\ured and predicted erodibilitie\ after applicat~on

of discrin~~nant functions 105 Figure 73-1: Diagram for the determination of LS fi~ctor\ 113 Figure 73-2: Examplec tor the determination of erosive \lope length 1 13 Figure 74-1: Mean annual ero\ivity d~\tribution for coa\t;il. tnland

and northern Catncroon 122 Figure 74-2: Influence of 111ulch on so11 lo\\ 125 Figure 74-3: Subfactor as influenced by canopy cover and crop hc~ght 126 Figure 74-4: C factor for d~ffcrent banana dencitics 135 Figure 75 1-1 : P factor+ for different r~dge he~ghtr 137 Figure 751-2: Isohyete\ for the I0 year \tor111 volume 138 Figure 752-1: P fitctors tor riparian butfer\trips of d~fferent width\ 155 Figure 11 - 1Annex: Eror- as cawed by I~near ext~apolut~on of erovvlty lh3 Figure 17-1Annex: Stornl erosivity a\ related to \tomi \olunle 2 1 1 Figure 21-1Annex: Water level for n~e;r\urement of \lope-length

and gradient 2 13 Figure 32-1Annex: Me\urernent of cover by the metcr$tich method 2 17 Figure 32-2Annex: Cord and knot method for cover mc;l\uremcnt 1 18 Figure 32-JAnnex: Marh~ng out ;rn area with cl~fferent de\ ice\ 2 19

Figure 32-4Annex: Sclecteci cover:ige\ for the ciilihl;ttiol~ of the cye 220 Figure 32-5Annex: Sighting l'rainc for me;i~urement of caiiopy irnd

mulch cover 22 1 Figure 32-6Anncx: Oh\rrvntiori error as ~ntluenced by ~ l i t ~ l t height 22 1 Figure 32-7Annex: Modified sighting fratnc for errorlec\

coverage a r~d cover height ~nea\ure~llent\ 222 Figure 32-8Annex: S ~ g h t ~ n g frame w ~ t h mirror for tall crops 223

Figure 33-1Annex: Canopy cover of Barnbara nut. cmavalia and cowpea 227

Figure 33-2Annex: Canopy cover of cawiva, ca\sava/mai~e mixcrop and pigeon pea 227

Figure 33-3Annex: Canopy cover of cotton, sunflower and tobacco 228 Figure 33-4Annex: Canopy cover of groundnut and soya 228 Figure 33-5Annex: Canopy cover of maize, rice and wrghum 229 Figure 33-6Annex: Canopy cover of tea 229

Preliminary note

Enhanced soil erosion research in Africa looks back on about 25 years of experimentation and recording with respect to soil loss prediction. A lot of information was gathered during this time which led to contradictory results about the applicability of the Universal Soil Loss Equation (USLE). This book tries to synthesize the latest knowledge and to evaluate it for practical soil loss prediction. This meant the gathering of a lot of data in tabular and graphic form which might be cumbersome for some readers. Nevertheless, we hope that it will be helpful to have these data assembled in order to save time for searching in many different journals which often can hardly be obtained locally. The ultimate aim of the book is to help understand the processes, to make the reader sensitive for recognizing them in the landscape and to allow him to quantify of the influence of agronomic measures on soil loss rather than to give detailed technical data and sketches.

People starting to get occupied with erosion problems will find basic knowledge about processes and effects and the necessary literature for more details. The book is also thought as a help for people concerned with the planning and realization of soil conservation activities. It allows to detect areas of high erosion risk which in turn facilitates the allocation of conservation efforts. The absolute results of calculations can be subject to substantial error whereas the relative differences between single measures are comprehensible. However, if a process varies in magnitude by a factor of 1000 (soil losses can be as small as 0.1 t/ha and as large as several 100 t/ha), an estimate which is wrong by a factor of 2 (= IOO%I error) is still a reasonable estimate. At the same time. this means that projects and research should continue to improve and enlarge the database in order to improve estimates. For this reason, the authors would appreciate to receive further data on measurements.

In Chapters 1 to 6 the reader finds descriptive information about causes, damages, processes, recognition and measurement of soil erosion. Chapter 7 is dedicated to soil loss prediction with the USLE. For the pure technical procedure of soil loss prediction the reader can refer to Chapter 7.

Causes for soil erosion

Soil erosion is a process acting over tens and hundreds of years. Its effects are normally only obvious, if they become disastrous. Until now, research focused on the physical causes of erosion. However, frequent failures of soil conservation projects showed that the causes were much more complex. A F A 0 study revealed that the lack of adoption of new conservation practices was a major reason for project failure even though technically sound practices were used (Hudson, 1991). Today, i t has largely been agreed upon that soil conservation will not be successful in many countries even by using the best available PI-actices if Inan and the social, economic and political conlext are not considered. Fones-Sundell (1992) summarized the problern very pointedly by saying: 'Neither engineering nor biological measures alone can eradicate erosion in a socio-economic system which makes non-optimal use of natural resources a necessary and often profitable form of behaviour for the individual.'

The cause-effect diagram in Figure 1 - 1 illustrates the problem of soil erosion. Soil erosion as caused alone by natural, physical factors (clim3te. soil, topography) is known as 'geologic' or 'normal' erosion (Bennett & Chapline, 1928). It is small enough to allow the sustainable growth of a natural eco-system. Geologic erosion ranges from several hundred kilogran~s per hectare for tropical bush and grass vegetation to below 100 kglha for tropical forest (Nill, 1993; Roose, 1975). Soil formation in the tropics is supposed to be in the same range. According to measurements in Central Africa, 150 to 400 kglha of new soil is formed each year (Owens, 1974). Dunne et al. (1978) found an annual formation rate of 150 to 300 kglha in Kenya whercas Kaye ( 1959) reported a rate of 15 t/ha on limestone i n Puerto ~ i c o l .

Soil loss becomes critical if socioeconomic and political factors favour erosion (man-induced erosion). The main factors are:

1 more detail\ a b o ~ ~ t soil torination ratcx arc given in C'hirpter 7.6.

Chapter 1

Figure I - 1: Physic-ml as n3c.11 as socio-ecot2ot~zic crnd political c-ctuses (ire clt the 01-igitz of crc*c.elerc~tucf ut-osiorz

Poverty of the farmers: Small farmers are obliged to cultivate their land as often as possible in order to assure their subsistence. The lack of capital hampers the application of intensive conservation measures and the use of inputs to restore soil fertility. Decreasing soil fertility leads to the extension of the cropping area, soil mining and finally migration of the farmers. I t is estimated that deforestation proceeds thirty times faster than reforestation (FAO, 199 1 ). Overgrazing, deforestation and agricultural use are major fxtors for soil destruction (Figure 1 -2).

soil erosion .-------------- r(

I

4

socio-economic physical causes

political conditions

man-induced erosion

geologic eiosion Y

human activity

A i A

v v 1 I

otl

'iculture (28.0%) -

over-grazing (35

llnbalanced population density: O\c~-lx)pi~l:it~oll u\uiilly I \ dctri~ncnt:~l lor the \ o ~ l ~-tt\oi~rcc. O\erpopi~lation i \ not only the rc\ult o f a high popillation ~sowt l i . LC11c\t cle\clopment\ \u,gge\t that segional ovespopulatio~i i \ oftell clue to 1111$1-:1tion cau\ed by stseam\ of' rel'ugee\ from bar \ os cn\i~-onme~lt~il cl~\a\ter\. For example, tlul-ing thc Sahel tl~oi~glit 01' the early 70\. 1 ~ i l ~ l l ~ o n Rl~l-hint.\e cclu~il to one {ixtli ol' the co~~n t r i c \ popillation. left the11 l i o ~ ~ i e \ (FA() . 1900). lJ~lde~-lxq't~lat~on al\o cau\e\ \el-lous clalliagc. Voscl ( I O X X ) clc\cr~bcd ~ h c dctc.t.~orat~o~i ot' a traditional re1-1-accd ago-5y5tt.m in Ycmcll al'tcr thc ~ i l i g r ; ~ ~ ~ o n o f the rural popi~l;~tion to ~ i e~g l ibo i~~- i~ lg coi~~itr ic\ . Soil c.o~l\cr\:~tion ~ o r h s 111 11ld1;~ bere ol'len :tbanclonecl clue to the rccruil~llenl ot thc [ii~rha. wh~cti wese the iilorc acllcc in \oil con\e~- ia t~on. 111to the K1.1ti\11 Arilly (Blaihic. 1OS5 1.

'I'he in\titutional frame: Go\ clnmcnl in\lili~tion\ ~n \ i~ t f~c~e l l t l y contsol the u\e of natul-,ll r.esoulce\ 'I lie I:~ch ot tlet~necl con\cscation ~ ~ o l ~ c y clilcl l abs to i.cpi~l,~tc the Lr\e of I,~ncl. to d e l ~ n e tllc o w ~ l e r \ I ~ ~ p ~illd to I~ i c~ l~ la l c llie co~ll~lle~-c~~~l~\~~tloll 0 1 yood\ p~-c\cnt\ t;ismer\ to apply \or1 con\er\ 1112 product~on method\

Politicians often give priority to short term benefits from export crops on the expense of soil conservation. Extension and conservation services are mostly inadccluately equipped and trained and suffer from a lack of cool-dination (Sheng, 1989).

The land tenure system: Only farmers who own their land or have secure access to them land for a long time arc interested In longterm rnalntenance of thi\ I-csource. Restricted access to fert~le land for social groups or farnily members (e.g. Y O L I I I ~ people, wornen) lead\ to the exploitation of stccp slope\ and mal-ginal, fi-agile soils. The traditional heritage system \ometimes favours an extensive fragmentation of the land which obstructs the adoption of conservation practices.

Tradition, believes and illiteracy: The degree of illiteracy intluences the adoption of new conservation practices and other cultural techniques. Small farmers commonly perceive erosion as a natural process and are not aware about its influence on productivity. Lack of knowledge exists along with effective local conservation methods (Tato & Hurni, 1992). The adoption of new practices always needs a11 effort and includes some risks. Farmers, as inost other social groups, need time to adopt new ideas.

It is wrong to conclude from these comments that measures should be solely applied riccording t o the socio-economic and political conditions. Soil loss by erosion is irrever-sible. Therefore, conservation activities can not be delayed unti l socio-economic and political conditions are favourable. Conservation thinking and conservation activities must proceed sirnul- taneously.

Damages caused by erosion

Surplus rain water leaving a fielti on the soil surfi~ce is called runoff. Runoff first causes dutnage on the field (on-site damage) by entraining fertile topsoi 1 and by reducing the available iumount of water for plant growth. Once left thc I'ield. the runoff is enlarged from adjacent fields ancl may entel- rills, ditches. s111all rivers, passes lakes and stl-earns and finally reaches the ocean. On its way, sediment is picked up arid deposited which causes further darnage outside the Sields (off-site darnage). Both, on-site and off-site cliumagc need t o be co~~siclered in order to assess the overall economic and ecologic effect 01'

Soil conservationists intend to protect the diverse functions of soils. Firnctions like ini'iltration and storage capacity of a soil to prevent tloocis and its filter function to purify water are o f tnajor concern for the urban environl~ient. In the rural cnvir-onment, however, it is soil fertility.

2.1 Damages in agriculture (on-site)

Soil proctuctivity depends o n a nutnber of physical, chemical ant1 biological soil properties. The ~nos t important physical ones are texture, structul-e ~unci depth o f the profile. They detel-mine the aniount of water aticl air stored in the soil, its capability to infiltrate ;~nd conduct water, its possibility /'or I-oot 2n)wth and the fines which can bind ant1 deliver nutrients t o t11c plant roots. 111 well structureel. deep soils even heavy storms infiltrate. Structural damage 011 sornc tropical soils is more severe than o n temperate soils as show11 by the influence of' bulk density on relative productivity (Figure 2 1 - 1 ). Profile tlepth and surface soil depth determine the water slor~tge and t11e volumc for water and nutrient uptake of thc roots.

Thc chen~ical fertility depencij on the aniount ot available nutrrenr\ rn :I {oil which 1 5 governed by \oil pH, 01-ganic matter content and othes characteri\tic\. Thc\e are greatly inl'luencad by parent r11atc1-ial and the conefition{ undet which a \oil wa\ So~.rned a\ well us by its 11~2.

rela

tive

prod

uctiv

ity [-

I

treatment

barefallow plow traditional notill

The etlrichlne~lt oj' the I'ille \oil fraction in the \edili~ent account\ for the hlghel- nutrrent content\ of the eroded wdimcnt a\ col-nparecl to the original \oil. Alli\on ( 1973; in Bouwtlian, 1989) reported ;I \edimcnt cnrichmcl~t ratio betweell 1.3 and 5 for 5011 organic carbon. Aina ct al. ( 197C)) ticmon\tratcd the enrichment of organic carbon and m;qor nutrient$ untiet. dil'l'cl-ent cropping \y\tenl\ (Table 2 1-2).

mean

cropping system I S O M 1 N I P I K I Ca / M g

grain size [mm] 4J.002

1 .6 2.3

1.2 1.5 1.7

barefallow cassava monocrop

0.002-0.0.5 1.6 2.6 1.6 I .O 1.7

maize-cassava mix-crop soybean-soybean

The iriipact of soil crosiorl on production depends o n the depth of the arable layer and the cluality of' the underlying hori7ons. Soil erwion i j

more cletriniental the shallower the \oil, thi \ being aggravated in area\ ol' irregular rainfall.

1.5

1.3

pigeonpea-pigeonpea mean

7 \ cd imrn~ r111-ichrricnt ratio = (prl-cc~itagc of a grain \i /c clabb in rhr \edi~iicnt) 1 (prrcrlitaye 01' rlic \ i /c i~ l a \ \ in the 01-iyinal boil)

0.05-0.125 0. 0.8

0.7 1.3 0.9

1.3

1.3

1.3 1.3

1.1

1.3

0.125-0.25 X 1 .O 0.7 0.0

0.0

0.9

1.3

I . I

1 . 1

0.9

1 .0

1.2

0.25-0.5 0 . 9 0.7

0.0

1 .O

0.9

0.')

1.4

1.2

1.4

1.2

0.5- 1 0.6 0.5 0.9

0.9 0.7

0.7

0.6

1-2 0.8

0 .3 0.0

0 .8

0.7

I .O

0.8

1.2

1.2

1.3 0.0

1 .O

1.3

1.1

1.3

Loss of fertile topsoil is most harmful on extremely leached Ultisols and Oxisols of the hulnid tropics where the subsoil contains very low amounts of nutrients and SOM compared to the less leached soils of the drier areas. SOM counteracts P fixation which explains why P fixation increases with increasing topsoil loss. Mbagwu et al. (1984) showed that the removal of 5 cm of topsoil reduced maize yield by 95% on a leached Ultisol but only by 52% on a less leached Alfisol. Maize died off at 30 cm height on an Ultisol in Canicroon which had lost its topsoil duri~ig 5 years of barefallow (Nill, 1993).

Some crops tolerate ero\ion better than other\. La1 (1976a) mea\ured 52% less m a i ~ e yield but only 38% leg\ cowpea yield i f 10 cm o f surface \oil wcre stripped off . Yield decrease\ are generally in the order gra~nineae > grain legir~nes > tuber crops (El-Swaify, 1990).

Fi,y~tt.c~ 21-2: l t ~ f l ~ i t v i ~ * t ~ o/ soil o r o ~ i o ~ z 0 1 1 p o t t ~ ~ ~ t i u l p ~ - o t l ~ i ( ~ t i \ ~ i t ~ ~ t.eltrtot1 to soil t\,pe (Pic.1-c,e r t trl., 198.3)

FAVORABLE SURFACE AND SUBSOIL HORIZONS

FAVORABLE SURFACE FAVORABLE SURFACE AND CONSOLIDATED AND UNFAVORABLE OR COARSE FRAGMENT SUBSOIL HORIZONS SUBSOIL HORIZONS

ACCUMlJl ATED EROSION (TIME) - The extcnt to which soil erosion efl'ects productivity depends on the

depth-distribution of fertility pat-a~neters in the prol'ile (t;igure 2 1-2). A dccp itncl honiogeneous soil acts with a slow productivity decline with increasing crosion whereas yields drop sharply with increasing soil loss on soils with unf'avourable subsoil properties.

The econorliic damage of soil erosion is alarming. In Zirnbabwe i t is estimated that farmers loose three times Inore nitrogen and phosphorus by erosion than they apply to their fields. 20 to 50 US $ o n arable land and I0 to 80 US $ on grazing land would be necessary to substitute these nutrient losses by fertilizer (FAO, 1990). I t must be st]-cssed that most erosion damages can hardly be cured (e.g. compaction, structure) or are completely irreversible (e.g. water holding capacity).

increased soil loss

degraded soil properties less plant cover

reduced structural stabilit and roots

Damages to agricultural productivity are not only caused by degrading soil properties but also by direct impact of runoff. Roots and seeds are washed ou t of' the soil. Seeds and seedlirigs on the foot-slopes are buried by the deposited sedil-nent. Kills and gullies hamper access to the fields. impede farm operations and transport. Deep rills and gullies form drainage syste~ns which drain the adjacent areas and lead to considesable loss of water.

reduced fertility reduced plant growth

I,oss of soil fertility by soil erosion is a self-enl~ancing process. Soil erosion reduces structural stability and soil fertility. Reduced structul-al stability decreases infiltration and may increase the amount of transportable rnateri:il. Reduced fertility causes poor plant growth, canopy covet- anti I-oot soil inter-actions. In till-n. runoff and soil loss are accele~-ated (Figure 2 1-3).

2.2 Off-site damages

Part o f the surfilce runof'f' ;inti the suspended sedimunt Icavc the ficlcts and g s a ~ i ~ l g lands and are concentratecl in the surficial drainage system. Depending o n the transport capacity of the flow. sediment is picket1 up or deposited. Rills are widcncd to gullies, channels arc deepenecl (clia~lnel erosion) and strean) banks ar-c undercut (stream b:uik erosion). Ditches. ro~ids and bridges arc damaged. The kist runoff leads to :i loss of' water I'ronl rhc landscape and results in a I xge fluctuation of' tlie rivers. Some ~-i \ .e~-s start to become olily seasonal. Tlie groundw:itcr table is lowel-cd w11ic11 al'l'ect\ the \ cgctation anti causes water shortages in wells.

Downstream st.clitiientation silts up irrigated fields. ditches. channels. clams and harbours. 111 areas with intensive agriculture. pesticicles ancl nutrients dissolved in ri~nofl'ol- ilttaclied to the sediment becorne a serious problem. Disasters at this extent arc dil'l'icult to quantify but national economics si~l'l'cr important expcnditur-cs 1.01- their restoration.

The amount ot 4cdimcnt transported by \olne \trcam\ can he enormoil\. For cx~ui~plc . the r i ~ e r per her^.;^ 111 Kenya rcccicc\ :in a\ el-age ot 105 1 1 x 1 - ye;11- 1'I-0111 eacli 1iecla1.c of' i t \ 13 10 hm' large uatt.l.shcc1 (13ilnnc. 1075 In: Walling, 1084). corre\ponding to an avcragc l o ~ c r i ~ i g of thc ~~a t e r \ I i cd by ahout 15 clll in I0 yeat-\. Ilam height\ on thl-ce 01' Moroccoq\ clam\ hat1 to be ~nct-ea\etl In order to maintain the \tot-age capacity. In 01-dcl- to prc\cr\c lllc curl-ent watel- \torage cap;icity in Mon)cco. one tirw ~ ; I I N ~ ~ i l l i

150 million rn{ nueci\ to he con\tructcd C;ICII yea1 (FAO. 1993).

The i'lcclucnt Ilood cl~\a\tet-\ In lt~clia arc anotl~cr \bell-hnomn cxamplc. They ,Ire expl;llncd by tile deforc\t:itio~l of the IIr~i~al :~qa\ Scdi~nent\ of tlie Hrahrnaput~-~~ i111d (;;lnge\ rlver In India hate tolmccl ,I 50000 h ~ i i ~ lar-gc \hallow In the Gult 01' k n g n l . I'hc KO\I Kt\ el 111

Klhal-Ilndi:~ II:I\ bur.ied 15000 hrii3 of icrtilc land wrth gr.a\cl :uid wnd ( Kollm:~nnspergcr. 1079)

Stitdie\ on the \ve\t-coa\t ol Sumatra \bowed that the \eci~mcnt lo~tcl ot tlie ~ - ~ \ c r \ wh~cli iricrru\ecl hy ciclorc\tntion. rllrnlng and channel

construction led to the destruction of the coral riffs off-shore (Hettler. 1994). Riffs are rich fishing grounds and help to protect the coast line. These are some examples, out of a large number which could be cited here, in order to \how the importance of off-site damage.

Present investments into soil conservation efforts are small compared to the i~nrnense investment in civil engineering aiming to repair the results of erosion. It is supposed that investing in soil conservation would have n higher cost-efficiency ratio and would protect both the soil resource and the down- stream areas and facilities.

3 The erosion process

Soil erosion can be regarded as a result of four processes (Foster & Meyer, 1972):

- detachment by raindrop impact - transport by raindrop impact (splash erosion) - detachment by the shearing forces of flowing water - transport in surface runoff (sheet or interrill erosion, rill and gully

erosion)

Rain falling on a soil causes increasing water saturation orland the formation of a seal at the soil surface. As a result of both processes, infiltration into the soil is decreasing. Water on the soil surface occurs as soon as rain intensity exceeds the infiltration rate. Before any runoff can occur, a small amount of water is needed to humidify the soil surface (detention storage). Once the detention storage is filled up, ponding occurs in the small depressions and irregularities of the soil surface (surface roughness) which form the retention or depression storage. Overflow of some depressions provides excess water to the ponds underneath. These, if filled up, in turn spill their water further down-slope. Thus, Inore and more water is

1

moving down-slope which may concentrate, dig out rills of increasing size and finally may cut deep gullies into the soil.

The amount of surface runoff (SRi) during a storm can be expressed as:

SRi = Pi - ( I + DS + RS) ( 1 )

with Pi rain volume of storm i [mm] I infiltration [mml DS detention storage RS retention storage

Sheet and splash erosion occur in areas of shallow sheet or interrill flow (few millimeters deep) whereas rill erosion is caused by concentrated rill flow. In the rills, fine sediment is transported as suspended load whereas coarser particles are dragged along as bedload.

3 \?el- definition rills can bc closed by normal farm ope[-ations.

22

3 The erosion process

The amount of transported soil and the size of particles depends on the transport capacity of the tlow. For a flow of given width, the transport capacity increases with increasing flow velocity (Figure 3-1 ) and tlow depth. Both depend on slope.

If the overland flow on a smooth surface is regarded as a water sheet with a certain depth and velocity, a unit volume of water can be picked out as an element with a defined weight (w) (Figure 3-2). W equals a force with a down-slope component fl parallel to the slope and a component f2 perpendicular to the slope. The down-slope force f l increases with gradient and speeds up the velocity of the element. In the Manning formula which is widely used to calculate flow velocities in channels, velocity augments as a function of the 0.5 power of slope (Figure 3-3) whereas transport capacity increases with the cube of flow velocity (Engelund & Hansen, 1967).

Figure 3-1: Tr(rnsport cnprrcity increases with runoff veloc-it,' as sholzw by the clr?!ount oftmn.(;yorted sedirnent (Auer.sw~nlc1, 1993)

mean flow velocity [m/s]

Chapter 3

Figure 3-2: Ruizc?f veloc-ity increnses with irzc.rea.sing grclclirnt due to t h ~ down .slope .fi)r.c.r ("'I) which is n con~ponrnt of the weight (ul) of'

gradient

3 The erosion process

Flow velocity is not equally distributed over the depth of the flow. In a laminar flow, velocity (v) increases to the square of depth (d) (Horton et al., 1934) which means that the water layer at the water surface is much faster than the layer close to the soil. This velocity distribution causes a force which lifts up soil particles from the ground and transport them. Depending on their size, they are either rolled and dragged along the soil surface as bed load or lifted up into the flow and transported as suspended load. In the shallow sheet flow, velocity is s~iiall. However, soil loss is enhanced by the energy supplied by the pounding rain drops. The drop impact causes turbulence in the tlow. Particles are heaved up, settle down and are heaved up again, thus, being transported towards the rills. In experiments of Mutchler & McGregor ( 1983), maximum soil loss occurred in tlow depths of 2 mni (Figure 3-4).

Figlire -3-4: A s~r~fircr cz7atrr lcryrr drc-r-eases or- itzc-rrcrses soil 1o.s~ drlwrzdirzg or1 its cleptll (Mutc.l?ler & MC Gregor, 1983)

surface water depth [mm]

Flow depth in rills i \ generally deep enough to minimize the influence of raindrop action. The amount of soil loss in the rills, therefore, depends altnost solely on the shearing forces of the flow and the saturation of its transport capacity. If the transport capacity of the rill tlow is saturated with 5ediriient from the interrill areas, the rills do not deepen. If the sediriient concentration i \ \mailer than the transport capacity, the flow picks up more sediment from the rills.

Chapter 3

Runoff is distinguished into two basic flow patterns attributed to differences in runoff generation. Horton flow occurs if runoff is caused by a rapid sealing of the soil which limits infiltration right at the surface (Horton et al., 1934). Dunne flow occurs if runoff is caused by saturation of the soil profile due to excess of rain, dense layers or shallow soil depth (Dunne 1978). Horton flow is characteristic for structurally weak soils which have enough fine earth to form a seal. Infiltration is rapidly decreasing after the onset of a rain even though the subsoil may still be dry. The runoff coefficient may reach 70 to 80% for single rains. Soil loss is limited by the amount of available sediment rather than by transport capacity. Dunne flow is characteristic for structurally stable soils rich in oxides, clay and organic matter. High infiltration rates can be maintained until the soil becomes saturated. Even if enough sedirnent of transportable size is available at the soil surface, soil loss is limited by runoff. Transport by splash erosion becornes more important.

Rain falling on a slope causes either runoff from almost the entire slope (sealing soils with Horton flow) (Figure 3-5a) or only from part of the slope (soils with high infiltration rate and Dunne flow) (Figure 3-Sb). Close to the upper slope end, runoff, even if present on structurally weak soils, is still too small and slow to transport soil. On structurally stable soils runoff seldom occurs on the upper part of the slope. It infiltrates into the soil and proceeds vertically or laterally in the soil. The lateral or interflow may add to soil saturation of the area further down-slope where runoff starts. Thus, runoff on both soil types but Illore so on soils with Dunne flow, leaves a 'zone of no sheet erosion'. Soil profiles on the watershed boundary are,therefore, relatively uneroded and can sometimes be used as a reference for the extent of erosion damage on the mid and down slopes.

3 The erosion pl.ocess

Figure 3-5: Runc?ff gufzercrtiotz on slol)c..c n'ith sec/lirlg ( ( I ) cuzcl l)errnecrblu ( h ) .c.oils (clftrr Chorluy, 19 78)

4 runoff area b

+ -

4

outlet drain

interflow

The runoff volurne produced by a rain depends on rain properties as well as soil and vegetation properties. Roose & Piot (1984) measured mean runoff coefficients (RC) of 20 to 40% and as high as 70% for individual storms. In own experiments with natural rain, RCs varied between as much as 30 5% on an Alfisol to as little as 1 C/c on an Oxisol. Small rains of' 2-3 mm could already generate runoff on sealing soils (Table 3-1) (Nill, 1993). Runoff' starts on some soils only some minutes after the beginning of rain. Pontanier et al. (1984) f'ound 1 to 4 min of artificial rain sufficient to generate runoff on hard setting soils ('sols hard&'), 2 to 20 min on Vertisols and 5 to 20 liiin for Ultisols. However, on the Acrustox shown in Table 3- 1 , 1.5 hours of rain with an intensity of 64 mmlh did not cause any runoff.

Trrhlc 3- I : Mew11 ~ - ~ i i ~ ( ? f c o < f f i c ~ i c ~ t ~ t ~ s fi-0111 i ~ o t ~ i r ( ~ l f -~r i i~ o 1 1 .\P\WI soils (Nill, 1993)

Soil covered by vegetation generally infiltrates tnore watcr than uncovered soil. Sabel-Koschella (1988) measured a 7 times higher infiltration volume undcr a natural savannah grass fallow ( 14 I0 ~nrnlh) compared to a sealed barefallow (2 10 mmlh) (Figure 3-6). If plowed and cultivated. thc raime (;oil infiltrated 450 mmlh.

soil

Paleustalf Andisol Kandiudalf 'l'rophumult Tropudult Hapludult Acrustox

With increasing arca. the total runoff volume becorner more. Ar rhown for plot4 of 70 to 550 rn' (Mutchler & Crees, 1980) and watersheds between 0.1 and 100 k~ i i ' (Dunne, 1978), runoff pcr unit area becolne5 rmaller with increasing watershed size due to longer travelling timc of the overland flow. The longer the flow stays within the watershed the more water and \oil can be retained in deprersionr or infiltrate.

number of storms

[-I 8 1 35 1

320 239

357 135 135

runoff coefficient

[%I 3 0 I8

I8

15 1 1

1 1

1

smallest runoff generating storm

[mml - 3 3

3 3 5 3 7

3 The erosion p roce \

infiltration time [min]

Experiments on the influence of gradient and slope-length on runoff led to varying results. The trials showed more, less or unaffected runoff volumes with increasing slope. In seven out of eight studies in the US, t'or example. annual runoff volume increased logarithrnical with gradient, whereas slope length had no effect on the amount of runoff per unit area (Wischmeier, 1966). One reason for a positive relation between slope and runoff is the decreasing surface retention with increasing slope comparable to a cup of water which is more and more inclined (Figure 3-7). In contrast to these results in the US. Poesen (1984) measured less runoff with increasing slope on sealing soils. The compaction o f the soil by impacting drops is less because the impacting force does not act perpendicular to the surface and the number of drops per unit area is smallel- (Figure 3-7). Thus, on steep slopes surface sealing is weaker and runoff can be smallel- than on gentle one's. Additionally, the number and depth of rills were higher on the steep slopes. The I-ills dissected the seals and enlarged the infiltrating surface area.

Chapter 3

Figul-c -1-7: Rrrirr ~ y o l u ~ r ~ ~ prr rrrrit area a d surfrrce \toroge r1ecrea.s~ 011 the slr~pe of lerlgtl~ AB with increrrsirlg gradient. Sl?lo.slz is alnrrj,~ trcrrz.\l7ortr~cl further don,t~slopo that1 upslope

rain

I surface /

4 Soil loss determining factors

4.1 Rainfall

One driving force for water erosion is rainfall. The raindrops which pound on the soil surface either infiltrate into the soil or leave the field as surface runoff. The rain volume which runs off on the soil surface not only depends on the properties of the soil, vegetation and topography, which will be discussed in the subsequent chapters, but also on the quantity, distribution and type of rain. Tnvestigations showed that soil loss is largely determined by rain volume, energy load, intensity and their distribution within single storms (Flanagan et al., 1988) and during annual seasons (Lal, 1990).

An example for the last effect was given by Temple (1972) who noted 8 times more runoff from a rain at the end o f the rainy season compared to a similar rain at the beginning of the rainy season.

Kinetic energy (E) of a storin is calculated by (Morgan, 1986):

with m mass of falling rain [kg] v terminal velocity of the falling drops [m/sJ

Terminal velocity of raindrops increases with diameter to a maximum of 9 to 10 mlsec for the largest drops which have diameters of about 6 rnnn (Gunn & Kinzer, 1949; Laws, 1941; Laws & Parsons, 1943). Drop diameter increases with increasing storm intensity up to intensities between 76 and 100 mmlh (Carter et al., 1974; Hudson, 1963). Pressures between 2 and 6 MPa are exerted to the soil for very short times (50 ms) when a rain drop hits the soil surface (Ghadiri & Payne, 1981). This pounding action destroys aggregates, displaces particles (splash erosion) and has a sorting effect which leaves a thin layer o f coarser particles at the soil surface. Thin water layers of 14 to 30%- of the drop diameter in thickness enhance splash erosion whereas thicker layers protect the soil (water mulch) (Mutchler & Young, 1975).

Tropical rains are characterized by high and distinct intensity peaks. Maxirna of up to 800 mmlh are reported from Jamaica (El-Swaify & Dangler,

1982). For northern Nigeria, Kowal & Kassatn ( 1977) measured common peak intensities of 120 to 160 rnmlh and showed that mean drop diameters where higher in tropical storms than in temperate areas. From western Nigeria, intensity peaks of 190 mmlh are reported (Wilkinson, 1975). Peaks occurred during the first five minutes in more than half of the storms. Hudson (1961) measured peak intensities of up to 340 mmlh in southern Africa. The erosivity of storms may additionally be enhanced by strong winds (La1 et al., 1980). In convective storms high windspeeds commonly coincide with intensity peaks (Raussen, 1990).

In order to predict soil erosion, Wischmeier & Smith (1958) found out that the product of a storms total kinetic energy (E) times its maxirnum 30 minute intensity islinearly related to soil loss:

R = ( E :,: 13(,) [Nlh] I = I and E = xn ( 1 1 .X9 + 8.73 logli) r; Pi lo-' [k~lrn'l

I= I

with R longterm mean annual erosivity [Nlh j E kinetic energy (kJIm21

I30 maximum storm intensity during 30 n~ in [mmlh]

I, intensity for storm interval i [mmlh] fot- 0.05 < I < 76.2 mm; for I > 76.2 rnm I = 76.2 mm

P, rainfall volume during interval i [mm] 11 number of storrn intervals with equal intensity [- 1 m numbel- of erosive \toi-111s pcr year 1-1

The R factor of Wischrneier & Stnith ( 1958) has proven appropriate for temperate areas. For tropical Africa, however, several constraints are to be faced. The calculation of reliable R factors depends on daily rainfall records over 22 year periods (Wischmeier & Smith, 1978). The necessal-y subdivision of individual storms into intervals of similar intensity and the I-ecognition of the rnaxitnum 30 min intensity asks for self-recording raingages with low paper feed rates. These data are nor~nally not available fot- a sufficient number of years and meteorological stations. The R factor overestimates large storms which causc only little runoff but underesti~nates small storms with much runoff (Foster et al., 1982; Laflen et al., 1985). Both occur- frequently on tropical soils. Therefore, other authors proposed a

number of different erosivity indices for tropical areas, which were either easier to calculate or which can be better applied to the local conditions. Fournier ( 1962) developed an index for river basins in West Africa:

where P,,,,, is the annual amount of rainfall and Pm,,,,, the rainfall amount during the wettest month. A regression of a modified version of Fournier's index with the R factor was used by F A 0 for the design of an iso-erodent map of Africa north of the equator and the Middle East (Arnoldus, 1978).

For southern Africa, Hudson ( 1 986) reported that only intensities above 1 inchlh (25.4 mmdh) caused significant splash. Therefore, his index KE>I considers only the energy of rain falling at intensities > 25.411im. For the calculation of kinetic energy he used:

with 1 storm intensity [mmlh]

The energy term as calculated by Kowal & Kassam ( 1 977)

with Pi storm volume [ m m 1

described soil loss better than EIjo (Salako et al., 199 1 ).

Delwaulle (1973) si~nplified the calculation of erosivity by substituting rainfall energy by rainfall amount (Pi) and Lal (197hb) additionally used shorter intervals for thc maxiinum intensity (I,,,;,,):

Chapter 4

For I,,,, he chose the maximum 7.5 min intensity. Sabel-Koschella ( 1988) obtained similar results for m values between 5 and 25 min.

Roose ( 1977) evaluated mainly 13 stations in West Africa and found a linear I-egression between the E130i and monsoon type rainfall (Pi) between June and September of:

and a curvi-linear regression for high intensity storms during the rest of the year. As an empirical approach for the estimation of erosivity in West Africa he proposed:

R = (0.85 (+/-) 0.05) :;: Pann [N/h] (10)

Roose (1977) verified his regression for 20 rainfall stations and drew an iso-erodent map of West Africa. Further iw-erodent maps were cornpiled for Zimbabwe, Kenya, Tanzania and Uganda based on KE > I (Moore, 1979; Stocking & Elwell, 1976). An iso-erodent map for Zambia wa5 supplied by Lenvain et al. (19x8) using:

with Pm mean monthly rainfall [cml b mean number of days with rains > lmm [-1 c mean maximum daily rainfall per month [cm]

For the iso-erodent map of South Africa (Smithen & Schulze. 1982) erosivity was estimated by 'effective rainedll', a modified Fournier's Index and a 'burst factor'.

4.2 Soil PI-operties

4.2 Soil properties

The influence of soil properties on soil loss can be ideally studied on runoff plots stripped from all vegetation for some years. Thus, it is assured that no influences of the former vegetation bias the results. Table 42- 1 demonstrate? the influence of soil properties on barefallow soils subject to 1200 mm/a. Soil losses are as low as < It/ha on an Oxisol and as high as 280 t/ha on an Andisol (Nill, 1993).

Tclhle 42-1: Soil lo.v.\ on diferetzr hczrqfnllow .soils c-orrt.c.tt.c/ to an L I I I ~ I U Q /

erosivity of' 800 N/lz ( (q?pr~x. 1200 r?znz/n).

However, the soil properties causing these differences are not evident as mostly a range of soil properties found in different soils and their combination are responsible. The important soil properties decisive for the extent of erosion are listed in Table 42-2.

soil type (US soil taxonmy)

Acrustox

Tropudult Trophum~~lt Hapludult Kandiudalf Paleustalf Andisol

Mineral composition, especially the content of metal oxides, is known to influence soil erodibility. Metal oxides act as binding agents between soil particles, thus increasing structural stability. Soil loss on subsoils decreases with increasing content of Al- and Fe-oxides (Roth et a]., 1974; Romkens et al., 1977). It is supposed that especially the amorphous part of the Fe-oxides is reactive. In experiments of Chauvel et al. (1976) kaolinitic clay mixed with > 5% iron oxides showed a self structuring behaviour (formation of shrinkage cracks) when drying out whereas at iron oxidc contents < 5% it formed a coherent matrix. Only 3% of the total Fe- oxides were actively participating in the aggregation process. Rapidly sealing

mean annual sail loss[t/ha]

0.5

12

20 5 7 8 9

147 280

Soil lo\a was calculated Srorn soil el-odibility values which were adjusted to 800 N/h mean annual erosivity (approx. 1200 m r n f a ) .

Chapter 4

soils generally suffer higher soil losses than non-sealing soils. The type of clay mineral also influences the formation of seals and the infiltration capacity of the soils. Soils rich in smectitic clay (e.g. Vertisols) swell and shrink with varying moisture content. Infiltration is, therefore, high in the dry state while cracks are open. In the moist state these soils become extremely sticky and plastic, cracks are closed and infiltration reaches very small values. Soils rich in kaolinitic-oxidic clay, on the contrary, are well aggregated in the dry and moist state. They are less susceptible to sealing than soils with 2: 1 clays (Levy & van der Watt, 1988; Shainberg et al., 199 1 ). The stable structure of the former enables high infiltration rates.

Table 42-2: Soil properties irzfluerzcirzg soil erosiorz.

soil properties permanent

mineralogy Fe-, Al-oxides

texture soil organic matter

pH and exchangeable cations aggregate stability and size

bulk density electric conductivity of soil water

soil temperature v antecedent rnoi\tur-e

variable

Type and quality of the parent rock act on the texture of the formed soil. For example, sandy soils form from granite whereas clayey soils form from basalt. Soils high in silt and low in clay and sand are highly erodible. Erodibility decreases with ;i decrease in silt, regardless whether the corresponding increase is in the sand or the clay fraction (Wischmeier & Manncring, 1969). The high erodibility of silty soils is explained by their weak structural stability. They rapidly form surfice seals upon raindrop impact. Erosion is less on clayey soils due to their better aggregation and on sandy soils due to their- non-sealing surface. Fine sand (0.05-0.1 Inn] diameter), however, behaves like silt and is therefore attributed to the silt fraction for soil erosion aspects (Wischmeier 8r Smith, 3978).

4.2 Soil propertie\

Soil organic matter (SOM) influences soil loss by improving soil structure. root penetration, water capacity and infiltration. With increasing SOM, el-edibility decreases (Wischmeier & Smith, 1978). SOM consi\t\ of very heterogeniou\ particle\ ranging in size between several Inn1 down to < 0.002 mln. Chelnically very reactive organic molecule\ compare with more inac t i~e one'\ and re\i\tant component\ with rapid decompo\ing one'\. The role of SOM as a binding agent i \ more important o n \oil\ deficient of other structuring coriiponents. Thei-ef'ore, the importance of SOM decreases with increasing clay content (Wischmeier & Mannering, 1969). Valentin 24 Janc~tu ( 1989) found that structural stability was only improved by organic matter if the ratio of organic matter to clay was 2 0.07. In tropical cr-opping systems SOM is high after the fallow ancl declines I-apidly during the cl-opping per-ioct. Tlii~s. erodibility changes during ;I cropping cycle from low values during the trtllow anti at (he t)eginlling of ci~ltiv;~tiori to higher- values towards the end of' culti~,ation. In own trials the erodibility during the Sirst year of bar-cfallow after bush and I'orcst filllows was only 40 CX ant1 80 (2 ;:. respectively. of the final erodibility which was rcaclicd al'ter about 3 years of bar-efallow (Nill. 1903).

Agy-cgate \17e and \lability have a pc~-~iianent and a Larlable coliij7onent, the latter of which retlect4. alllong other influences. vegctatioli allcl ~nanagcliiont. Eroclibrl lty dccrca\cs it11 lncl-ca\ing aggregate stitb~lr ty as \cal formation i \ clelayed and ~nl'iltrat~on incrca\ed. H o ~ c v c r . the effect ol aggregate \ i lc i \ le\\ clear. Mo\tly \oil lo\\ wa\ found to becorne \ln;~lle~- L+ ~ t h incre:~\ing aggl-egatc (Ehwue. 199 I ; Falayi KL LaI, 1979) for. , I Y Y I - ~ ~ , I ~ ~ C L cliumctcrs hetwccrl 0.5 and 5 0 mm. 1,uh ( 1083) tcstcd aggrcgatc cla\\c\ betueen 0.5 and 30 mm and tounci higher splash and sheet cros~on I'rorli larger than frolii smaller agg~-cgate\. Wi\chmeier Kc Smith ( 1978) al\o attributed Il~pliet erodibilitte5 to larger aggregate\. Howevcl-, Aniba\\:i-Klhl & La1 ( 1093,) only Sound a \oil lo\\ decl-ea\c up to 10 rnm aggl-egatc diameter-. For :~ggrcgate\ between 10 and I00 rnm 110 effect wa5 rnea\u~-cd.

The el'l'ect ot the cxcliange:tble ca t~on\ i \ e\pecially Important on Ic\\ \beathered \o11\ ol' t l~c {emi-arid to arid tropics. Tllc$e \oil\ are wtxthlq \trircli~~-ed cli~e to low SUM ;uid oxide co~ltents and have often \ancly 10 loamy ~c\tul.e. Na. a\ a monobalent c ~ ~ t i o n , ha\ a pronounced d~\pcr\ ing influence o n \oil \t~.ucturc. 3 to 5(/( 01' N a o n the exchange complex arc enough to cl~\per.\c the \oil (Shainberg. 1985) ant1 crodibility incrcam with irlc~.ea\iny Na content (S~ngcl- et al, 1980). Mg \aturated \oil\ wcr-c Ihund to be mor-e el-od~hle than C'a sati1i.atc.d soils cau\cci by the lal-gel. hydl-atton shell 01' the

Chapter 4

Mg ion which weakens bonds between soil particles. The stronger aggregation in the presence of exchangeable aluniiniurn explains the higher stability of acid soils. The electrolyte composition of the soil solution also exerts an influence on soil loss through tlocculationldispersion effects. For example, saline soils rapidly disperse after dilution of the soil solution at the on-set of rain.

On previously rnoist soil runoff starts earlier and reaches higher runoff volurnes than on initially dry soil. For this reason, rains occurring at the on-set of the rainy season generally cause less runoff and soil loss than rains at the end of the rainy season. Not much data are available about the influence of soil and water temperature on soil loss. With increasing temperature water viscosity decreases. Aggregate destruction caused by the pressure of encapsulated air during I-apid wetting of the aggregates is enhanced if the wetting velocity is higher. Water rnulch by less viscose water (= ,,more liquid") will be less protective against raindrop impact. Auerswald ( 1992) explained a soil loss difference of 17 % between artificial rain applied in the morning and in the afternoon with a temperature difference of 8 "C.

4.3 Topography

Topography intluences soil loss by the length, gradient and shape of a slope. Soil loss increases very sensitively with gradient and commences already on slopes < 1 9%. Mutchler & Greer (1980) measured losses up to 5 tlha from dry <oil and up to 1 1 tlha from wet soil on a 0.2% slope when a simulated 60 min storrn was applied. In Senegal, annual losses from groundnut fields on a 1 C / r 1

slope reached 15 tlha (Fournier, 1967).

Uncertainties arise, however, where the influence of gradient has to bc quantified. Most studies propose an equation for soil loss tiom interrill areas of the form

with A soil loss S gradient a, b constants

4.3 Topography

Values for b between 1.35 to 2 were suggested (Hudson, 1986; Hudson & Jackson, 1959; Musgrave, 1947; Zingg, 1940). A value of b = 0.67 was suggested for soil 105s from rills (van Liew & Saxton, 1983). In the USLE the influence of gradient is described by

S = (65.41 " sin2 u + 4.56 * sin u + 0.065 1-1 with u slope gradient [degrees]

More recent analysis of slopelsoil loss data revealed a change of the relationship at > 9 % slope (McCool et al., 1987). Soil loss for very low slopes was found to be overestimated by the LS factor (Murphee & Mutchler, 198 1 ). Runoff on low slopes flows slowly and quickly forms a water layer deep enough to act as surface mulch. It further became apparent that soil loss depends on the ratio of rill to interrill erosion. Soil loss is higher on soils very susceptible to rilling (McCool et al., 1989) and the potential for rilling is greater on steep slopes (Mutchler & Greer, 1980). S factors for the Revised Universal Soil Loss Equation (RUSLE) are, therefore, calculated by:

with u slope gradient [%]

Equation 14 and 15 are used for slopes > 4.6 m long and gradients of < 9 5% and > 9%1, respectively. On slopes < 4.6 m long rill erosion is negligible on most soils and equation 16 is to be used.

Increasing slope length enhances soil loss as more runoff can accu~nulate on long slopes. For slope length, the following term is used (Wischmeier & Smith, 1978):

with 1 slope length [m] rn slope length exponent I-]

The product L S is called the topography or LS factor. The LS factor wa5 derived from $011 lo\\ data of \lopes ranging from 3 to 18 '/r and 9 to 90m (30 to 300ft) long (W~whrneier & Smith, 1978). Beyond the\e range\ n o mea\urement\ were taken. However, the equation wa\ regarded applicable by the author\ to \lope\ 300 m long and 5 0 % steep (cf. Chapter 7.3). Foster et al. (1982) est~rnated that the 1,s factor can be appl~ed in the trop~c\ to slopes up to 25 C/c whereas Hurni ( 1980) u\ed the LS factor tor \lope\ > 50 C/c . The LS Eactor wa\ verified In West Africa on \lopes between 3.5 and 23.3 (/c (Rome & Sarrailh, 1989) wherea\ Shcng (1990) reported an overe\t~matlon of \oil lo\\ by the LS factor on 30 54 slope\.

The effect of slope length on soil loss is interrelated with slope steepness. This is expressed in the slope length exponent 111 of the LS factor which is 0.5 for slopes > 5 'X and decreases to m = 0.15 for s l o p e s 2 0.5 54 (uf. Chapter 7.3). In the earlier development of the ecluation, an exponent of 1.6 was used (Zingg, 1940). Dangler & El-Swaify (1976) reportod an underestimation of soil loss by the L ['actor as used by the USLE. However. on soils from West Africa Lo,? was found to give better results (Rooso 24 Sarrailh. 1989). In the RUSLE, m varies between 0.02 and 0.83 depending o n rhe soils susceptibility to silting (McCool et al.. 1992).

Soil lo\\ i \ al\o ~nflucnced by the (hape of a \lope. It dccrea\e\ In the orcler convex > regular > concave \lope l'or~ii. On a convex \lope. whcrc the gradlent increa\e\ in the order up-slope < mid-\lope < down-(lope. a large runoff volu~i~e coincide\ with the maximum gradient (down-\lope). 0 1 1

the contrary. on a concaLe \lope the m a x i m ~ ~ ~ i i gradient I \ up-\lope uhcrc runol'f I \ \ t i l l \~rialler.

4.4 Cover, tillage and protection techniques

Cover. tillage and protection techniques depend on management. i n contrast to rain erosivity, soil erodibility and slope. This r~litkcs them of fol-emost irnportat~cc to soil conservation. Managcmcnt pl-actices can be dictinguishcd according to the basic erosion processes that they inrlucnce:

I . Methods which reduce the runoff' volume or the sediment transport capacity

All ~llethod\ which incrua\c water inl'iltration o r reduce runoff vclocit y al\o seduce soil lo\$. I f ri11101'1' i \ \lowed down. the wares \ray\ longer on the t'icld

4 4 ( ' o ~ c r . t l l l i~ge and protcctlon t c c h n ~ q i ~ c \

and gc15 tnol-e time to infiltrate. Additionally, the water layer on the \oil \urf'ace becomes deeper and protect\ the \oil from raindrop impact (cf. Chapter 3).

These methods include:

D reduced tillage D no-tillage Ci tillage and planting acres\ the slope (contouring) C> soil cover by inorganic or organic mulch C> surface fhl-nning practice\ (ridging, tied-ridging, bedding)

11. Methods which reduce the slope length

Thereby, the ~-unoff procluc~ng up-\lope area and I-unoff volume are I-educccl. So11 tnaq \till be tr-an\pol-(ed bill phy\ical ob\tacle\ divcrt r-unol'f andlor cauw depo\ition. Thi\ group includes:

P hillside-ditches [> filter-strips with grasses, hedges or tree rows D earthen and stone bunds I> terraces

The single methods mentioned above are described and discussed in the following.

4.4.1 Reduced tillage and notillage

Tillage breaks down soil ugpt.cyiites. disturbs soil structurc, pore continuity. and biological activity and produces transportable soil material. Reducing tillage intensity and frequency increases the number of continuous pores. maintains :iggrepation and reduces organic tnatter decomposition. Thereby. binding agents between soil pal-ticles like fine roots, fungal hyph:~e. root and bacterial exudates are conserved. Thc soil stays more consolidated as colnparcd to tilletl soil. Thus, infiltratio~l and resistance against impacting drops and shearing forces of rhe water are higher than on tilled soil.

Chapter 4

4.4.2 Contouring

Contouring necessitates that all tillage and planting operations are carried out across the slope. These operations produce a low surface relief across the slope. Runoff is slowed down and the surface storage is increased. With increasing gradient, the surface storage capacity decreases (cf. Figure 3-7) and the risk of spilling over with consequent rill formation increases. Therefore, the efficiency of contouring reaches a maximum on slopes between 3 to 8% (Wischmeier & Smith, 1978).

However, the effect of contouring is uncertain in handtillage systems where the soil is tilled from the bottom of a field moving up-slope and were a general down-slope movement of the soil from hoeing can be observed. Thus, tillage in such systems is not on contour in the strict sense. Only contour planting can be achieved.

4.4.3 Soil cover by organic mulch

Surface cover is one of the most efficient measures for soil loss reduction. Organic material is easily available in areas with sufficient rain. Residues of the previous crop, weeds and additional mulch material from outside the l'icld can be used (leaves; twigs fro111 bushes, hedges; straw; wood cuttings; organic household waste like peelings, shells and husks). Left at the soil surface, they protect the soil against the pounding drops and prevent seal f'orniation. The stalks and leaves form barriers where the water ponds (= water mulch). Runoff is slowed down due to twisted pathways. Additionally, residues and mulch reduce variations in soil ternperatul-e, humidity and thereby biological activity and improve structure and infiltration. Earthworms which move to the soil surface in order to pick up food create large continuous pores. Mulch efficiency depends o n the surface area covered by the material (cf. Figirre 74-2).

Incorporation of the residues diminishes the coverage. The deeper the residues are incorporated into the soil, the less protection they offer against erosion. The efficiency of surface mulch may also be reduced on soils with extre~nely unstable structure ( e .g soils with high sodium saturation , or hard setting soils) where aggregates already disperse on moistening . On such soils superficial incorporation may provide higher infiltration rates because the

A t e t can be can-ied OLII hy \ubrnel-ging dry aggregate4 o r fragrncntx of 1-2cm diameter in water. Unstable aggregates break down immediately

3.4 Cover. tillage and protection tcchniqi~cs

decomposition products of the residues stabilize soil aggregates and the residues act as stabilizing framework.

4.4.4 Inorganic mulch

Several forms of plastic foils are used in intensive agriculture and gardening to protect the soil, reduce evaporation and suppress weed growth. Howevcr, these mulches play a marginal role in s~nall scale agriculture. An important natural mulch tnaterial are stones and gravel of various sizes. Stone pavements and surficial gravel concentrations on soil surfaces frequently indicate erosion processes. The gravel was enriched at the surface by the selective removal of the soil. With increasing cover of the surface. the soil i11lde1-neath becomes protected. Howevcr, the active use of gravel as rnulch ~natel-ial deserves much more attention than rccently given. In highly leacheti soils o f the hi1111id tropics, the use of gravel from basic rocks may additionally deliver sorne nutrients.

4.4.5 Surface forming practices (contour ridging, heaping, tied ridging, bedding etc.)

The\e method\ create physical ob\tacle\ (Figure 445- 1 ) which reduce e\pecially the slope length. Runoff is \lowed down. \topped or deviated \idew:ly$ on a reduced gradient. Alike contouring. the protective effect of thew methods depends on the gradient of the \lope. the \ide \lope and he~ght ol' the ob\tacle\ and their distance frorn onc another. Contour ridges are \mall earthen dam\ of about I0 to 30~117 height placed across a \lope.

bed

The protection by ridges is low on very low slopes hecause soil loss is genet-ally low. On stccp slopcs. i t is low because the amount of' water which can be retained by a ridge decreases with increasing gradient. A 15 urn high ridge does not store any water on slopes >25 C / c (Foster et al. 1992). At the same time the risk ol' spilling over- and ol' break throughs in the ridges is enhanced. Thus, maximum efficiency is obtained on medium slopes (cl'. Figure 75 1 - 1 ). The eff'iciency of' ridges also depends on ridge height and the side slope. The higher the ridge, the more water can be stored. The lower the sidc slope 01' the I-idgc, the slower the runoi'f. Meyer- & Harrii(>n (1985) \howeci that on side slopes < 0.5 C/c the suspended sedinlent in the runoff is cleposited and most of' the setiiment originates from the ridge-sides. Above 254 sidc slope deposition ceases and the sediment is moved out of the field. With sick slopes of' 5-6 % rilling of'thc I'LI~I-ows commenced.

The ci'lkct of' l.iin-ows also depends o n stol-m size. Large storms may \uspass the carrying capacity of' the I'LII-rows and cause overflowing of the I-idge top. Overflowing Inay cause very severe damages and should be a\,oided in any case. Thus. efficiency is less for areas with frequent lar-ge storms. The 10 year storm which is the largest. I-egularly occurring sto1.m within 10 year-s. can be i~scd to calculate the efficiency of contour ridging for an area.

A\ the ridge-sides act as runoff' producing area for the furrows. the ri~nol't' ~ x o d i ~ c i n g area increase\ with the length of the furrow\ and so doe\ the ri1t101'1 volutne. In order to avoid overflowing of' the ridge top\ 01- rilling. f'i~rrow length {hould be limited.

Not muc11 i \ hnown about the effect ol' ridges placed along the slope. Mea\urement\ on \lopes of 7 t o 13 Cjc indicate an ero5ion enhancing infl~~erice ( P f.ictor\" between 1 and 6). On \lope\ o f 13 to 20 fZ the ncgativc effect was Ic\+ pronounced ( P = 0.31 to 3.4) but \till important (Reining. 1091 ).

Heaps of' v;~ryirrg sizes are frequently used especially for tubers but also for other crops like groundnut or bambara nut. The loose. rich topsoil usecl for the heaps is I'avourable hi- tuber fon?iation. Watcr logging is ~ w c ~ i t c d by the heaps and mincrali7,ation is enhanced. However. not moch is known about the effect of heaping on soil erosion. Unfortunately. sizc and LII-rangenlent of heaps on a slope arc inostly nut described in literature.

A P Factor < I indicates less erosion compared to a harcfallow I'ield wherca\ P > I 1nclic.aL~4 n o ~wotection. Morc int'ormation i 4 given in ( ' l lapt~r 7.5

Heaps on a slope A -r C2 enlarge the average gradient a by the gradient U-cx on7the sidewalls ( F i g ~ ~ r e 445-2). Taking the r i laxi~n~~rn angle of about 40" (a+B) into account which forms if topsoil is poured on a heap. the actual gradient of the slope is changed on the heaps into a gradient of about 80 C/r (= 40') on the sidewalls. This also implies that the heaps become lower with increasing slope as demonstrated by heap height (h l+h2) for heap ABC I on level ground compared to h2 for the heap ABC2 in Figure 445-1. Runoff from the heaps enters into the furrows aniong the hcaps and mo\,es

downward in a concentrated flow. Runoff volume depends on the size and arrangements of the heaps. Compared to small heaps. largc heaps have a larger runoff producing sidewall area and less drainage paths between the heaps. If the heaps form furrows along the slope, rapid water movement results. If they are arranged in quintuples, the water has t o flow around the heaps and a reduction of the flow velocity can be assumed. If the heaps are arranged up and down-slope, runoff is directed straight down-slope and will reach a higher speed.

' The m~iximum alopc anglc fol. I~capcJ LIP aoil (angle of I-cpocc depend\ on the particle \ i /c di\tribution. Surface \oil < 3 0 m m = 38.7": aggrcgate~ 1-2 Inrn = 40.2' : <urlacc \oil sietetl to < ?mnr = 37.6" ( i2 i ic l -~ alcl. 1903 ).

4.3 Cover. t l l la fe and protectton (echn~yucs

The influence of heaps on soil loss further depends on surface soil depth. On soils with a thin surface horizon, all soil is scraped together for the heaps. The underlying soil which can have very different properties is exposed. A pattern of very different soil erodibilities is created this way which can include, for example, a less erodible surface soil on the very steep side-walls of the heaps and an erodible subsoil between the heaps. Overall soil loss should be notably increased in this case.

4.4.6 Bufferstrips

Bufferstrips (filterstrips) are < I ni to several m large strips of' planted grasses, hedges or natural vegetation on contour. They slow runoff down and maintain higher infiltration rates within the strips as co~npared to the adjacent field. If runoff occurs, soil is transported within the cropped area and deposited in the vegetated buf'ferstrip. The runoff either infiltrates completely in a bufferstrip or crosses the strip. If all runoff infiltrates, all transportecl sediment is deposited. If part of the runoff passes through the strip only a part of the sedirnent is deposited. First, the coarser, heavier sand particles and aggregates settle whereas the srnaII particles of clay and organic matter use furtl~er transportecl and may leave the strip on its lower side. Thus. quantitatively a large part of' the sediment can be retained by a bufferstrip while an irilportant amount of' fertile soil is still lost. Cornpared to temperate soils. this is more relevant on leached tropical soils. Their low cation exchange capacity (CEC) is largely associatctl with the organic matter. Another- inconvenience of strips with incoriiplete infiltration is that the water which leaves on the lower side can speed up again and pick u p new sedirncnl.

1 to 4 n~ large buffer\trip\ on 4-20 C?c slope\ can reduce soil lo\\ by 10 to 00%. Strip\ w ~ t h natural fallow vegctatioi~ can already be \pared out when cultivating the field. Compared to i ~ ~ c h 4tr-ip4, planted $trip$ have a lower efficiency in the I st yeas. Efficiency decline< after an opt~murn due to increasing \ed~rnentation in the strip\. A 40 rn wide bufferctrip for example dropped from 99 to 75 cji, el'f~ciency during 9 month\ (Barfield & Albrecht. 1982). In agreement with other authors. Sc t ia~~der & Auer~willd ( N 9 2 ) could \how that a 30m wide grass stw on a 8 54 slope retained 64(L of the sediment which entered the strip (ca. It/m strip width). The efficiency increaced with incl-ea\ing \t~-ip width (cf. Figure 752- 1 ).

Bufferstrips are Inore acceptable to fanner\ if they give solne yield. Introduction of fruit tree\ or woody specie$ may be of more interest than PLISC

gra\\ strips and can encourage farmer\ t o protect the strips :tgain\t t ~ r c j . Some conimon g1.21\\ and tree species u\ed for bufSer\tripc and biological control are li\ted in Table 42- 1 Annex. Ari extensive databanh on sutted

'r

woody spec~e \ is ~i\ailablc from ICKAFJ Nairobi .

4.4.7 Contour bunds (stone hunds, earthen hunds, diguettes)

Buncl\ ai-e a form of Iiigli ridges which arc mounted at a distance o f \evcral m from one anotlicl- (Figure 447- 1 ). They can be constructed fi-om \oil or \tone\ or both co~nbined (earthen core wltli stone\ o n the outside) (Figure 447-2). Bund5 are il\ed t o control ero\ion and to conserve water. Impermeable bund\ (I'rom earth or with ii1I earthen core) stop runoff completely and direct it \idcways. They are ~i iorc rigid in tlicir action and are to be con\tructed more \olicl than pcrriieable bund\. Waterlogging in front of' impermeable bunds can be a problem for 4ensitlvc crop\.

Runoff hi t t~ng a permeable bund (e.g. \tone bund) i\ \lowed down, \lowly penetrate\ the bund and leave\ partially on its lower side. Reduced celocity al\o f a ~ o u r 4 infiltration on thc lower \ide. However, the runot't' may regain velocity and pich up new sediment. Thus, the ~trca below a bund may be eroded wlierea\ thc area in front of thc butids is wdiment-enriched. Therefore. \mall terrace\ fonii after a couple of years.

Bi~ricls ;ise recotiiii~ended o n slope\ of less than 12 % (Table 449- 1 but are

also u\ed on steeper slopes with \ucce\\. In order to diminish maintenance mot-k. the bundi ihould be planted to permanent vegetation (gra\\es, woody \pecie\, tree\ ).

' h1ultipur~x";e Trcr K: Shrub Datahasc (ca. 110.- \IS$) IC'IIAF. P.O. Box 30077 Nail.ohi. Kenya

4.1 Co\ el-. rlllagc ;111tl protectloll ~ c c h r ~ ~ c l ~ ~ c \

permeable bund

semi permeable bund

~mpermeable bund

Dykes are a large version of bunds which is especially used in semi- arid to arid areas to store water and to slow down torrential floods. They are the transition to the even larger darns. Alike bunds, dykes are constructed as permeable (digues filtrantes) or impermeable obstacles (digues dkversantes). Between the two extremes there are a couple of intermediate solutions with impermeable lower and permeable upper parts (Figure 447-2). In the first case, the water is slowed down and ~nomentarily stored while it percolates through the dyke. In the second case, the water is stopped and stored behind the dyke. The water quantity exceeding the dykes storage capacity either flows over the top of the dyke or is conveyed by a weir or spillway. Behind the dyke, water infiltration is increased and sediment deposited (Figure 447- 3). The deposits are either used for irrigated cultivation in the border zone while the water is retreating or for a crop which uses the water stored in the soil. Clayey deposits serve for brick construction.

SHED

4.4.8 Protective ditches

Several types of ditches can be established on contour to slow down runoff and collect eroded sediment. Drainage ditches are constructed by disposing the excavated soil down-slope of thc ditch (Figure 448-1 ). The ditches are laid out on contour or with a slight side-slope of 0.4 to 0.5% (see Hi~dson. 1975for planning principles). Drainage ditches reduce slope length into several segmenls. The down-slope concentration of runoff is thereby avoided. The sediment spilled into the ditches nceds regular excavation. Maintenance efforts are therefore high.

The Fanya Juu terraces are a modit'ied version of dl-ainage ditches especially used in East Africa (Figure 448-1 ). The excavated soil is disposed up-slope thereby forming an earthen bund which traps further sediment. Hillside ditches (Figure 448-2) are a form of reverse slope or level bench terrace. The bench is generally not used for cultivation but as foot path or road.

tr rrac7e.r

drainage ditches

Riser protected with grass

Hillside ditch

( foot path) Tall grass .I

Cultivation is carried out on the graded interterrace area along with further protection measures. Hillside ditchc\ may be used for slope\ of up to 47% (Sheng, 1989). They divide the slope into \liorter segments and divert runoff at non-erosive velocity.

A version of the hillside ditch is the broad based terrace used in mechanized agriculture. It can only be used on gentle slopes. The interterrace space and the terrace interval on the graded terrace is used for cultivation. The terrace is kept as low as possible in order to allow the passage with I'arm equipment. Figure 448-3 shows how these measures are laid out in a watershed or farm. The runoff collected by the ditches must be disposed saf'ely by constructed waterways or by conveying i t into densely vegetated areas.

4.4.9 Terraces

Terrace\ are described by a number of characteristic\. Important features are the vertical height. the horizontal length, the ratio of the raiser b/a, the revert

\lope a, the side slope. the terrace interval and the interterrace interval (Figill-e 449- 1 ). They are used for a range of purpo\ec a\:

D to divide a slope into shorter segments D to reduce the slope angle on the terrace interval I> to convey the surfilcc runoff to controlled water ways at a non-erosivc

velocity L> to harvest water from intertcrrace intervals for water conservation D t o store water for paddy cultivation b to store sediment eroded l'rom the interterrace interval

A number of different terrace type\ was developed to cope with these tasks. Most oc them can be described as riiodified bench terrace4 (Figure 440-2 ).

The level bench terrace has a wide-spread use f'or paddy cultivation whereas the reverse slope ancl outward slope bench terrace are favourablc for upland crops of the hurnid tropics. The conservation bench terrace is used in semi-arid to arid areas for water harvesting. B5nch terraces are used on slopes between 12 and 50C/r and are built by hand, animal drawn equipment or machines.

7- i- y b horizontal length j + '

~nterterrace :, terrace -&- Y

interval ~nterval

1 L E V E L BENCti T E R R A C E S 3 CONSERVATION BENCH T E R R A C E S

Runof f C o n t r ~ b u t ~ n g Area ~ O r ~ q ~ n o l Land Sur face

O ~ , Q , ~ O I Land Sur face '---qv~ .. , . ee$ D Y k e .-\, . q,+ ~ e v e l Bench

- . . . .

2 OUTWARD SLOPING TERRACES 4 REVERSE SLOPING TERRACES

Original Land Surface

Toe O r o . Toe O r o ~ n

4.4 Cover. t~llage and pl-otectlon techn~cliie~

The intermittent terrace is used if terracing is not completely carried out for the entire slope (Figure 449-3). The design is carried out as for bench terraces but only every 3rd terrace is constructed. This leaves the option to later construct further terraces in between which gradually transforms the intermittent terraces into bench terraces.

Orchard terraces are used for tree plantations 011 very steep slopes in order to facilitate access and maintenance (Figure 449-4). The terrace interval is not planted to trees but stabilized by grasses. The distance between the orchard terraces is determined by the spacing of the trees which are planted in the interterrace interval. In combination with orchard terraces individual basins can be used to plant the trees in the interterrace interval (Figure449-4). The individual basins prevent erosion and loss of fertilizer and herbicides. They conserve water especially if mulched.

The area between the individual basins is vegetated.

Chapter 4

Some characteristics and applications s u ~ n m a r i ~ e d by Sheng ( 1989) are listed in Table 449- 1 .

terrace type width of terrace interval reverse dope land slope

[mJ L%l 1941 bench terraces (hand made) 2.5-5.0 5 12-47

intermittent terraces 2.5-5.0 5 13-17

natural terraces

(caused by bunds) 8 -20 < 13

orchard terraces 1.8 I0 37 -58

individual basins 1 5 ~ - o u n c l 10 < 58

hilhide ditches 1.8 2 0 I0 < 17

4.4 Cover. tillage and protection tcchniclue:,

The length of the terraces is generally < 100 m and a side slope of l CTc is proposed. The terrace interval depends on slope and soil depth. The gentler the slope and the deeper the soil, the larger the terrace interval. The slope of the raiser is built with a ratio of 0.75 : I .

Indicators for soil erosion

Erosion leaves finger prints which also give information on the type and intensity of the processes. Some of these finger prints are dramatic and hardly to be overlooked while others are less distinct and hidden. Such parameters and finger prints can be studied and provide useful information in a first survey on the general erosion risk.

The age of a landscape indicates its erosion susceptibility. Old landscapes are characterized by gentle slopes, plateaus and plains whereas young landscapes show a rugged relief with steep slopes and deeply incised valleys (Roose, 1975) resulting in higher erosion potential. Long periods of 'normal' or 'geological' erosion cause a lowering of the landscape, the final form of which is a peneplain. A peneplain is characterized by a low and gently undulating relief (Figure 5- 1 ). However, often the process is

J

interrupted by tectonic upheaval or tilting of a landscape. The base level is lowered and a new erosion cycle starts. Tectonic movements and several erosion cycles create a landscape of plains at different altitudes ( pediplains ) which are separated by distinct scarps (Figure 5- 1 ). The oldest surface corresponds lo the highest surface. Remnants of the older surface were separated from the faster eroding pediplains and fhrrn isolated steep hills (inselbergs) or plateaus on the pediplain below which occur frequently in the savannah areas of West and East Africa (Thornbury, 1985). These remnants were maintained because they were resistant to erosion. An examplc ~I-OITI

Camcroon shows how pediplanation has created four levels during > 60 million years (Figure 5 - 2 ) (Segalen, 1967).

Gully erosion leave\ very striking features in the landscape and destroys agricultural land. In Lesotho, for example. 11 is estimated that 3% ot

the arable area is occupied by gullies (Wenner. 1989). Gullies vary greatly in slze. They are defined us deep enough in order that crossing is impo5\ible with agricultural machines while rills can still be closed by ordin;~ry tillage 111ethods (Hudson, 1986).

lowcct point of thc landwape to which thc water c;un Ilo\v

58

5 Indicators lor soil erosion

Figlo-e 5-1: Corzc,e/'t~ of rr ~ P I I C I ) / L I ~ I I ~l'itll ,yor~tIe, ~4rld~llr1tirlg relief' ( u / > o ~ v ) r111rI ( 1 /?e(iipl~lirl ~ ' i t l l LI d i~ f i11~ ' t S C . L I ~ ) hct~lrerl t\tlo lc.~~el\ (c!ftet- T / I ~ ~ I Z ~ ~ L ~ Y - \ - , / 98.5)

Chapter 5

k'iglrro 5-2: Podiylnir~s crt citflkl-erzt n1titudtl.s I(.\ fhl-rllecl irz Cnnzc~r-ool~ h?' ~llor-tj tlllrrl 60 r~zilliol~ !,etrl-,\ c!f'rr-o,siol~ (after- Scgnlc7t1, 1967)

Altitude

>1500m

900-1500 600 - 900

< 600 x x

X i.

o Maroua 2 ++

++ * x x X X , ' ~ x ~ ~ ) ( * x :+*

AFRICAN REPUBLIC

Large gullies reach several tens of meters deep and wide and several kilometers long. Gully incision starts were large runoff volumes are concentrated into linear llow. Possible sites are runoff convergence points of several fields or spillways from roads (Moeyersons, 1989). The water from thc inipcrmeablc road surface collects i n the road ditch and, instead of entci-ing in ii~tervals into a reinforced evacuation ditch, is often led into the aclsj:iccnt area where i t triggers gully forination. Lowering of the base level or large storms which coincide with a sparsely vegetated soil can also initiate gully formation (Oostwoud Wijdencs & Bryan, 1 99 1 ).

Gully l'ormation is facilitated on soils with a coarse textured surface soil underlain by clay-rich subsoils (Lal, 1992). Concentrated lateral subsur-face tlow creates subsurface pipes which in turn can start gully erosion (Firth & Whitlow, 1991). Once a gully is initiated, it is enlarged by regressive gully head cutting along with undercutting and collapse of the side-walls. The gully head moves more and rnorc up-slope and secondary gully branches forrn. Soil cracks form parallel to the gully side-walls. Surface water enters the cracks which increases pore water pressure and decreases soil coherence thus destabilizing the side-walls. It was demonstrated on sodic soils that gully

Chapter- 5

head advance wa5 determined by rainfiill, antecedent \oil moi\ture, headcut height (plunge-pool effect) and runoff contributing area (Stocking. 1981 ). Gully development depends on the depth of the weathered layer and the water\hed area. Gully inci\ion stops if the wlid rock ~lnderneath is reached. Vegetation which forms during less erosive years can also \top further gully enlargement. However, thi5 may only be temporarily. Heavy \term\ or man- made damage to the protective vegetation can reactivate the gully. I f the runoff producing area becomes \mailer with progressive head cut regre\sion, a mature stage o f the gully is reached (Figure 5-3) (Hoebl~ch, 1992). Gully reclamation is laboriou\ and costly. It is more eff~cient to avoid concentrated flow than to protect the \oils against its damaging effect.

Landslides are another form of easy recognizable down-slope soil transport which causes disasters. 2 1.000 people were killed in 1970 in Peru. when an earthquake st;irted the mover-nent of 25 million m3 o f earth which destroyed two entire towns (Schustcr, 1978). L,andslides occur if the weight of u sloping soil rnass cxceeds the shear strength of the soil. Cracks occur o n the upper side of the soil mass and the soil slumps down-slope along a weakness plane. Such weakness planes within a soil or geologic substrate c;ui be duc to different layers of the soil (e.g. permeable layers over less pcnneable layers) or natul-a1 layering of the geologic substrate (e.g. schist). I~nbalances are caused either by increasing the soil weight on the slope (c.,g. construction. water saturation) or the instability of the weakness planes (undercutting by roadcuts, water saturation). Landslides are classed according to material (soil, stone). humidity (e.g. ~ i~ i~df lows) , the type and direction of the movement and its velocity (c .g . cr-cep, flow). An example of a translational and a block slide is given in Figure 5-4.

Majot determ~nant\ tor land\licles are slope, climate, geologj, layering and hydrologic properties, seismic activity, vegetation and human activities (Ga\\er & Zobl\ch, 1988). Mocyer\on\ (1989) reported that landslidec in Rwanda occur especially on elope\ > 58C/( and 011 5chl\t whereas C ~ I I sand ctonc and quartzilic rocks no \ l~de \ were ob\erved. Slldc\ on \lope\ < 5 8 % oonly occurred 11. road con\truct~on cauced \lope in\tabll~ty. Landslide\ were more frequent on \lope\ > 62% in Uganda (Temple & Rapp. 1972) and on \lopes > 53 C / c in New Zealand (O'Laughlin. 198 1 ).

5 Ind ica~ors for boil ero\ion

Figlrrv 5-4: Lrft: bloc-kslicle; I-igllt: tr-crr~.\ I~rtioilcrl 5lidc (ir7: Grr.\,\c>t- & Ziihi\c.lz ( 1988); ~ ( f t c r Griggs & Gilc81zhir.c.t (1 977) N I Z C / SL' /ICIL~PI- ( 197-5))

Landslide frequency is determined by a rainfill duration-intensity threshold which varies due to geology and climate (Figure 5 - 5 ) (Larsen & Simon, 1993). Long duration, low intensity rains cause deeper landslides on volcano-clastic material in Puerto Rico whereas short and intense storms causc shallow slumps (Larsen 8( Simon, 1993). Vegetation decreases landslide frecluc~icy. Plant roots increase the shear strength of the soil. A perm"nial vegetation consu~i~es an iniportant part o f the rain as interception water and f'or transpir:ition. Thus, a vegetarcd soil is drier than an ~~nvegetated soil. These influences of vegetation outrule the physical weight of the vegetation and the weight increase of' the vegetated soil due to incl-easctl infiltration.

An inclication for soil creep is the 'sabre growth' o f trecs. Fl'hc slowly down-slope moving soil inclines the trces which in turn redirect thcir growth upwiu-ds. The result is a trunk forln which resembles a sabre (Figure 5-6) .

Construction of terraccs and roads is ot'ten at the origin of' Ir~ndslides. The c ~ ~ t s weaker1 the slopc stability or locally increase watel- inf'iltration (e .g on the foot of reverse sloped terraces) which changes soil coherence. A similar influence is exerted by overgr;lzing of land. The grass cover is locally destroyed and livestock paths cut the slopc and clestabili7,c i t (Wenncr, 1989).

Chapter 5

Figurcl 5-5: Critiool ~1lrlucs of' \torin d,rrlrtio~, rrnd i/lturritj- ,for i / ~ c r e o ~ i / ~ j i lrrr~rlslide ~frer/rrc/zcy (Wilson et ol., 1992; L n , ~ r / l & Si/lloir. 1993; Ccrirlr, 1980) (fi-om Larserl & Simorl, 1993)

storm duration [h]

Another informative source for erosion susceptibility is the geologic map. In-situ soils are formed from the parent material underneath. Areas with parent materials which form medium to light textured soils are more endangered by erosion than those with materials that generate clayey or very sandy soils. A tentative classification for different parent materials is given in Table 5- 1.

5 Indicator for \oil ero\lon

T~rblc) 5 - 1: Soil or-oclibilitj, c!f'.soils cleri~!ecl,fron~ d i forr~l t l~nr-erlt rllcrtericrl,\.

Soil classification as well gives rough indications f'or soil erodibility. With rcspect to USDA Soil Taxonomy erodibility increases in the order Oxisols < Ul tisols < Alf'isols, Vertisols, Mollisols < Aridisols (Chi-ornec et al.. 1989). Andisols were found highly variable (El-Swaify. 1990) and Vertisols proved more erodible than lnceptisols (El-Swaify 8L Dangler, 1982).

1

erodibility

Erosion susceptibility decreases with increasing organic matter content which in turn increases with soil moisture and decreasing temperature. Therefore, a soil will have more organic matter in a cool highland climate and will bc less erodible than a similar soil in the lowland.

low

basil t

gabbro \hale

coar\e/gravelly \and depo\its carbonate rocks

A very evident indicator for soil erodibility is soil colour. Structural stability and erosion resistance of red hematitic soils is higher than of yellow goethitic soils. In other words, with decreasing hematite content, as seen fsom the redness rating (Torrent & Barrcin, 1993), structure becomes weaker (Chauvel et al., 1976; Muller, 1977). Organic matter content is also roughly estimated by soil dar-kness (Munscll values).

medium

gneis\ diorite ande\i te

high

volcanic a\h

granite rhyolite granod~ori te fine sand depo\it\ \ i l t stone lee\\

5 Indicators for \oil el-osion

mean anual runoff [mm]

A first iinprecs~on of \oil erosion can be gathered fro111 the sediment concentration in the river water. Rivers coming out of forested watershed$ carry very low amounts of sedi~iients colnpared to cropped and gra7ed watersheds (Figure 5-7). An extreme example is seen from a plane approaching Madaga\car. The high sediinent load of the rivers leaves a red corona close to the estuaries and coast.

Finally, a very good indicator for erosion processes is population density. Agriculture creates soil erosion and agricultural systems are often not conservative. Thus, information about the population density combined with knowledge about cropping systems, quality of the soils and cli~uate give already an idea of the erosion potential.

Chapter 5

Further indicators like washed out root\ and top\oil depth are sometimes already useful to quantify soil loss. They are therefore discu\\ed in Chapter 6.3.3.

6 Assessment of soil erosion

Methods for soil erosion measurement depend on the scale applied and the accuracy needed. Measurements are carried out on plots of less than 1 m2 to watersheds of several hundred km7. The accuracy ranges from an estimate of the erosion dimension to ineasurements precise to the kglha. Choosing a measurement system, therefcjrc. needs a clear definition of the problern to be investigated and the accuracy of answer needed.

6.1 Rainfall simulator studies

Rainfall simulators are used in the laboratory or in the field in order to apply storms of controlled length, intensity and drop size distribution to erosion plots. Today, a nurnber of different rainfall simulators exist which apply perrnanent or intermittent rain from needles with drop-formers, small hoses or nozzles to plots of varying sizes. Plot size is limited by the size of the sirnu1;ltors and the availability of water in the field . Simulators can be as

( 1

large to fill a big truck or as small to be carried by hand (Crouch & Collison, 1989; Kamphorst, 1987). Largest and smallest plots of field simulators actually used in Germany and Switzerland are 42 and 0.38 m2 (Auerswald et al., 1992b). A comprehensive review on rainfall simulators is given by Meyer ( 1988) and USDA (1979). A detailed description and discussion of sinlulators used in Germany and Switzerland was published by Auerswald et a1. ( 1992a. b, c) , Auerswalcl 81 Eicher ( 1992). Bechcr ( 1990) and Kainz et al. ( 1992).

Using rainfiill simulators, soils can be tested fairly quick, under standardized conditions and independent of hazardous natural rainfall. The air-dry soil (simulations are advantageously carried out during the dry season) is exposed to several rains with varying duration and intensity. Intensity can be atljusted to the local conditions or to the international standard of 63.5 rnmlh. The latter facilitates comparison with other studies. The standard treatment comprises a first storm of 1 hour. 24 hours later a second 30min storm and after a 15 min break a third 30min storm. This stor111 sequence represents rain o n dry, moist and wet soil as it occurs undcr natural rain.

l 0 As a thumbrule about 10001 of watcr is nccded for a 60rnin storm of 63.5 m ~ n / h on a 10 m2 plot if the intensity is rnensurccl bcfore and after- the storm.

Runoff and soil loss from moist and wet soil are generally larger compared to dry soil. In order to calculate a mean soil erodibility for all soil moisture conditions during the year, soil loss from dry, moist and wet soil is weighted in a ratio of 1 : 0.31 : 0.23 (Wischmeier et al.; 1971). This ratio proved valid for the climate in mainland USA. An attempt to adjust the ratio to other climates was made in Hawaii (Dangler & El-Swaify, 1976) by taking the number of dry and wet" months to weigh the storms on dry and wet soil. A storm on very wet soil was not carried out. Correct results were obtained in Ca~neroon by applying a ratio of 1 : 1 for dry and wet soil (Nill, 1993).

6.1.1 Laboratory studies with simulated rainfall

Laboratory tests are used especially for the study of single erosion processes like surface sealing, rill and interrill erosion, splash erosion, intluence of mulch layers and different rain intensities. Relative differences in the erodibility of soils can also be evaluated. However, if quantitative inSol-mation about soil loss is needed, the results frorn laboratory tests are better calibrated with to data from larger erosion plots under natural rainfall. In laboratory tests only a part of the soil profile (generally thc surface soil) is used and the soil is disturbed i n its natural structure. Therefore, results on ri~noff and soil loss can only be compared t o in situ soils if the runoff volu~ue is determined by the surface layer (= rapidly sealing soils). If runoff is especially determined by less permeable subsurface horizons or the degree of presaturation of the soil, laboratory test are of limited use. This is of'ten the case for well structured soils rich in oxides, clay and organic matter.

The advantage of laboratory tests are the controlled conditions of \lope angle, water temperature and quality (wme test\ are carried out with disrilled water), rain intensity and antecedent soil moi\ture. The \mall plots can easily be handled which facilitates repetition\. The comparatively \ma11 amount of soil needed allows the collection of very different soils distant from one another.

I I A wet month was characterized by the median rainfall exceeding class A pan evaporation lot- the month.

6.1 Rainfall siniulator studies

plan view

flume

side view

For the tests, a layer of soil is packed into a flume about 10 cm deep where it is compacted to its natural bulk density for which a roller can be used. A large nurnber of different flumes and rainfall simulators are used. Therefore only some general features will be given here. A flume is a wooden or metal box with an inner and outer area (Figure 61 1 - 1 ). The inner area is connected to an outlet which delivel-s runoff and sediment into tl

container. The outer area is also exposed to the artificial rain. The idea of an outer area is that the amount of splash which leaves the inner measurement plot is replaced by the amount splash from the outer plot entering into the measurement plot. The bottorn of the flurne is perforated in order- to allow percolation of' the infiltrating rain. The flume can be adjusted to several slope angles. Some flumes are also variable in their length.

6.1.2 Field studies with simulated rainfall

Field studies with rainfall simulators are more tiresome and expensive than laboratory studies. The whole equipment and the necessary crew must be transported to the site. If water is not available in the vicinity it needs to bc carried from several kilometers away and stocked beside the experimental site. Test conditions are not as controllable as in the laboratory. Antecedent nioisture, slope, water quality and temperature can riot be standardized. On the other hand, the soil stays rather undisturbed and the whole soil profile is tested instead of a single soil layer. For determining soil erodibility, the plot and a 50 crn wide strip around the plot are tilled to maize seedbed conditions before the rest. The tilled strip around the plot serves for the same purpose as thc outel- area of the laboratory flume. As the soil docs not need to be transported, the plot size is generally larger in field studies cornpared to laboratory tests. Generally the field si~nulators need to be calibrated on runoff plots with known erodibility.

Field simulators also allow the testing of the effect of various factors 011 erosion such as vegetation cover during different growth stages and seasonal variation in strt~ctural stability. Use of field simulators also gives most realistic infiltration data as it reflects closely the intluence of natural rain. In order to reflect the variation in soils and treatments two to four repetitions are carried out in most studies.

6.1 Rainfall simulator studies

Figure 612-1: Mobile rainfall simulator unit in the field. 1: simulator; 2: manometer; 3: spraying nozzle; 4: supply hose; 5: outlet hose; 6: 300 1 tanks; 7: electric pump; 8: electronic control system; 9: 5000 1 tank; 10: 1 1 sampling bottles; 11: 5 1 beaker,- 12: outlet; 13: dissembled simulator; 14: 120 1 barrels for water transport

Chapter 6

nietrurl 40

dry-run

60-

40-

20- ediment concentral

0 0 0 1000 2000 3000

time [s]

1 .o I wet-run

The plots are bordered by metal sheets which are driven into thc ground to 10-15 cm depth. At the bottom end a metal triangle is put into the soil which collects the runoff into a tube and finally into a graduated bucket (Figure 6 12- 1 ). Standard reported data are:

D antecedent soil moisture before the rain D surface roughness D startofrain(time=O) D time for the first runoff D time for every litre of runoff and runoff samples depending

on the volume (e.g. lst, 2nd, Sth, 7th, loth, 15th, 20th, 30th litre, etc.) for sediment determination.

D end of rain and end of runoff D initial and final rain intensity

With these data runoff/soil loss diagrams can be set-up which demonstrate the erosion process (Figure 6 12-2).

6.2 Runoff plots

Measurements of soil erosion were originally conducted on runoff plots under natural rainfall. A standard plot of 22.1 n~ length. 1.87 m wide on a uniform slope of 9% was taken as 'unit' plot. It served as reference in comparative studies. The 'unit' plot was tilled up- and down-slope to maize seedbed conditions. Seals were regularly destroyed by further tillage (Wischmeier & Smith, 1978). Thus, all conditions were set to attain maximum soil loss. For soil erodibility measurements, a period of at least 2 years under barefallow was recommended to exclude all influences of the former vegetation.

Today, experiments are carried out on plots of different dimensions and on different slopes. However, these plots can be corrected to 'unit' plot conditions with the Universal Soil Loss Equation (USLE). Nevertheless, a minimum length of 9- 10 r-n is recommended for erosion plots. Calculation is facilitated if the sill-thce area is equal to an even fraction of an hectare (e.g SO. 100, 500 or I000 n12) (Shcng, 1990). The plots should be large enough to contain a representative unit o f a cropping system or treatment. Runoff plots range between < 100 m' to about I ha. On larger plots different slopcs and soils as well as deposition within the plot create a more and more complex sitiration which is difficult to interpret. Additionally, i t becomes difficult to contt-ol and measure the large amount oC water and sediment.

The runoff plots are bordered (metal sheets, bricks, earthen ridges planted to grass) in order to prevent outside runoSf frorn entering the plot. Runoff and sediment are collected by a drainage ditch at the bottom of the plots which leads to an outlet. Here, the volume and sediment can be measured by a divider tank system (Figure 62-1) or a Coshocton wheel. A steady rneas~~rcinent of the runoff rate is possible by using a standard t lu~ne and a watel-level recorder.

The tank system consists of a large tank which can collect all runofl'. I f thc expected runoff volume is too large, a system of several interconnected tanks or ban-els is needed. If the first tank is filled the overtlow is separated into a large aliquot (e.g. 90%) which spills into an outtlow ditch and u small aliquot (c.g. 10%) which enters the second tank and so on. Most of the sediment, especially the coarser material. is deposited in the first tank whereas suspended soil particles will be found in the second tank. All tanks should be provided with an underground outlet to hc i litate water evacuation.

Chapter 6

Figr~~-(> 62-1: I l i l ider tmrk \jystc>ilr fi)r tilt) i~rr~rslrwirrrilt of r u i l o f o ~ r d o i l lo.~.c (Snhrl- K o , s c ~ l ~ c l l ~ ~ , 1988)

HS - F l u m e ( \ - foo t ) Connection Tube

Concre te P i t

Coshocton wheels are installed in the waterspill underneath a flume. The water falls onto the whccl thereby making it rotate. A slot in the whccl which is connected to a barrel passes underneath the spilling water taking each time a srnall aliquot of runoff. Coshocton wheels can be purchased for about 1500 to 2000 US$. Their advantage is that they can be installed in several places, while cemented tanks become worthless after the measurement in one place.

6.3 Erosion measurement within existing fields

If less incasure~nent accuracy is needed, there are a number of simple devices which can bc used to estimate soil loss.

6.3 Ero\ion nlcasul-emcri~ within cxi\ting I'ield\

6.3.1 Erosion nails

Erosion nails also called erosion pins can be hammered into the soil until a defined length ( e . g 20 crn) stays out above the soil surface. If this length is reduced or enlarged, sedimentation or erosion has occurred i.e. the soil surface has increased or decreased. The nails are placed along the slope of a field with an interspace of some meters and a lateral displacement of 10 to 20 cm (Figure 63 1- 1 ) in order to avoid any interference on I-unoff fro111 onc nail to the nail bclow (Ziibisch, 1986). Length measurement of the nails rnust be carried out with high prccision. An crror o f 1 nlln in length means an error of' 10-1 5 tlha if bulk density is supposed to be between 1 .O and 1.5 gIcm3. Therefore, a metal plate 5 ctn in diameter is slipped onto the nails to compensate for random roughness of the soil surface. The length is measured with a slide calliper precise to 0.01 rntii (Figure 631-2). The mean length of' all nails is con~pared to the mean initial length. The soil loss can than be calculated by the missing soil height if the soils bulk density is known.

A\ the measut-ement error can be appreciable, this tilethod is e~pecially suitable for measurernentc over a period ol' several year\ or for \ite\ with a high erosion potential.

Chapter 6

I

Oel % lateral displacement to avoid interaction of nails

I I I I I 2 m I I I

slope 2 m

erosion nails

6.3 Erosion rneas~~rement \ within exi\ting field\

Slide caliper erosion pin

\ \ iron

/ soil

- -il

surface

6.3.2 Sediment traps

Sediment traps are simple and cheap devices which permit the measurement of runoff and soil loss. They were extensively used in ~iieasurements in Kenya (Zobisch, 1986). The traps consist of a 50 x 50cm metal box closed on threc sides by a 5 cm high I-im. A IOcm long extension of the bottom is left at the open front part. This extension can be pushed into the soil in order to allow runoff to freely enter the fnr end of the box which is installed at a slight angle. The box is covered by a lid ;to avoid direct access of rain water. At the far end a funnel is attached to the trap which is connected to a 301 reservoir where runoff and sediment are collected.

6.3 Erosion mcawremcnt within exi4ting I'ield4

The sedinient traps are installed in places with homogenous slope. A conversion ditch 10m in front of the traps diverts runoff from further up- slope. The catchment area of 5 m' for the sediment traps is given by the width of the traps (50cm) and the slope length (10m). It is assumed that runoff which enters the area laterally equals the runoff volu~ne which leaves laterally. A sketch of a sediment trap system is given in Figure 632- 1 .

A system with this de5ign collect\ about 6 mm of runoff. I t needs to be emptied after each rain. For larger storrns either the re\ervoir mu\t bc bigger or the catchment area reduced.

6.3.3 Diverse techniques

Root growth of many trees is evidently inhibited on the exposed root parts as thc bark and cambium are damaged. This was found for Pillus aristata (La Marche, 1968), Pinus cdulis, Juniperus scopulorum and J. osteosperma (Carrara & Carroll, 1979). Soil loss is calculated as the depth since exposuse of the upper root surface which is observed on the growth rings (Figure 633- 1 ).

Figlcl-c) 633-1: Ccrlculafi~zg fho time sir1c.e exposure of' tree roofs hj' t l ~ c growth rirzgs mot ~ e r t i o ~ ~ s

eroded part

of root

Dunne et al. ( 1 978) used a similar approach in Kenya by measuring the height of the mounds underneath Acacia drepanolobium, A. tortilis. Scricomopsis pallida, Olea africana and Acocanthera species (Figure 633-2). These tree species were chosen because they do not develop any superficial roots. However. caution must be given that the mounds were not formcd by water and wind erosion, termites or the trees themselves. The age of the trees was given by a regression between number of growth rings and diameter of thc stem. Acacia drepanalobium is reported to develop a physiological mark on the stem at the levcl of the original soil surface (a bulge, branching or change of bark colour). An indicator for very erodiblc sodic soils is Colophospermum mopane (Stocking, 1988).

0.3 Erosion measurement within existing I'iclds

minimum level of former soil surface -------------

resent soil surface

Some neth hods can orlly be used very site-specifically. Rhoton ct al. ( 199 1 ) used the gravel concentration of the surface soil in order to calculate the eroded depth of a soil with rather homogeneous gravel content.

Chapter 6

6.4 Sediment yield from river basins

Suspended sediment yield of rivers is calculated by measuring the cross section of rivers, the waterlevel, the discharge and the sediment concentration. Sediment yield, however, only gives a rudimentary estimate of' the soil loss from fields within the basin. Several factors bias the result:

- Sediment yield measurements generally only cover the suspended sediment load. The bedload which is carried close to the river boltorn is neglected. The bedload of African rivers accounts frequently between 5 and 1 0 % of the suspended load (Walling, 1984).

- The soil lost within a watershed is not entirely transported into the river. Sedimentation occurs on foot-slopes, depressions and well vegetated parts within the watershed. The sediment delivery ratio (SDR) which gives the suspended sediment load of the river relative to the total soil loss in the watershed varies with watershed s i ~ e (Figure 64- 1 ). The SDR from large watersheds is s~naller than fro111 s111a11 watersheds as the mean gradient declines and sedimentation increases with increasing watershed size.

- The soil lost from fields in the watershed can be subject to several cycles of deposition and remobilization until i t reaches the river outlet. Thus, measured suspended sediment load may reflect soil erosion of' formel- periods.

- Suspended sediment is not only derived fi-orn sheet erosion within the watershed but may also stern from gully erosion, landslides, channel and streambank erosion.

river

Watari Bun\~u-u Senegal

1 Faleme , Ham~nanl Kebir Oue5t

1 Meunu

country I basin area (suspended sedimen

Nigeria Nigeria Mali Mali Algel-ia Algeria Ethiopia

(krn2J

F A 0 soil loss estimate load

[t/(ha;J:a)j

6.3 Sedinicnt yield I'ro~n I-i \rr ba\in\

The difference between suspended sediment loads of rivers and estimated soil erosion rates fror-n the respective fields is demonstrated in Table 64- 1 . Suspended sediment load is generally an order of magnitude lower than the estimated total soil loss.

Figltt.rCi4-1: Relcltiotl.~l?ip brfit'retz \t,crtur.sl~ecl tlrctirzcrgo at-eu C I I I L I setlit~~erlt cleli\vrj rcrtio as 1tsc.d hjr the U.S. Soil Corzset-11trtio1l S~t-i'ice ,fi)t- the c'etltt*er/ nrz~l ee~stem USA ( f i ~ 1 1 7 2 WCIIII'II~, / 9 N H )

0 01 0.1 1 10 100 1000

Drainage Area (km2)

7 Soil loss prediction with the Universal Soil Loss Equation

Erosion has already been noticed in ancient times. Plato already described the disastrous effects of the denudation of the hills around ancient Athens more than 2000 years ago (in: Herkendell & Koch, 1991 ). However, more attention to the problem was only given by the 1920s when the menacing extent of soil loss in the US became aware (Bennett & Chapline, 1928; Lyon & Buckman, 1922). As a consequence the US Soil Conservation Service was created in 19.35. Soon i t became insufficient to notice, describe and measure soil erosion. For a deeper comprehension of erosion and its assessment under varying conditions, i t was important to understand the basic processes.

The development of mathematical models started with the equation o f Zingg ( 1940) which related soil loss to slope length and gradient. Smith ( 1941 ) included factors for the influence of crops and conservation practices on soil loss. The addition of a rainfall fiictor resulted in the Musgrave equation (Musgrave, 1947). Finally data collection and analysis of 10.000 plot years from 49 locations led to the 'Universal Soil Loss Equation (USLE)' (Wischti~eier & Smith, 1978) which, today, is still the basic tool for soil conservation in the US and other countries.

The USLE is an enipirical model with widespread use in land use planning, extension and the design of cropping systems and conservation practices. I t allows to estimate soil loss under varying climatic. topographic and management conditions on different soils with a set of relatively sirnple parameters. The basic idea was to measure maximum possible soil loss of a specific soil on a control plot with standard size, gradient and treatment, - the 'unit' plot. The 'unit' plot was 22.1 m long on a 9 % slope. Soil loss as caused by gradients, slope lengths and management conditions different from the standard conditions was examined relative to ~naximum soil loss on the control plot which was achieved by barefallow tilled up- and down-slope to maize seedbed conditions. The equation is expressed as:

with A rilean, longterrn annual soil loss [t/ha*a] R erosivity of rain (Nlh]

7 Soil leas pl-ediction

K erodibility of a soil, i.e. its susceptibility to erosion It+h/N:. bhal

L slope length factor [ - I S slope steepness fdctor [-I C management factor [-I P support practice factor I-]

Soil loss (A) gives the mean annual soil loss in tlha on a longterm basis. Soil loss of a specific year may differ considerably from year to year. Rainfall erosivity (R) is calculated from rainfall charts for single erosive rains during a period of 22 years and represents the mean annual erosivity for this period. Soil erodibility ( K ) indicates a soil's susceptibility to the erosive forces and gives the amount of soil loss per unit erosivity. K was defined constant for a specific soil. L, S, C and P are expressed as ratios of soil loss on a given plot to soil loss on the unit plot. For example, an L factor of 2.1 for a 100 m long slope of 9% means that this slope will suffer 2.1 times the soil loss of the 22.1 tn long unit plot if all other conditions (climate, soil, management etc.) are alike. A C Factor of 0.2 for a crop signifies that soil loss under this crop is only one fifth of the barefallowed unit plot provided that all other factors remain constant.

The model parameters were calculated from a defined set of natural and management conditions in the US. Therefore, it was not surprising that the application of the USLE has led to contradictory results under tropical conditions (Lal, 1980; Mtakwa et al., 1987; Ngatunga et al., 1984; Roose, 1977; Vanelslande et al., 1984). Part of the differences were however caused by treatments very different from the one's defined by Wischmeier 8( Smith (1978). Recent data show, that the USLE can be directly applied to a wide range of tropical soils and corrections can bc made for most other soils (Nill, 1993). The most urgent need exists now in obtaining reliable data on tropical cropping systems.

Today, several deterministic models exist which try to con\ider the numerous, complicated processes which determine erosion. Mostly they need a large amount of infortnation on climate, soils and management. Often they are not tested under differing conditions. Compared to these models, the USLE convinces by its simplicity, the large data base which was used fur its development and its widespread application. Although empirical in principle, it still includes all irnportant factors which influence soil loss. Its parameters, possibilities and limitations will be outlined in the following chapters.

Chapter 7

The USLE was designed to predict longterm annual soil loss from a given slope under specified land use and management conditions (Wischmeier, 1976). It can be used for watersheds, if these are subdivided into smaller units where the USLE factors apply. Using mean gradients, erodibilities and slope lengths for the whole watershed may cause important errors in the estimate. Soil loss, as estimated by the USLE should rather be regarded as best available estimate than as absolute data. Soil loss from a specific event can not be calculated with the USLE. Even annual soil loss of a specific year may vary largely from longterm mean annual soil loss. The USLE does not account for deposition of sediment along field borders, ridges or on foot slopes and can not predict gully erosion.

Beside the USLE, a second important prediction model is applied in southern Africa. The 'Soil Loss Estimator for Southern Africa (SLEMSA)' (Elwell, 1980a) predicts mean annual soil loss (Z) on a given slope by:

with K mean annual soil lo\s from a 4.5 O/c slope, 30 m long under conventional tilled bare soil

X adjustment factor for different slope length\ and -gradients C adjustment factor for the influence of crop cover derived

from the annual energy distribulion curve and growth c ~ ~ r v e s of crops

An appreciable database was collected for this model. For furthcr details rcfer to Elwell ( 1980b), Elwell ( 1984) and Elwell & Stocking ( 1976).

7.1 The erosivity of rain

7.1 The erosivity of rain (R factor)

Wixhrneier & Smith (1958) found that soil loss increased linearly with a storm's total kinetic energy (E) times its inaxiinuin 30 minute intensity (I3()):

with R longterm mean annual erosivity [ ~ l h ] " E kinetic energy of a storm j [kJ/rn2] I,,, maximum storm intensity of storm j during 30 lnin [mm/h]

for I,,, > 63.5 mrnlh: I,,, = 63.5 mm/hl' In number of erosive storms j per year [-1

The energy of a storm is calculated by:

with I i intensity for storm interval i [rnmlh] for 0.05 < I < 76.2 mrnlh; for I > 76.2 rnrnlh 1 = 76.2 mm/hl

PI rainfall volume during interval i [mm] n number of storm intervals i with equal intensity [-1

R is calculated from raingage charts. Each storm is divided in i intervals of constant intensity ( I ) . For each interval intensity, volume and energy are calculated. The total storm energy is the sum of energy of all intervals. An example is given in Figure 7 1 - 1 :

'' R is oftcn given in IJS units us [hundreds foot tons 'I' inlac ph] or as Ifoot tons ": inlac ph] . Multiply by 1.735 or 0.01735, respectively. to receivc INthl. 1 [Nth] = 10 IMJ :!: mmtha :': h ]

1 ,,, was limited to 63.5 rnrnlh because col-I-elation coefficients between crosivety and \oil loss

irnproved by introducing this threshold (Wischmeicr & Smith, 1978)

The rnaxirnunn intensity was limited to 76.2 mmlh because drop diameters do not increase any rnorc for \.cry high intensities (cf. C'hapter 4.1)

Figcue 71 - 1: Strip rlztrt-t of rr 20 m n ~ storrzl registered n~itll tr .\elf' rc.c.or-cling rcri~ gcrge rtyith cr pcrpe).,fLotl t-trtc of'hO t11171/11

A 20mr-n \term was registered by a raingage with a papenpeed of 60 mmlh and a cylinder which emptied automatically after each l0mm of rain (= vertical drop of the line). The storm \tarted at 17.00 hour and lasted until 19.48 hour. It was divided into 4 intervals of approximatcly equal intensity (= slope of the ascending curve). Energy is cornputed as follows:

The maximum rain volume during 30 min was 7.8 mm in interval 2. Thus, I,,, equals 15.6 mmlh and R = E :!: I,,, = 0.4 9 15.6 = 6.2 N/h.

Only 'erosive' storms are used in the calculation. For the US, they were defined as storms with at least 12.5 mrn (112 inch) of rain or, if less, a maximum 30min intensity of at least 12.5 mmlh (Wischmeier & Smith.

energy

[kJ/m2]

0. 13 0. 22 0. 04 00.008

0.4

intensity

[ m d l 6. 6

12. 8 4. 7 1.4

7.1

rain volume

[mml

7 10 2.4 0.6

20

interval

1 2 3 4

1-4

duration

[min] 64 47 3 1 26

168

7.1 The erosivity of rain

In Germany. the limit for erosive storms was set to 10 mm height or I0 mm/h as maxirnurn 30 min intensity (Schwertrnann et al., 1987). The 10 mrn threshold also proved to be valid for stations in Cameroon. Nigeria and Kenya (Nill. 1993; Ulsaker & Kilewe, 1984; Wilkinson, 1975) and was used for all further computations. Storms separated by less than 6 hours are considered as one storm (Wischmeier & Smith, 1978).

The tedious procedure for the energy calculation is easier done by cornpoter and d ig i t i~ ing board . Providing erosivity data on a nationwide b;i\i\ i \ an important task for the national meteorological kervices.

In practice, the calculation o f reliable R factors faces several constraints. Ideally, the calculation is based on daily rainfall records over 22 years (Wischmeier & Smith, 1978). However, in most countries it is already \,cry satisfying if 10 to 15 years of complete data are available. The obligatory subdivision of individual storms into intervals of similar intensity ancl the recognition of the maximum 30 min intensity demands self-recording raingages with high resolution. Very often these requirements are not met and several estimation procedures and indices have been developed in different countries in order to rcplace the R factor (cf. Chapter 4.1 ).

Determination of the R factor:

Erosivity for many locations in Africa must be estimated from available data of different origin:

- In some countries erosivity is calculated for single sites. - In other countries regressions exist which may be extrapolated

to the surroundings. - For some countries national or regional erosivity maps

(iso-erodent maps) are available.

For countries where no erosivity data are available Eljo must be derived fro111 rain data or rainfall distribution maps.

l 5 A \ol'~\zar.c psogr-anlln for d ig i t i~ ing and analy/.ing sainl.all chart\ i \ a ~ a i l a b l e Iiorn: Dr. M.'. Martin. Bahcri\che\ Gcologisclies Lancle\rimt. Hel3str. 118. 80797 Miinchun. Ger-rnun

The quality o f the obtained erosivity values will be no re reliable for sites or areas where E13,, was directly calculated from rain data (provided that the measurement period was sufficiently long). If regressions are used. the reliability decreases with increasing difference in climate and increasing distance from the stations of which the regressions were derived. National erosivity maps generally will be more precise than regional maps. For estimates of erosivity from rain data or maps, several regressions can be applied (Table 12- 1 Annex).

For the Sahel countries Roose's regression is recommended (Roose, 1977). The regression of Bresch ( 1 993) was developed from 18 stations in Cameroon with 700 to 4000mrnla. Its use is proposed for the semi-humid to humid parts of West and Centl-a1 Africa. The equations for Zambia (Pauwelyn et al., 1988) and Zimbabwe (Stocking & Elwell, 1976) are based on a large and well described data base and are recommended for areas of southern Africa with comparable climate. For the highland areas of East Africa, the regression for Rwanda (Durand, 1983) and Kenya (Moore, 1979) can be used.

If regressions for near by neighbour countries are available, the user may decide whether the climate can be co~npared to these countries and whether these regressions may provide reasonable estimates.

In order to obtain El3u for a particular location, look for the country in Table 7 1 - 1 and check if the site or a site nearby is listed. Table 7 1 - 1 indicates the reference tables in the Annex where you can find the EIjo values. If the site is not listed, look up a national or regional erosivity or rainfiill map as indicated in Table 7 1 - 1 I".

A given rain falling on low slopes (between 0.2 and 4 % ) is not as erosive as the sarne rain o n steeper slopes due to the formation of a protective water mulch. Correction o f erosivity on low slopes demands the erosivity of the 10 year storm (EI,JIo)". E1,,410 was found to be inore suited than inran annual erosivity as runoff depth is especially determined by the intensity of'

Daily rain data for a11 \lalion in Benin. Ruskinn Faw. Cameroon. Central Afsic:ui Repuplic. Cliaci. Congo. Ciabon, Ivory ('oast. M:~li, Nigcr, Scnegal and Togo are a~a i l ab l e 1'1-0111 the Cornit6 In(erafr-ican d'Etude\ Hydl-a~tliquc (CIEH). B.P. 3hO. Ouag,.aJougou. Buskina 1:;lso 111

luo \erich. Sesie I : Station\ cctablislieci until 1965. Sesie 11: I965 - 1080.

" ,411 c\timation method lor E17,JI0 is descl-ihed in Annex 1.7.

~ndividual storm\ (Renard et al., 1992). With EI,,,/I 0 a correction fiaclor can hc scad fro111 Figi~re 7 1-2. Enter the chart vel-tically from the x-axis with the EI:,,/IO of the site. Choo\e the gradient of the slope and read the corrected ero\ivity value by moving hori~ontally to the y-axi\.

Figllt-c~ 71-2: Not~logt-trph fi)r. tllt~ c'ort-c~c.tiotl of oros i~ i t j* for f l ~ e eff2,c.f o/

nvlirer r?r~ilch or! ,rlopc.c h e t ~ v o o i ~ 0.2 nlld 4 C/c (Fot tor , per-so~lrrl c7ot?lt?zllnic.otiot~)

100 200 300 400 500 600 700

El30 of 10 year storm [Nlh]

If hail i 4 a frequent event a \ i n ronie mountain area\. the annual cro\l\ity \hoi~ld be con-ec~ed by estimating the percentage of annual pl-ccipi~a~ion as hail. This percentage of the annual ero\ivity mu,t be multiplied by 2.5 (Hurni, 1980) and added to the remaining annual erosivity.

Chapter 7

Example:

It is e\timated that 20 52 of the annual rain falls during hailstorm5 in an area with a mean annual erosivity of 800 Nlh. 20 5% of 800 Nlh corre\ponds to 160 Nlh. Multiplied by 2.5 = 400 Nlh. Thus, the annual erwivity corrected for hail i5: 640 Nlh (= 80 %) + 400 Nlh = 1040 Nlh.

country site

I I map map map

MU, lN/hl

I sites regression national regional

Annex

Algeria

rain volume [mml

1 .1 1.2 1.3 1.4

X Ciour~u-i (1shcr h i ~ \ i ~ l ) Hcri/ (Isser hnsin) Macljouclj (Isser basin)

Sidi Mohiirnccl Cherif ( 1ssc.r hasin )

Angola Benin Botswana Burkina Faso

Boho-Dioulu\\o Dori

Fada-N'Gou~.rna

F~iraho-Ba Garnpela ncar

Ouagacioi~gou

Gaoua ( ;on\& []car

Ouagadougou

Mog~cdo

Nii~ngoloko

Ouagaclougou Ouahigouya

Saria (Mctcn)

Burundi Mashit\i (Gihcta)

national

X X X X

X

X

X X X X

X X

X X X X

X X X X X

X X

7.1 The eso\i\ ity of rain

country site

Cameroon B. f ' . 'I 1'1

B~lmenda Rang;itigte

Batouri Dilxinlha

Dollala Dwhang

Garoua M L I ~ O L I ~

Mciganga Nachtig;ll

Ngaoi~ndGrC Nhount i ja

Penha Michel ( B a n w a ) Poli

Y ~1ound6

Y o k o

Central African Repuhlic Chad

Congo Eg\ pt 13quatorial Guinea

Ethiopia Gahon Ganihia Ghana Guinea (;uinea-Bissau I\or! Coast

4hldl'in Ar,lpu16

Bo~lithk

countrj site

Ivory Coast Divo

KorI1ogo

Kenya Eldoret

K a t i ~ ~ n a n i

(Macliako\)

Kisurnu

Kitale

Lodu iur

Muli~idi

Mornhasn

Nuiruhi

( Kahctc) Nakul-LI

N:uiyirhi Naroh Voi

Lesotho Liberia Ldihya

hladagascar Hefandriiu~ia Malawihlnlr

hlorocco klauretania Xlocamhic~ue Niger

AllohoLo

Nigeria Alore

Culabal Enugu

I bad~ui

Ih orn N\ i~kha

7.1 The e t o v L ~ t y ol ralu

country site

1 Nigeria -

Onitahu O u e r r i

Port- Harco111-t

Umildike Rwanda

Butare Galiir(a

( i iwnyl

K a ~ n c m b e

K i p 1 1 (~lir[>(3rt)

Kuhengeri Sao 'l'ome and Principe Senegal

Barnhe!

s 6 1.21

Sierra Leone Somalia South Africa Sudan Tanzania 'I'ogo Tunisia Llganda Zaire Zamhia

Chipatii

Kabompo K a b u e

K a l ' ~ ~ a Polder

K:r\:~rrla Mwir11Iung;l

Ndola Sed ieke

Chapter 7

rain volume Ln1mI

national map

country site

Zimbabwe Bcibridgc

Chipinga

Chisumbarijc

Delt

Eastel-n Ilistricr

Enkeldoorn Forr Victoria

Gokwe H i g h ~ e l d

In) ungn

t i u r o i

L o w ~ c l t l

Lupane Middleveld Si~lishury

Tjolotjo 'Tuli

Regional: Sahel

El.,,, [ N h l

sites regression national regional map map map

X X X

X X X

X

X

X

X X

X

X

X X

X X

7.7 Soil erodibility

7.2 Soil erodibility (K factor)

The soil erodibility factor K of the USLE expresses a soil's susceptibility to erosion. It is defined as '... a quantitative value experimentally determined. For a particular site, i t is the rate of soil loss per erosivity unit as measured o n a 'unit plot'. A 'unit' plot is 72.6ft long, with a uniform lengthwise slope of' 9'k, in continuous fallow, tilled up- and down-slope.' (Wischrneier & Smith, 1978). Crusts on the soil which form during rains have to be regularly destroyed by furrher tillage. In order to exclude influences of the previous vegetation, the unit plot is kept under barefallow for at least 2 years before determining erodibility. It is assumed, that by then soil loss is primarily a function of inherent soil properties and increases linearly with the rainPall erosivity. Erodibility is considered to be a specific constant for a soil and is calculated by:

K = A

R L S C P I t h/N ha]

On a unit plot L, S, C and P equal I and the equation can be written as:

A K = R [t h/N ha]

This basic concept o f erodibility can also be applied to tropical soils.

As shown in Figure 72-la, erodibility initially increased after clearing of the vegetation. After 1000 to 2000 N/h (which corresponded to 2-3 years in the example) a steady erodibility value was approximated. However, on some soils erodibility may still increase or decrease after some years of barefallow. Erodibility of the soil in Figure 72-lb, for example, started to decrease slightly after 4600 N/h. This is the case if a surface horizon is partly or cornpletely eroded and tillage inixes the underlying horizon with a lower erodibility more and more into thc initial surface horizon. Tropical soils rnostly have surface horizons of less than 15 cm depth. Partial or complete truncation

Figrrrc~ 72- 1: Chotzge ofsoil erodibility with cr4i~z~rl~rti~~e e r o . ~ i , ~ i t ~ ~ oil [rtr

Ultisolfi-oi)i Cl11~0rooi7 ( ( 1 ) cltzd ail Alfisol f;.oi~ Nigeriu ( 1 1 )

soil soil loss [cdal

5.8 Andisol over basalt

[Yha*a] 698

7.2 Soil cl-odihility

ol' a barefi_lllow \oil i i poq5ible under tl-opical rain within a few year\ a \ ihown by annual cro\ion depth\ of ioriie Canieroonian soils (Table 72- 1 ).

A decrease in erodibility occurs if the surf.iice soil or subsoil contains coarser particles like quartz or iron oxide gravels. With the selective rcriioval of the l'i~ic-earth. the gravel is enriched on the soil surflice and protects the soil. Soil loss estimates for gravel-covered soil need, therefore. to be corl-ected 1)s the protective influence of' thc covcr. An increase in erodibility takes placc i f an irnstablc sirbsoil ( e . g with high sodicity) is more and more incorporated into the surtlcc soil.

Measuring erodibility is time consul~iing arid expensive. Wischrneier & Smith (1978), therefore, cariie up with an equation to calculate erodibility from sirnple soil properties which arc moasured routinely:

K = 2.77'!' 10-0 MI 1-l ( 12-OM)+O.O43 (SC-2) +O.O33'!' (3-PC)

M here M I-] = (\i+fl'S) ( 100-cl)

~ i t h c1 clay [ ( X 1 \ i \ i ~ t [ (x 1 ' ' ffS very fine sand (0.05-0.1 ~ i i m ) 1% ] O M organlc rnatter [C/c I SC str-ucturc clas\ 1-1 PC perrneability cia\\ [ - I

The equation shows that soil erodibility incl-cases with increasing silt plus very fine sand content of the soil. It decreases with increasing clay and organic niatter content.

Structure class of a soil (Table 72-2) docs not refer to the actual structure of the soil surface of a field but to structure after 2 years of barefallow. Therefore. sorne experience is needed in order to assign a str~icture class to a soil. Soils with an ~rnsrable structur-e develop coarse f ragr~ient~i f ' t e~- prolonged barefallow periods whereas stable soils maintain an aggregated surface. The coarser the final struct~rre, the higher the structure class and erodibility.

structure class

The permeability of a soil describes its inf'iltration capacity and ability to conduct water. Per-nieability classes (Table 73-3) must be detel-mined for all hol-i~ons down lo XO cm depth. For each horizon a permeability class is chosen. The permeability class of the soil is deterrnined by averaging the permeability classes of all I~orizons.

structure size [mm]

I 3 - 3

4

It the h o r i ~ o n with the Iowe\t pcrrncabil~ty I \ within the upper 30 cnl, i t \ pesnieab~l~ty is counted twice before averaging. If the lea\t permeable hor17on I \ found within the uppel- 20 cm, i t determine\ the permeability cia\\ of the \oil.

mean aggregate

I

For field use, the permeability of a soil can be cstilnated by using information on biological activity or structul-e in the profile descr-iption. An exarnple is given in Tublc 72-4. However, use of such data needs experience and should only be considered carefully.

very fine CI-umb tine crumb

medium to coarw crumb hlochy. platy or mas\ive

< 1

1-2 2 10

> 10

7 . 2 Soil erodibility

tigrlr-r 72-2: C'oirq?rrri,\orl 01 c.rrl(.r[lo~rd (K,,,,,) r r ~ ~ d i~zen,sur-ed (K,,,,,, ,) soil er-ociibilit~, for 28 .soil.\ fro111 Ccz~llcroor~ crild Nigc>r-io

description

very few pore< few pore4 common poi-e\

many pored pot-ou4

very porous very high biological activity. very porou\

Equation No. (24) wa\ applied to soil\ with < 65 '/c \and and < 35 % clay (Wischmeier et 31.. 197 1 ). K factors for \oils beyond the5e limits need to be deternl~ned from a nomograph. However, ;I recent investigation indicated n o quality lo\\ if erodibility wa\ calculated by the above equation for \oil\ beyond the\e textural propertie4 (Ni l l . 1993).

permeability class

I 2

3 3 5 6

Chapter 7

Erodibility measurements on 28 tropical coils t'rom Cameroon and Nigeria showed that equation (24) can not be applied to all tropical soil\ but needc correction factors for 3 different \oil group\ (Figure 72-2). For part of the \oil3 (group I ) , erodibility as calculated by equation (24) underectimated the measured erodibility whereas group 3 wa\ clearly overe\ti~nated. For group 2 (about half of the soil$), calculated erodibility agreed well w ~ t h meaqured erodibility.

A d~scrirninant analysis can distinguish between unknown sods by using two di\crim~nant function,". 89 % of all sollz were correctly cl;l\\ed by uqing:

- bulk density of 5-10 c111 depth [g/cm7]; meajured one day after the soil had been tilled with a hand hoe to \eedbed condition5 of m a i ~ e

- \ilt content of the \urface \oil I(/( ] - organic matter content of' surfllce soil [%] - pH in water o f the \u~-face \oil; measured in 1 : 2.5 soillwater

~ 5 p e n s i o n after 18 h - amount o f air-dry aggregate\ of the surface \oil (0-5 cln) bi th

0.6-0.2 rnrn diameter [%I rnea\ured by dry-sieving

Wrong classification only becomes dangerous if a soil's erodibility is underestimated. If it is overestimatcd, too much conservation efforts may be the result which means an exaggerated input of labour and money but no virtual danger.

All soils with a very high erodibility (group I ) were correctly classed. 9% of low erodible soils (group 3) were classed into group 2 and would receive more conservation than necessary. 17 % of the medium to high erodible soils (group 2) were assigned to group 3 and would receive insufficient conservation.

Based on the two discriminant functions, Fisher's Linear Discriminant Functions were derived which facilitate classification. They were:

'('function 1 = 13.03:i:BD(5- 10) + 0.169:!:si + 0.265:!:OM - 1.62:::pH - 0.066::agg(06-02) - 5.62 eigenvalue = 1 . O X , Wilks' Lambda = 0.19 sign. = 0.000 1

eigenvalue = I .OX, Wilhs' 1,ambda = 0.48 sign. = 0.0038 with BD(5-10) bulk density in 0 - 5 cm Iglcrn']

si s i l t [%] O M organic rnatter [ % I agg(O6-02) dry sieved aggregates with 0.6 to 0.3 m dian~elcr ('A 1

The group centr-aid\ for function 1 and 2, respectively. were -3.6 anti 0.93 for ~ I - ~ L I P I . 0.938 and 0.77 for groi~p 2 and -0.76 and I . 1 X for group 3.

7.7 Soil erodibility

group 1 = 1.118 si + 90.5 . BD(5-10) + 0.416 OM + 13.4 pH + 0.724 agg(06-02) - 100.6 (26)

group 2 = 1.7 . si + 134.7 a BD(5-10) + 1.395 OM + 7.577 .+ pH + 0.478 r agg(06-02) - 114.4 (27 )

group 3 = 1.343 4 si + 114.2 *BD(5-10) + 1.617 *OM + 7.816 * pH + 0.378 .;; agg(06-02) - 90.5 (28

In order to assign a soil to one of the groups, the three functions must be solved. The soil belongs into the group whose function yields the highest value. Once the group is selected, the erodibility of the soils (K-trop [t - h/N -ha]) can be calculated by the following regressions:

Chapter 7

group 1: Kt,,, = 2.3 K,,, + 0.12 (29)

group 2: Kt ,,,, = 1.1 , K,,,

Applying the three regressions to the set of 11-opical soils mentioned above, explained 92% of the variation in measured soil erodibility (Figure 72-3). The third regression was not significant. However. soils in group 3 have very low erodibilities (maximum K,,,,;,, = 0.026) and prediction errors lmay be tolerated.

How can the typical soils for groups I to 3 be characterized? Table 72-5 shows average properties for the groups. Group 1 contains soils with more sand, less clay and a higher amount of aggregates in the size fraction of 0.6 to 0.2 mrn and have a slightly higher pH than soils in ~ I - O L I ~ 2 itnd 3. Bulk density is lower than in group 2. The surface soil of group 1 readily seals. Their low bulk density and high ariiount of transportable material enables high soil loss rates. They are characterized by an early occurring runoff, high runof'f' rate and coefficient. The agronornically very important volcanic ash

soils belong to this group. An alternative ecluation to calculate erodibility was developed by El-Swaify & Dangler ( 1977) for a group of seven residual soils and five volcanic ash soils from ~awa i i " ' :

20 dimen\ion lol- K: [[(on ;: ac1.t. .:: IIOLII./ hundrt.d\ ol'ac1.c .:: fool t on \ ;:: inches]: in ol.tit?r to at't'I\'e at I r :i: IIIN ::: ha1 tnultiply wirh 1.3.

7 . 2 So11 erodibility

Tcrhle 72-5: A\wrcrge nrtlojfl .coil lo.~.s t r ~ e / soil pi-opurtius ,for t l~u tl~roe rrodihilit~ groups (1~cr1zte.s $tirll different lettens crre .sigil~fi'c'crrlfl~. c/(ffet.et~f crt the 0.0.5 /oI)o/);

wit11 LT250 In percentage of soil which passes a 0.25 t11r11 sieve by wet-

parameter

\and [(k] very fine \and [(/c I \lit 1% 1 clay [ % I organic matter [ % I pH bulh den\~ty ( 5 - I0 cm) {g/cm3] bulh dens~ty (0-10 cm) lg/cmz]

aggregate4 (0 6-0.2 mm) ['A] Ktliea5 [ t h / N ha]

$tart runoff 141

111ean runoff rate I llmin m']

~naxlnium runoff rate [llnlln ,111'I

runoff coel'l'ic~ent 1 % ot l-aln]

$011 lo\\ [tlhu] mean $o~ l lo$$ rate (g/l I m ~ n ] maumun~ $011 lo55 rate [g/l min]

slevlng MH (\iuid > 0. I 111111) (silt + Lery fine \and) H C ba\cx \ ; i l~ l~- ;~ t io~i i l l 1 11 NH40Ac at pH 7 \ i percent \ i l t s ;1 \and > 0.1 111111

Soils in group 3, as the other extretiie, tcnd to be more clayey iuid richer in organic 1-naltcr than soil$ in g ro~ lp 1 and 2. Bulk density is slightly lower than in ~ r o u p 2 ancl there are le\\ aggl-cgates of 0.6 to 0.2 mnI ilinrnetel.. Thew soil\ h:it-dly \eal and I-unoff start\ very late if rain fall\ o n dry soil. Soil lo\\

group I 5 2"

2.8"

1 6" 32"

4.2" 5.3"

0.87" 0.88"

19"

0.2775"

X60a

3.1

5.2" 3 2,'

8.0

33,l 49%'

means for: group 2

43,l

3. I " 1 8" 40G'

3.8" 5. l L 1

1 . 0 6 0.97 'I'

2sh 0 0969''

1339"

2.2" 4 3"

23" -. 3 9''

- 3?s1 -

36'

group 3

31"

3.0" 1 7" 52h 6.3" 5 .O"

0.87' 0.79 -- 37'1

0.0077L

3X73b

0.3" 0.8

3h 0.1'

3" 5"

Chapter 7

under natural rain OCCLISS especially after sequences of several storms which presaturate these soils. Their runoff-soil loss behaviour is not determined by surface sealing but by the permeability of the profile. The poor I-elationship between surface soil and profile properties explains the insignificant I-egression in Figure 72-2 for this group. Drop impact and storm have little influence on soil loss. As runoff in group 3 requires a presaturation of the soil, occasional rains during the dry season are of no danger. Especially clayey, iron oxide rich soils from basic parent rock are found in this group.

Soils in group 2 tend to have more silt and very fine sand, along with mediurn clay and sand contents (Table 72-5) . The organic matter content is lower than in the othel- groups and bulk density higher. Their sul-face seals as in group 1 but sealing and runoff occurs later during a rain resulting in lower rnean runoff and soil loss. This is indicated by smaller differences of ~naximum runoff and soil loss rates of group I and group 2 soils compared to mean rates. It can be assulned that maximum runoff is reached at the end of a storm for soils in group 2. Croup 2 had 7 1 % of the mean runoff rate of soils in group I but reached 83% of the maximum runoff rate of group 1 . The values for mean and maximum soil loss were 7 0 and 74%, respectively. Thus, with increasing rain volume the difference in runoff and soil loss between group I and 2 became smaller. However, mean and maximum soil loss rate did not differ as rnuch as mean and rnaximuln runoff rate. This suggests that runoff increased more than soil loss. Typic soils in group 2 are formed from metamorphic basement rocks.

Soil taxonotny gives some, though not very safe, indications. Oxisols frequently are to be found in group 3 although some occur in group 2 as well. Ultisols, Inceptisols and Alfisols are especially found in group 2. However, some soils in group 2 also have low erodibilities (Figure 7 2 - 2 ) and rather stable structure.

7.2 Soil credibility

L)etenninatioiz ofthe K factor:

1 . Calculate K,,,, according to cquation (24). Silt, clay, very fine sand and organic matter content ar-e taken from soil analysis o f the surface soil. Structure class is choscn from Table 72-2. If doubts exist which class to choose, erodibility can be calculated for two different classes in or-der to receive the range in soil loss. Permeability is calculated as explained on page 96 and shown in thc following cxamplc:

In soil I , the lowest permeability corresponds to the deepest horizon within 80 cm depth and per-rncability class of the soil is calculated as mean permeability of all horizons to 80cm depth. In soil 2, horizon Btl has the lowest permeability and lies within 40cm depth. Therefore, i t is counted twice (sum of all classeslnurnber of horizons = 1615 = 3.2).

2. I n order to decide into which erodibility group a soil belongs, equations (26) to (28) must be solved. The group with the highest result is assigned to the soil.

3. Erodibility (K,,,,,,) is calculated for group 1 and 2 soils from equation (29) and (30), respectively, or can be read from Table 72-6. The regression for group 3 soils (equation (31)) was not significant. It is recornmended to use the maxinlurn erodibility K,,,,;,, = 0.026 found for group 3 soils. As 3 0 % of the group 3 soils had erodibilities between 0.01 and 0.026 and 70 %, of the soils erodibilities < 0.01, most of the group 3 soils are overestimated by this procedure.

soil 1: soil 2:

horizon depth permea- permea- tcml bility bility class

A 0- I0 very h igh 6 Bt l 0 - 0 ~nedium 3 Bt2 50-150 med~um 3

tllean perrneablllty ot soil: 4

horizon depth permea- permea [cm] bility bility class

A 0-10 very h ~ g h 5 A/B 10-25 high 4 Btl 25-60 low 2 X 2

BC 60-150 medium 3 mean pertmeability of 5011: 3.2

Tlrhlp 72-6: Corn1e,:cior, of K,,,,, ro K,,,,,, ti),- soils in gt.orq, I (old 2 (l/et-i~,cd fro~lr ecluntiorl,~ ( 2 9 ) rulcl (30))

Kcq<,

0.00 1 0.002 0.003 0.004 0.005 0.006 0.007 0.008 0.009 0.0 1 0 02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0 . I0 0 . I I 0.12 0. I3 0. 14

0.15 0.16 0.17 0 . I8 0. I 0 0.20 0.2 1 0.22 0.23 0.24 0.25 0.26 0.27 0 28 0.29 0.30

group 1 group 2

0.122 0.00 1 0. 125 0.002 0 . 127 0.003 0.129 0.004 0.132 0.006 0.134 0.007 0 . 136 0.008 0.138 0.009 0 . 14 1 0.0 10 0.14 0.0 1 0 . 17 0.02 0. I9 0.03 0.2 1 0.04 0.24 0.06 0.26 0.07 0.28 0.OX 0.30 0.09 0.33 0. I0 0.35 0 . I I 0.37 0 . 12 0.40 0.13 0.42 0.14 0.34 0 . 15

0.47 0.17 0.49 0. 1 8 0.5 1 0.19 0.53 0.20 0.56 0.2 1 0.58 0.22 0.60 0.23 0.63 0.24 0.65 0.25 0.67 0.26 0.70 0.28 0.72 0.20 0.74 0.30 0.76 0.3 1 0.79 0.32 0.8 1 0.33

K,,,,,

0.3 1 0.32 0.33 0.34 0.35 0.36 0.37 0.38 0.39 0.40 0.4 1 0.42 0.43 0.44 0.45 0.46 0.47 0.48 0.49 0.50 0.5 1 0.52 0.53 0.53 0.55 0.56 0.57 0.58 0.59 0.60 0.6 l 0.62 0.63 0.64 0.65 0.66 0.67 0.08 0.6'1

K,,,,, group 1 group 2

0.83 0.34 0.86 0.35 0.88 0.36 0.90 0.37 0.93 0.39 0.95 0.40 0.97 0.4 1 0.99 0.42 I .OO 0.43 1 .OO 0.44 1.00 0.45 1 .OO 0.46 I .OO 0.37 1 .OO 0.38 1 .OO 0.50 1 00 0.5 1 1 .OO 0 52 1 .OO 0.53 l .00 0.54 I .oo 0.55 1 .OO 0.56 1 .00 0.57 I .OO 0.58

1 .OO 0.59 1 .OO 0.6 1 I .OO 0.62 l .OO 0.63 1 .OO 0.64 1 .OO 0.65 1 .OO 0.66 1 .OO 0.67 1 00 0.68 1.00 0.69 1 .OO 0.70 1 .OO 0.72 I .OO 0 73 I 00 0 74 1 .OO 0 75 1 .OO 0.7t7

7.3 Thc topogl-aphic tnctol-

If the analytical data for the solution of the discriminant functions are n o t available the experience that volcanic ash soils (Andisols) are often in groi~p I , soils from acid basement rocks are frequently in group 2 whereas soils from basalt and other basic parent rock are often in group 3 can be used for a crude soil loss estimate.

7.3 The topographic factor (LS factor)

Soil enlsion is fiivoul-ed with increasing slope length and -gradient1 (cf. Chapter 4.3). The slope length factor (L) gives soil loss on a given slope length relative to soil loss on the USLE unit plot. The hctor for gradient ( S ) gives the ratio of soil loss on any given slope t o that of a 9% slope. The combined topographic factor (L'I'S) allows to adjust soil loss on a given slope length, gradient and slope form to that of the control plot. I t is calculated by (Wischrneier & Smith, 1978):

with

or

1 \lope length (m] rn slope length exponent 1-1 ;1 gradient [ " I

with \ gradient I c X ]

Thc slope length exponent (111) depend\ on the gradient and i5 \rnaller for low {lopes than for steep slopes (Table 73- 1 ).

' I GI-;ltlient can be ~ ~ ~ c a s u r c d by illclinon~elel-a ( 3 s \peci;~lly ccll~il,pccl compaisc.,. A icr.! ,irnl?le tie\,~ce to Inca\lu.e \lope - length and - g~-;lrlicnt i i illust~-~~tc.tl in Anrlcx 7. 1 .

On low slopes, m becomes smaller because low obstacles as rills and clods (surfiice roughness) produced by tillage slow down runoff. Thus, rnore water stays on the field for a longer time and water depth on the field increases. Tinie for infiltration is longer and at least part of the soil surface is protected against drop impact by a water layer. LS factors can directly be read from Figure 73- 1 . In order to a?just for less splash erosion on low slopes and thc PI-otective water layer, an additional correction of the annual erosivity is proposed on low slopes in the successor model of the USLE, - the Revised Universal Soil Loss Equation (RUSLE) (Renard et al., 1992). This correction factor can be obtained from Figure 7 1-2.

gradient f % 1

< = 0 . 5

0.6- 1 .O

1.1-3 4 3.5-4.9

> = 5

An exponent In < 1 shows that soil loss increases to a smaller extent than slope length. Nevertheless, in contrast to erosivity, soil erodibility, and slope-gradient, slope length can be influenced easily by Inan and is an important parat~leter for soil loss reduction. Slope length in the USLE is defined as the distance fro111 the point where runoff begins to the point where deposition occurs or where runoff enters a well-defined channel (Wischmeier & Smith, 1978). As demonstrated in Figure 73-2, the lower slope end may be presented by a small ditch or ridge along a field bot-der, a road ditch or a drainage channel. In case of small rivers, the slope end generally does not correspond to the river border because deposition generally starts earlier-. The upper slope end can be formed by the watershed boundary or by ridges. channels or deposition zones which limit a slope above. In general, the definition of an upper slope limit is met if no runoff from slope seg~nents above enters the slope.

m

0. 15

0.20

0.30 0.40

0.5

7.3 'fht: topographic tactor

slope - length [m]

Kunofi' volulne and velocity increase along the slope. This causes an increase of' soil loss pel- unit area with increasing distance down-slope. In order to calculate soil loss on a segr-nent of the slope, the slope is divided into a small number of segments i with equal length and approximately equal gradieiit. The segment on top of thc slope corresponds to i = 1 . The ratio of soil loss on each segmenl to soil loss of the total slope can be described by:

with Ai relative \oil lox\ of segment i I-] I (egmcnt nurnber N nurnber of segllients with cqi~al length 111 \lope length exponent

On a uliiform \lope of 6C/( ( m = 0 . 5 ) , for example, which wa\ divided into 3 \egments of equal length, the upper segment would provide 19%. the middle and lower 5egment 35 and 46 C7r of the total coil los\ on the \lope. Ke\ul ts of ecluation 35 for different \lope exponents and \egmcnt numbers are glven in Table 73-2.

A\ \o11 lo\\ I \ not equally distributed alol~g a \lope, \lope torm a\ well determine\ \oil lojs. On a concave \lope, the up- and inld-\lope part\ arc \teeper than the foot-\lope wherea\ 011 a convex \lope the foot-\lope ha\ a hrghcr gradlent Thc large runoff volutne which arrive\ down-\lope meet\ a low gradient on the cot~cave but a h~gh gradient o n the convex \lope. Submitting the \amc average gr-adlent, \oil lor\ on convex \lope\ I \ . therefore. Inore \evere than on concave \lope\.

7~1h10 7.5-2: Soil lo\\ of \lol~e \r~/lzollt, u,itll oylrcrl /e/lgtll or1 1~/11fot-111 \lope( /'rltrtr\~> to \or/ 1 0 5 s o f t o t ~ ~ l \Iol?o 10)- L / I ~ P I - C ~ I ~ ~ / ~ L ~ / ~ T / ? P I - of \cJglllcilr 5 clrlc/ c/iffe/-o/lt slope C ~ ~ J O I ~ C I ~ ~ ~ ( 1 x 1 ~ ~ 0 O I Z o~ll/(ltro/l (35))

Determirzation of the LS factor:

Rcad the LS lactor for un~lorm slopes fTrom Figure 71-I?'. In order to correct soil loss for the effect of slope form, an irregular slope is d~vided into :t small number of equal length segments with approximately unilorm gradient. LS values l'or each segment are cho\en from F~gure 73-1 by sing the slope length of the entire slope and the gradient of the segment. The \o der~ked LS value\ are wc~ghted by niultiply~ng them with the value\ from

Tablo 73-2. Summation of the prod~1ct5 gives the LS f ~ c t o r for the whole \lol'e

number of segments

3

3

4

5

Exa~nple: A 60 rn long convex slope is divided into three 20 In long segment\

wrth unifol-m gradlent of 10, 15 and 20% for the up-, mid- and down-dope \cgment (segn~ent\ 1.2 and 3 in Table 73-3). I-e~pectively. The LS factor for each \egrnent i \ cho\en from Figure 73- 1 by uslng a \lope length of 60 ITI and rl~c gradient of each \egnlent (column 3, Table 73-3):

27 ,I C ' O I I \ C ' ~ ~ ~ I O I I table Stom degree$ to ~)c'l-cerit 14 y l e n In Annex 7.1

segment number

I 2 1 3 - 3 I 3 - 3 3 I 3 - 3 4 5

slope exponent

m=0.5 m=0.4 m=0.3 m=0.2 m=0.15 0.35 0.38 0.4 1 0.44 0.45 0.65 0.62 0.59 0.56 0.55 0.19 0.2 1 0.24 0.27 0.28 0.35 0.35 0.35 0.35 0.34 0.46 0.43 0.4 1 0.39 0.37 0.13 0.13 0.16 0.19 0.20 0.23 0.24 0.24 0.25 0.25 0.30 0.29 0.28 0.27 0.27 0.35 0.33 0.3 1 0.29 0.28 0.09 0.1 1 0.12 0.14 0.16 0.16 0.17 0.18 0.10 0.19 0.2 1 0.2 1 0.2 1 0.2 1 0.2 1 0.25 0.24 0.23 0.22 0.22 0.28 0.27 0.25 0.23 0.23

Thew LS factors are weighted by the values from Table 73-2 for a \lope 2 5 74 (111 = 0.5) and 3 segments (column 4 in Tablc 73-3). The product\ of all \egriients (column 5 ) are \ul-nrned up and give the LS factor (= 3.1 1 ) fol- the ilope. Thi\ means that on a soil on this slope, \oil loss would be 4.11 tirile\ the (oil loss of the same soil on a 22.1 m long \lope of 9 C/r.

1

segment

I 2

3

Soil erodibility change\ on a \lope can be con\idered by thc \ame prc~cedure. Soil erodibility for each segment (colurnn 6) I \ multiplieci w ~ t h the weighted LS factor\ Sor the segments (co lu~nn 5 ) w h ~ c h give\ a KLS t'actor for the {lope of 0.7 (column 7).

Change\ of the crop and management factor are dealt alike a\ long a i n o deposition i \ induced by thc changes.

2

The slope length f ~ c t o r also allows the calculation ol' a n iax i rn~~m length if the maximum tolerable soil loss (T) is known:

If LS i \ known, the maximnm length can be chosen from Fiiguse 73-1. A tolerable LS value of 2 on a 10% slope, for example. yields a maxilnum slope length of 65 m in order to keep soil loss within the tolerable lirniti.

3

gradient

I0

15 20

4

LS factor

1.92

3.53 5.33

5

weighting factor

0 I9

0.35

0.46

{urn:

6

corrected LS factor

0.37

1.23

2.50

4.1 1

7

K factor

0.02

0.13 0 .2 1

corrected KLS factor

0.007

0. I h

0.53

0.70

7 4 The co\er ancl rnan,~gerllcnt (C') ICic~tor

7.4 The cover and management

The cover and management C of the USLE gives the ratio of soil loss on a cropped plot to soil loss on a barefallow control plot of identical size, slope length, gradient and soil. In contrast to the barefallow control plot where soil loss per unit erosivity (= erodibility) is supposed to be a constant (see Chapter 7.2). soil loss on a cropped plot is subject to changes over the year which depend on crop growth and rnanagement. After planting, the growitig canopy increasingly protects the soil surface while litter fr-om senescent parts l'alls 10 the ground and forms a tnulch layer. The weeds in the crop stand develop additional canopy covet- and act as mulch after weeding, i f left in the field. The protection of the soil surfi~ce depends o n the amount and cli~ality of coverage. Both are crop and management specific.

Howcvel.. ;in uncovered soil surface is only endangered if erosi\,e storms occur. Therel'ore, in order to calculate the influence uf crop cover on soil loss. the distribution of erosivity during the year must also be considereti. As thc annual erosivity distribution is site specific, the salne cropping system will cause different soil loss at different locations because of different distribution of erosive rains. The incan annual erosivily distribution is the11 assigned to thc different crop stages (Table 74- 1 ).

- -- -

crop stage

crop (scedbccl to germination) I0 (/( canopy cover until SO(;/( cover (ectablishrnent) 50 ' X to 75 (k canopy cover (development) 75 % canopy cover to harvest h;u-ve\t 1 0 next plowing or seeding

description

F S B

SB - 10

For each crop stirgc ( i ) a coil loss ratio (SILK) i $ cnlculatud ac \oil Ios\ of thc cn,pprcl plot (Acrop,) relative lo soil lo\\ of tlre control plot (Ahare,) during the same pel-iod:

rough fallow (F) al'ter primary tillage (coarse tilth) to seedbed (SB) preparation (= secondary tillage; fine tilth) ;~fter. seedbed preparation ~lntil 10 74 canopy cover of' the

Chapter 7

Acrop, SLR, = -- I - ]

Abare,

The soil loss ratios indicate the degree of soil protection by a specific crop stage. They are independent or site specific climate.

In order to avoid short term soil loss variations, the longterm mean soil loss of the barefdllow is used instead of the actually measured soil loss. I t is calculated by:

Abare , = R, : K (tlha) (38)

with K soil erodibility [t"'h/N'l'ha]

R1 mean erosivity during crop stage i [Nlh]

The term (R, 'I: K ) give\ the mean soil loss of the barefallow control plot during crop \(age i . In order to reflect the site \pecific ermivity distribution, the erosivity during crop stage i relat~ve to the annual erwivity i \ calculated:

Rrel, = RI

R I - I

with Rrel, proportion of annual R (relative erosivity) during crop stage i [N/h]

R mean annual erosivity [Nlh]

The \oil loss ratio\ for each crop stage of a rotation are multiplied by thc corresponding Kreli'\. Summation o f the products and \ub\equent division by the duration of the rotation rcsults in an avcragc annual C factor:

with n number of crop stages i per year j t duration of the rotation [a]

7.4 The cover and management (C) fitctor

An example for the calculation of a groundnut - maize rotation is given in Table 74-2. Mean planting date of groundnut and maize was 15th March and 5th August, respectively.

The crop stage duration is taken from the growth curves of the various crops (Annex 3.3). The Rrelils (column 4) are obtained from the mean annual distribution of erosivity (Table 74-3). Alternatively, the erosivity , - distribution can be estimated by calculating the relative rainfall distribution- . The C factor is calculated by summation of the product of the relative erosivity (colurnn 3) times the soil loss ratio (colurnn 4) for each crop stage (column 5 ) .

" Estinlation of the el-osivity distribu~ion from I-ainfall clistribu~ion I'or 18 \tations in Cameroon I - ~ ~ L I I ~ c ~ in a mean ancl maxiln~lm crrol. of 1.3 ancl 13.6%. respectively (Bresch. 1993).

3 4 5 6 I . relative soil loss column 4 * 5

erosivity erosivity ratio (Ci)

[relative to (Rrel,) (SLR) mean annual] I-] [-I

0.06 0.0 1 1 51 0 03

0.12 0 07 0 63 0 04

0 13 0.02 0 02 0 00

0.24 0 10 0 07 0 .OO

0 54 0.29 0.03 0.00

0.49 0.06

0.68 0 14 0 56 0 O X

0 70 0 I I 0 51 0 06

0 84 0.06 0 32 0 02

0 99 0 15 0 05 0.0 1

1 05 0.05 0 00 0 00

0 51 0.17

1 0 23

column 1 1 2

gronnd-tint

make

crop stage duration

[dl

S13 - 10 16

10-50 18

50-75 7

75 - H 3 7

H-SH 5 5

SR - SR 143

SR - I0 3 I

10-50 I9

50 75 8

75 H 44

H - SB 120

SB S B 222

Total 365

Chapter 7

High contribution to the C factor results from crop stages where little surface cover coincides with high erosivity: This was the case for crop stage SB - 10 of maize which received 14 5% o f the annual erosivity (Rrel, = 0.14) in Table 74-2 in a state of little cover. 35 % of the annual soil loss (column 6 in Table 74-2: 0.0810.23) occurred during this period. The soil loss ratios are generally high during the initial crop stages when cover is poor. An SLR > 1 for crop stage SB - I O of groundnut (SLR = 1.52) signifies that soil loss on the cropped plot exceeded the mean soil loss of the barefallow plot during this crop stage. This was due to compaction of the cultivated plot during the planting operation and sealing by early rains. On the barefallow plot, seals after a rain are raked (per definition) and no planting takes place.

The duration of' the crop stages shows the fister growth of groundnut which needed 88 days from seedbed (SB) to harvest ( H ) compared to 102 days for maize. Groundnut in the example received 49 Cr/c of' the annual erosivity (sum RRi of groundnut (SB to SB) = 0.49) compared to 5 1'/( for maize (sum RRi (SB to SB) = 0.51). The contribution of groundnut to the C fiictor was 0.06 (SB to SB) which corresponds to 26 C?c coriipared to 74 54 (0.17) for make. Thus, in the example, groundnut was more protective for the soil than maize. Protection mea\ures (e.g. mulch) would thu\ be more effectively applied during m a i ~ e cultivation.

7.4 The coccr and inanagerncnt ( C ) fiictoi-

~ I F , I I ~ , . C P ~ ~ S - ~ I F - ~ I ~ ~ C P X ~ S - ~ I ~ , I t , , kc c a c d c d a a G ~ . - r - P 1

r-1 rl rl P I CI r-I r i ri ri P I C I CI r~ rr 'cT; r-i PI cr rr rl PI CI cr rl rr T I PI rr rr rr

The difference of erosivity distributions is shown by the three sites from Camel-oon in Figure 74- 1 . On the coasi (Douala), the very humid ocean cli~nate has rather uniformly distributed erosivity during 9 months. The dry season lasts about 3 months. The inland of southern Carneroon (Yaounde) has two distinct rainy seasons separated by a dry spcll whereas in the north (Maroua) nearly all erosivity is concentrated in a few months.

To establi\h soil lo\s ratios for different crop\ and management systems needs field rneasurcments which are co\tly and time consuming. Sol1 lo\\ ratios for the rn~ijor c ~ ) ~ s - ' in the USA have been experimentally determined for a range of management options2'.

'' maize. aoybcans, coltona a~nall grain. iorghum, wheat. rycgraas, potatoes. paitul-t.. range uncl icllc land ancl forest

'' plow. notill. chi\el plow. contour tillugc, \(sipcrop. ridging. with and \vjtho~~t I I I L I ~ C ~ or rcsiduei

7.4 The cover ant1 management ( C ) factor

In order to calculate soil loss for further crops and systems. Wischmeier (1975) proposed to divide the influence of the cropping system into subf'actors. He defined a sub f~c to r for:

1. the influence of the canopy cover (c 1 ) 2. the influence of mulch or of' vcgctation close to thc soil \urliice (c2) 3. tillage and residual effects of the former vegetation (c3)

The C factor is calculated as the product of all 3 subfactors:

For trop~cal countrie\, the \ubfactor niethod i \ especially valuable becau\e for Inany cropr no experimentally determined data are available. A further complication i \ the large variety of small holder sy\tetns which are difficult to compare to American standard\ ( e . g hand till~tge, mixed cropping. heaping and bedding etc.).

Data for the rubfi~ctor calculation also are often not available but can rather ea\ily be collccted. The procedure for \ubf;ictor deter~ilination i \ \~~b\ecluently explained.

Subfactor

The intluence of canopy is calculated by (Foster, 1 982):

with CC, effective canopy coverage I-] He cf'fective canopy height [rnl

The canopy height effects thc velocity of drops falling of'f' the leaves ant1 thereby the energy of the drop irnpact on the soil. As drops may be formed by lower and higher leaves on a plant and drops from higher Ier~vcs may he intel-ceptecl by the lcaves below, the effective canopy height is used which represents an average valuc. For practical considerations, He is estitnated as:

with H,,,,,, mean height of the uppcrrno\t horizontal leaf of the plants in a crop \tand [rn]

Thc second variable in equation No. (42) - canopy cover - enter\ also as effective canopy covei-. D r o p which fall f roi i~ the canopy may not directly hit the soil s~ i r f~ ice but may fall on mulch material underneath without cau\ing \oil loss. Ac a cover fl-om ~nulcli is 111ol-c protective than from canopy, the effect of mulch is conjidered to 100 C/( wherea\ only the canopy cover with no mulch underneath, i.e. the effective canopy cover (CCe), is taken into account. It i \ calculated by:

with CC canopy cover 1 - 1 MC 111uIch cover [-I

I f canopy cover is 80'/(, for example, wit11 a mulch cover of' 20'A. the effective canopy cover is 0.8 :I' ( 1-0.2) = 0.64.

Subfactor Q

The intlucnce of lnulch cover (c2) can be calculated by (Yodel- et al.. 1992):

bquation No. (45), which rcllect\ the curve u\ed by Wi\chmeier & Snlith (1978) glve4 a consei-vatrve e\tlmate of the mulch etfect. Mea\urements by numel-ous other author\ (Duma\, 1965; Kaln7, 1989: NiII, 1993) revealed a hlgher efficiency (cf. F~gu re 74-2). Nevci-thela\, ~t I \

Indicated to continue u\ing eyuation No . (45) in order to arrive at a cautlou\ estimate of so11 los\ reduct~on by mulch. Some stmple iiicthod\ for \oil cover measurements are illu\trated and explained in Annex 3.2.

7 4 Thc co\ cl. and managcmcnt (C') 1';11,1o1-

Fi,qli/-o 74-2: / ~ ~ / / / i o ~ i c ~ o o f 111itlc.11 011 . $o i l 1 0 , s ~ (/,\ o \ ~ / l / i ( ~ t e ( / 173, ~ l i ~ o ~ - ~ i i t cl/itllor..\. S l l l? f i l c ' to~- c.2 gi\1c>s thc /*[rtio of' \oil 1os.s 0 1 1 ( I

~ ~ o \ ~ ~ ~ / - ~ ~ ~ ~ ~ l l o t to (111 /i11~*0\1Ol"c~d / l IOl .

mulch cover [%I

Subfactor c3

Not much data are available to determine subfactor c3 for tropical agl-o- \y\tem\ which account, for the re5idual effect of the prcviou\ vegetation. Ou.n mcasuremcnt\ re\ultcd in an average c3 of 0.8 for the 1st year after 11)rc\t I':illow and 0.4 after pl-ajs fallow (Table 74-4). A mean c3 01. 0.67 for the first 2 ye:lrs after gra,\ fallow can he c\timated from data of Kilewc & Mbuvi (1987) by the ratio of e m d ~ h i l ~ t y during the l in t 2 year\ and clr)dihility of thc 3rd to 5th year. For practical purpo\c\. the c3 value\ in T;ihlc 7 1 % are proposed. The influence of. the gl-ac\ fallow re\iduc\ come\

very clo\e to the residual effect\ dew-ibed by Wischmeier & Smith ( 1978) for turned \od. For the first year after plowed grassland they propo\ed 0.4. 0.45. 0.5 and 0.6 for crop stage\ SB - 50. 50-75. 75 - H and H - SB. re\pectively. The same cnlp \tages during thc second year were weighted by 0.8. 0.85, 0.9 and 0.95.

Figl,,.e 74-3: Sl,hfictor c . 1 a\ i17/lu~,rr>r~l 1 7 ~ 3 efi~c-ti~~~ crrr~ol)\, collri- liiril c . f i ) / ~

height ([1/iet. Fo~trr-, 1982 rrrlll Wi.$chtrlrier, 10751

effective height

effective canopy cover [%I

7 4 The cover and managetnent (C) factor

At the mornent, not enough data are available to calculate SLRs for the multitude of' tropical cropping systel-rls. Nevertheless, soil loss c~u1 be esti~nuted by the available data. In most experiments pi~blished in literature, soil loss was measured o n a cropped plot and cornpared to soil loss on an act.jacent control plot. Such data supply soil loss values for single cropping seasons or years without considering different crop stages. C f'actors which have a high variability due to a low number of repetitions can be calculated fro111 such data. However, some crops have been lested in several experiments and by comparing and averaging the results some reasonable ti-ends can be observed.

fallow type

fore\t

- -

gsa\$

- -

- -

Such annual C factors from ciif't'erent locations include an unknown, \ite specific variation caused by the el-ojivity distribution which can not be accounted for. By using them in different sites, the same soil loss will be predicted irrespective of the site specific erosivity distribution.

&st K factor c3 [t*h/N*ha] [-I

0.0 105 0.0 135 0.78

0.0886 0 . I I00 0.8 1

0.0 1 15 0.0236 0.49

0.0660 0.2000 0.33

0.1620 0.3450 0.47

T o estirnate the error caused by ignoring the erosivity di\tribution, C factors were calculated for a mixed cropping sy\tern measured in Cameroon by using erosivity distribulion curves from cites with an annual erwivity between 750 and 3231 N/h and mono- and bimodal rain distribution. The maximum difference was \r-nall (16'k ) (Table 74-5) (Petri. 1902). Furthermore, the liriiited ecological range of' 111o\t crops will also contribute to keep the difference within certain limits becauw very large differences in climate are generally also accompanied by a change in crop\.

The systern rain - canopy cover - soil loss can be regarded as sclf- stabilizing within certain limits. More rain after seeding or germination will enable faster and more growth provided that water is a limiting growth f'actor as i t is in many regions at the onset of the rain. Sucli an auto-regulation also favours similar annual C factors despite site specific differences in ternporal rain distribution. However, some crops can be found in very contrasting climatic zones. Groundnut and maize, e.g., are as well planted in the teainforest as in much drier environments. In this case it is safer to choose a C factor which was measured in a climate comparable to the site for which calculations shall be carried out.

site mean mean annual C factor annual rain erosivity

Bamonda 2470 1395 0.26 1 04

Bafia 1370 XI8 0.29 116

ecological zone

hulllid rainli~rest humid highland hurnid savannah savannah/ forest transition - -

Determination of the Cfactor

As previously clescribed, C factors can be derived from available cxperimental data or be calculated by using subfilctors.

I. Ilerivation of C factors from experimental data

Choosc a table from Tables 74-8 to 74-1 8 according to the main crop and look for a similar management system as your own in the descriptions:

table no. Ttrhle 74-8

title C /tlc.torcfila forest, hlr $11 errlel gnis 5 \ ~ g - c2tcltlorl (fclllon>~, p~r\t~rre) crrlcl \rthf(rc.tor, for rr.5 iclralrl e f ec - t~ Etart1171c. of trltrrrlcrti\~e ~ ~ o t h o t l fir deter- 171inlrtioll of C f c~~ tor for. the I st \ ucrr- for grtl,\ \, c - ~ \ ~ c . t - c 1-017s crt~d hri\/~ / ~ I I I o c L ' \ C f i r ( tors for. hcrtlcrtlcr C' fcic-for\ for /?lile~1/?/710 C ftrc tors foi c tr ,(r \ l r r

C fcic tot.\ foi. tur sc~rllcrrloorr,, ~ ) o - c j / ~ ~ ~ i t i l c.r-ol~c C fcrc.tot.\ fol 81-orrrltlr~rrt C ftrc tot-, fol- tlltrr;e C' fcrc tot-c. fot t?rrllc't N I I C I \oi.g/lro?l C ftrc.tot-5 /oi ~ l l ? l r i ~ ~ ~ l r-lc.c C' fcrc.to)-s for / l r r s c c~llrrrloo~ic ( r o l ) ~

Table\ 74-8 to 74-18 contain average values der~ved trom t l~c detailed data in Annex 3.4 (= \oLlrce refer\ to the line\ In the Annex table\). The detailecl C tactor\ given in Annex 3.4 are n o t advised for unexpericncecl il\er\. They were included Tor people who seek Inore inf'ormation uncl in orclet- to allow control and improvement of the data-ba\e and the derived value\ in the u\er wction ' . If you doubt about what to choose. take a11 average v;ilue.

With wme routine. col-I-ection\ for tiiffel-ences between de\cribed and own \y\tem can be applied. If, for example, your crop is e\pecially well developed, a \lnaller C factor \hould be chosen within the range given a \ 'extremes' .

If a notill option is not i nc l~~ded in one of the management systems, the C factol- for the clean tilled variant can be taken and multiplied by one of' the values in Table 74-6 which were derived frorn data in Table 34-8Annex:

?(' 11' >oil hayc li~e~-a,ur-c a\.ail;tble oil the sub,jec~ u hich is not inclucled in lhe tablcs 01' ,~ l l l lca 4.4. the auIl~or\ would be g~-a~cl ' i~ l for intlicutions or- a cop).

Chapter- 7

If a certain mulch cover is maintained in your systcm, you can choose the C factor for the system without mulch and correct i t by multiplying with a mulch factor (c2) from Table 74-7.

no.

1 2

If two crops are planted during the year, two C factors must be chosen from the tables. In order to arrive at an annual C factor, the two C factors and the periods between the two cropping seasons rnust be weighted according to the erosivity which they receive.

Example:

notill system

without residues with re\idue~

A rotation consists of groundnut which is planted during the first rainy season and is followed by plowed maize. In the dry spell between the two cropping seasons, the field is left to the natural weeds. The C factors for each crop and period arc lnultiplied by the relative amount of erosivity which falls during the respective period i.e. 30 '%I of the annual erosivity falls during groundnut cultivation, the dry season receives 10 %, maize 50% and the 2nd dry season another 10% of the annual erosivity. The sum of all products gives the annual C factor of 0.35:

C factor

m a n extremes 0.65 0.45 to 0.8 1 0.22 0. I to 0.4 1

literature (lines in Table 34-8Annex)

mean of no. 1 to 7 mean of no. 8 to 12

product

0.1 17 0.0 19 0.195 0.0 19

0.350

period

g r o u n d ~ ~ i ~ t

dl-y \enson

rnai7e

dry seawn

total

C factor for single periods

0.39 0.19 0.39 0.19

relative erosivity

0.3 0.1 0.5 0. I

1 .O

7.4 The cover and management (C') factol

I n order to judge a systcrn, not only the cultivation period i \ regarded but thc whole rotation which includes the fallow period. I f in tlie above example the groundnut-mai~e year i \ followed by two years of' grass fallow, the annual C Pactor is (0.35 + 0.19 (Table 74-93, line 2) + 0.004 (Table 74-9a, line 3))/3 = 0.1 8.

11. Derivatiorl of Cjactnrs by subfactors

The C factor can be calculated from subf'actors by:

with subfactor: c l influence of canopy cover c 2 influence of nlulch cover c 3 residual intluence of former vegetation

In order to derive c I to c3, the following information i s needed:

1 . the canopy cover curve and the canopy height to calculate \ubfactor cl (equation No. (42))

1. the ~nulch cover curve for s~lbfactor c2 (cquation No. (45)) 3. the residual influence of the fernier vegetation 3. the relative distribution of the annual ero\ivity

l'hc influence of notill can additionally be considered by lnultiplyi~lg with the notill subfactors in Table 74-6.

D I . The canopy curve is either deterini~~ed by imeasunng canopy coverage for the \y\te111 (methods in Annex 3.2) or by u\ing the typical growth curves given in Annex 3.3. However, it \hould be kept in mind that the variability included in the mean growth curves due to growing condition9 a id cultivars may be appreciable. Calculations and ~neasurenients can be crrrried out for crop stage periods (Table 74-1) or with a finer resolution i.e. 10 day or weekly intervals. The effective canopy coverage is calculated by equation No. 44. The canopy height can be rneawred or estimated from experience and is used to calculate the effective height by equation No. (43). With the effective height and the effective covcr si~bt'actor c 1 can be read f'rorn Figure 74-3.

D 2. Mulch coverage is determined frorn the mulch cover curve which shows mulch cover in the cropping system during the year. Subfactor c2 can be directly read from Table 74-7.

C> 3. Use c3 I'rorn Table 7 4 % . C::> 3. Calculate the mean relative erosivity distribution fro111 as tnany years as

available (as in Table 74-3). 11' no erosivity data arc available. Llse the relative rainfall distribution. Gcnerally. mcan curves are calculatecl by ~i~jcraging weekly or I 0 day intervals for as many years as possible.

mulch coverage [%I 0

subfactor c2 [-I

1 .OO

mulch coverage [%I 50

subfactor c2 [-I

0.28

7 4 The co\ er :und management ( C ) factor

7irhlr 74-8: A\,~r(rgr C t o . o o r hrr~h orld gf-lrr \ )e~ot~l f iotz (frrllo~rs. p ~ r ~ m r r ) lrrrd .\uhfirc-ior.s,/i,r tlrrir residrrrrl qtfr.ct.\

& - Z k b c3 0 G

.c, rl 2 Z Z r E n:& * g i a J c - y 5 *= c - 2 "Go E G&-LI+

gc! yt-

.-

> , s -m - - y = , S S .s ';: >

ZSC gs ",'5 1 1 '3 - '- - c / " E*? ";! c U c ' = c3 -a.= o '2 A A ,- , - r S a ? , &,zS Y , P La-=, G g c 3 u . c

- c 3 ; : g

" 2 p OJ u S U 3 L E La LC f 5 ,z 2 '" ?J

'3 c ,z O E . - 4 ' >, aJ - . - % Z + . L 2 - 0 2 5 O C C 0 E 5 " - .

7;rI)Io 73-9: E . I - ( I I ? I ~ / ( ~ o/ c117 c ~ / t ( ~ ~ - ~ ~ c r i i ~ ~ ~ I ? I P ~ / I O ( / f i ) t a t11e ~Ie~t~~1-11li11c/tio11 of' cr C' filc'tor- for- t11c 1 \t I ' O ~ Z I - of'gru.\.\, c-ollel- c'l-op,\ crrlcl hlr\h flrllo\c

coverlperiod duration erosivity SLR weighted SL during period (SLR*relative

erosivity)

[dl L N / ~ I 1-1 1-1 C-I

I SLI 111 365 1450 / 1 0.24 I

Thc C fiictol- lor the 1 st year i \ the \ u ~ n o f thc weighted SILK'\ (0.24).

description

I Iea\cx plncccl around rrunhx :unci on

conlour; \pacing 5 x 3 m o n contour

, .\Itcrnatively I'or :I young plantation I xt ycar)

2 :I\ aho1.c hut \pacing 7 x 3 m

3 ax above hut \pacing 3 x 3 m

1 4 a\ a h o w hut spacing 4 x 3 111

5 LL it11 con~plctc ITIUICII cover

I C factor 1 literature mean I extremes (lines in Table

34-2Annex) I

For- other- \pacings/ den\ities bctweerl 5 x 3 m (= 660 plantdha) and 2 x 3 111 (= 1650 plantslha) C' can bc taken from Figure 74-4.

7.4 'The coves and riiun:~gemcnt (C) I.;~ctol.

Fi,qlrt,r 74-4: C' ftrc.fot- for- diffi)r.clrlt Ocrtlcrrltr t/tvl.\itie.c.

Chapter 7

* 1

leve

l gr

ound

= n

o m

ound

s or

rid

ges

wer

e fo

rmed

3

C fa

ctor

m

ean

extr

emes

0.

36

0.12

to 0

.56

0.18

0.

02 t

o 0.

59

0.63

0.

57 to

0.6

9

desc

ript

ion

plan

ted

on l

evel

on

lev

el g

roun

d: ln

terc

ropp

ed

wit

h m

aize

. ric

e. c

owpe

a. s

oya,

phas

eolu

s or

coc

oyam

for

inte

rcro

ppin

g w

ith m

aize

w.

liter

atur

e (li

ne in

Tab

le 3

4-4A

nnex

) m

ean

of n

o. 1

-5

mea

n of

no.

16

to I

8 an

d 20

to

26

no.

27

cass

ava

1 2 3

mon

ocro

p

inte

rcro

p

subf

acto

r

Chapter 7

Chapter 7

7.4 The cover and Inanagetncnl ( C ) t l~cto~-

Tlrl?lr 73- 18: A ~,r.r-c~grd C,firc-tors for rlli~c-ellnneous c.r-op.\

no.

I 1 bean and jack bean 1 1

C factor m a n extremes

0.43 -

0.27 0.47 to 0.16

crop ' description

planted as monocrop o n contour 011 4 out of 8

Balnbara nut

bean\

3

literature lines in Table 34-7Annex

plowed: 3.5% {lope; jpacing 30 x 30 cm data from mung bean, red

cabbage

no. 1 no. 2 to 6

4 /chili I 1 0.33 - 1 no. 8 1 different Indone5ian soil types 1 0.6 - 110. 7

5 6 7 8

no. 18 to 20 planred along \lope notill. plu\ re\tdues of f.01-rner

I rnai7e. planted along slope 1

cotton - - - cowpea - " -

0.005 0.0004 t o 0.02

planted along slope 211d cycle plowed: without residues plus residues ol' former maire;

10 I I 11 13

I I residues humed: soya 1-esiducs I I

0.29 -

0.5 0.24 0.2 1 to 0.27 0.06 0.002 to 0.28

In\h potatoe lemon ~ r a \ \

1 4 15 16

17

no. 9 no. 1 I

no . 12 and 13 no. 14 to 1 7

papaya \oya

surficially incorporated rotation a\ ilbove but with notil I 0.04 -

-

- " -

sweet potatoes tobacco

wheat-soya

incorpol-atecl as above but all residue\ no. 33

wlthout cover crop

0.05 -

no. 34

0.22 -

0.434 -

notill without residue\ -

2nd cyclc

rotation on 12% slope. wheat

20 /wheat-m;~i~e / a\ :ibove, conventionally tilled.1 0.1 - 1 no. 35

no. 2 1 no. 22

2.1 -

0.26 0. l to 0.4 no. 23

no. 24 to 27 0.103 -

0.23 -

0.5 -

0.1 13 -

no. 36

no. 28 no. 29

n o . 30 and 3 1 no. 32

residues incorporated as above but notill 0.0 14 -

yam - -

(residues maintained) on heaps o n heaps; intercropped; with

residue ~iiulch

0.23 0.16 to 0.8 0.07 0.04 to 0.09

no. 38 no. 39

7.5 Thc cflicr of protective methods - Support practicc factor ( P )

7.5 The effect of protective methods - Support practice factor (P)

Protection measures must be adjusted to the possibilities and resources of each farmer. For nearly each individual situation a set of suitable physical and biological methods can assure sufficient soil protection.

7.5.1 Contouring, contour-ridging, tied-ridging

Generally speaking contouring means that all tillage operations and planting are curried out across the slope. Contour tillage and planting with mechanical tools leaves a roughness of the soil surface that is oriented across the slope. This may be considered as micro-ridges on contour. Such a formed surface redirects and retards the surface runoff. The efficiency depends on the degree of roughness (ridge height), the side slope of the tillage marks and the gradient of the overall slope. There is no clear limit between contouring and contour-ridges or bunds. The latter could be regarded as extreme roughness. Contouring in its original sense occurs ~ ~ n d e r mechanized tillage with crops planted in rows. In handtilled systems only planting can strictly be achieved on contour. Tillage with the handhoe is generally moving up-slope. The blade of the hoe is placed on contour but no continuous roughness is created. There is no information whether the roughness left by the handhoe marks can be compared to contour tillage.

Contouring reaches its maximum protectiveness on slopes between 3 and 8% (Tablc 75 1 - 1 ). It is less efficient on slopes below 3% where runoff' velocity is slow and a protective water mulch forms. On slopes above 8%. the protectiveness declines as the water storage capacity of the ridges becomes smaller with increasing gradient. For slopes > 2574, no protection is reached. P factors which were calculated from recent soil loss s t~~d ies (Table 4 I - 1 Annex) support the values in Table 75 1 - 1 :

: ' I The 111akimum hlopo Ienyrh lnay be incsoascd by 7Sfk i t ' se\idi~c co\cs ;~l'tt '~- planting reyul~~sly exceed\ SOc/(

slope [%I

1 - 2 3 - 8 0- 12

13- 16

17-20 21 -25

>25

A P f'actor of I for slopes > 25% was based on the assumption that a typical 15 crii high ridge in mechanized systerns retains no rnorc water on a slope of 25% (Foster ct al.. 1992). If the storage capacity of the ridges is large enough to prevent overf ow, maximum slopc lengths need not to be applied. As the effectiveness of contour ridges depends o n their- storage capacity, it must also depend o n storm size. In locations with frequent I~irgc stor~ns. contour-ing is less effective than in locations with smaller storms. Theref'ore. thc 10 year storm volume is chosen for ridge design purposes (Foster ct ul.. 1992). If thc furrows can only carry the rnaximurn 2 year storrn, length limits arc applicable ( Wischmeier. & Smith, 1978).

The procedure to estima~c the intlucnce of ridges applied by the USLE gives a rough estimate and does not allow to distinguish between dil'fcrent ridge heights. Ridges. however, play a11 important role in tropical ~igro-systeriis. A more refined estimation is possible by using the P factors in Figill-c 751 - 1 i~sed in the Kcviscd Universal Soil Loss Equation (RUSLE) (Renard ct a].. 10'32). The curves were calculated on the basis o f a 10 year storrri of 86 to 190 111111. hydrol~gic soil groirp C - I and clean tillage for row

P factor for contouring

0.6

0.5 0.6

0.7

0.8

0.9 I .O

'' hydl.olori~~ soil g1-oi1p C includes xoil\ with low inl'ilts~~tion sate5 wllcn wet. ~no \ t l \ ~ i t l l i~npendinp luyors or ~noclcratly fine roxturc (IISIIA. 1073: SCS Nalional Enginerr-ing I 1;untlhook )

maximum slope length Em]**

122 9 l

(3 1

24 1 X

15 13

7.5 The el'i'cct of ~,l.orccri\e ~nerliod\ Support p~.aclicc I'itcror ( 1 ' )

F~g~~ t - c l 751-1: P jilc.for-s f o r . difor.rrlt ritlgo hc>igl~t\ 101- trt-clri\ r~,rfll 10 \.t>tit. \to/.llls bct\c.cerl 86 L I I I ~ I 190 111111 (111ii / ? \ ' ~ / I - O ~ O , ~ I ( . \oil qr-0111) C ( I . o rtrr- e f ( I / . , 19913)

CI-opi u it11 n o c ~ L ~ ~ I - tlnd mlnlmum roughneis (cover-rnanagemcnt condit~on 6) . In Weit Afr~ca iuch I0 year itorms are found approximately In the belt h c t ~ e e n I h northern latit~ide (north Senegal, north Burkina 1-aio) and the co~ri t line (Figure 75 1-7). For other areas no infonilation on the 10 year itorm M ; ~ i found.

I2nr areas with a lower 10 year storm volume (e.g. north of I6 Intitode) the I-idpc cfficicncy will he underestimated by Figulc 75 1 - 1. whereas f-ol- ai-c;is with higher 10 year stol-m volume ;in overestimation is possible. Regarding the soils. hydrologic soil gnliip C may be applied to the Aridisols. Alfisols. Inccptisols and VCI-tisols o i the semi-humid to semi-arid/ arid area. For the LJl~isols ancl Oxisols of the humid to s c ~ l ~ i - l ~ u ~ ~ ~ i d areas thc cfficicl~cy i s underestimated by hydrologic soil groilp C.

Chapter 7

Figrrro 75 1-2: Isoh?.ctc~ for - tlrr 10 ?vor { tonn ~lolirt?~e (CIEH. 198.5)

Contouring and contour ridges are mostly not exactly on contour. In practice, they have a side slope either accidentally or in order to evacuate excess water. For side slopes < 0.5%, all soil is deposited in the furrows (cf. Chapter 4.3). For steeper side slopes the efficiency of contour ridging is reduced. P factors corrected for side slope effects (Table 75 1-2) W ~ I - c calculated by (Foster et al., 1992):

P, = P,, + ( 1 - P,,) ( \ , 1 5 , ) ".i

m it11 P, P I'actor I'or ol'l'-gn~de contouring

P,, P l'actor for on-~ra t le contouring

, grade along the furrow\ (sine 01' slope angle)

\, \teepness of thc I~und (sine of \lope angle)

Mea\urcd \slue\ for ridges are given in Table 4 1 -'Annex, line I and 2. The table al\o indicate\ the di\astrou\ effect of up- and down-\lope ridges ( P = 0 0 to 4.3). practical purpo\e\ a P factor of 2 can be ~ ~ \ e d for thi\ practice.

slope [%I

4

8

12

16

uncorrected Pfactor

I 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0 . 1

I 0.9 0.8 0.7 0.6 0.5 0 4 0.3 0.2 0 . I

1 0.9 0.8 0.7 0.6 0 .5 0.4 0.3 0.2 0.1

1 0.9 0.8 0.7 0.6 0.5

corrected P factor

side slope of furrows [%I 0.5 1 .O 1.5 2.0 3.0 5.0

1.00 1.00 1 .OO 1 .OO 1 .OO 1 .OO 0.94 0.95 0.96 0.97 0.99 I .OO 0.87 0.90 0.93 0.95 0.98 I .OO 0.8 1 0.86 0.89 0.92 0.97 1 .OO 0.75 0.8 1 0.85 0.89 0.96 I .OO 0.68 0.76 0.82 0.87 0.95 I .OO 0.62 0.7 1 0.78 0.84 0.94 1 .OO 0.56 0.66 0.74 0.8 I 0.93 I .OO 0.49 0.62 0.7 1 0.79 0.92 1 .00 0.43 0.57 0.67 0.76 0.9 1 1.00 1 .OO 1.00 1.00 1 .00 1 .OO I .OO 0.93 0.94 0.94 0.95 0.96 0.98 0.85 0.87 0.89 0.90 0.92 0.96 0.78 0.8 1 0.83 0.85 0.88 0.94 0.70 0.74 0.77 0.80 0.85 0.92 0.63 0.68 0.72 0.75 0.8 1 0.90 0.55 0.6 1 0.66 0.70 0.77 0.87 0.48 0 55 0.60 0.65 0.73 0.85 0.40 0.48 0.55 0.60 0.69 0.83 0.33 0 42 0.49 0.55 0.65 0.8 1 1 .OO 1.00 1 .OO 1 .OO I .OO I 00 0.92 0.93 0.94 0.94 0.95 0.96 0.84 0 86 0.87 0.88 0.90 0.93 0.76 0.79 0.8 1 0.82 0.85 0.89 0.68 0.72 0.74 0.76 0.80 0.86 0.60 0.64 0.68 0.7 1 0.75 0.82 0.52 0.57 0.6 1 0.65 0.70 0.79 0.44 0.50 0.55 0.59 0.65 0.75 0.36 0.43 0.48 0.53 0.60 0.72 0.28 0.36 0.42 0.47 0.55 0.68 1 .OO 1 .OO 1.00 1 .OO 1 .OO 1 .OO 0.92 0.93 0.93 0.94 0.94 0.96 0.84 0.85 0.86 0.87 0.89 0.9 1 0.75 0.78 0.79 0.8 1 0.83 0.87 0.67 0.70 0.72 0.74 0.77 0.83 0.59 0.63 0.65 0.68 0.72 0.78

7.5 'l'he cffect of protective methods - Support practice factor (P)

slope [%]

16

20

24

28

uncorrected P factor

0.4 0.3 0.2 0. I

1 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1

I 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1

1 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0 . 1

corrected P factor

side slope of furrows [%] 0.5 1.0 1.5 2.0 3.0 5.0

0.5 1 0.55 0.58 0.6 1 0.66 0.74 0.42 0.48 0.52 0.55 0.6 I 0.69 0.34 0.40 0.45 0.49 0.55 0.65 0.26 0.33 0.38 0.42 0.49 0.61 1 .OO I .oo 1 .oo I .oo 1 .OO 1 .oo 0.92 0.92 0.93 0.93 0.94 0.95 0.83 0.85 0.86 0.86 0.88 0.90 0.75 0.77 0.78 0.80 0.82 0.85 0.66 0.69 0.7 1 0.73 0.76 0.80 0.58 0.6 1 0.64 0.66 0.70 0.75 0.50 0.54 0.57 0.59 0.63 0.70 0.4 1 0.46 0.49 0.52 0.57 0.65 0.33 0.38 0.42 0.46 0.5 1 0.60 0.24 0.30 0.35 0.39 0.45 0.55 1 .OO 1 .OO 1 .OO 1 .OO I .OO l .OO 0.9 I 0.92 0.93 0.93 0.94 0.95 0.83 0.84 0.85 0.86 0.87 0.89 0.74 0.76 0.78 0.79 0.8 I 0.84 0.66 0.68 0.70 0.72 0.74 0.79 0.57 0.60 0.63 0.65 0.68 0.73 0.49 0.52 0.55 0.58 0.62 0.68 0.40 0.44 0.48 0.5 1 0.55 0.62 0.32 0.37 0.40 0.43 0.49 0.57 0.23 0.29 0.33 0.36 0.42 0.52 1 .OO 1.00 1.00 1 .00 1 .OO 1 .O0 0.9 I 0.92 0.92 0.93 0.93 0.94

0.83 0.84 0.85 0.85 0.87 0.89 0.74 0.76 0.77 0.78 0.80 0.83 0.65 0.68 0.69 0.7 1 0.73 0.77 0.57 0.60 0.62 0.64 0.67 0.72 0.48 0.52 0.54 0.56 0.60 0.66 0.40 0.43 0.47 0.49 0.53 0.60 0.3 1 0.35 0.39 0.42 0.47 0.54 0.22 0.27 0.3 1 0.35 0.40 0.49

Determination of the P factor for contouring and contour ridging

Contouring and contour-ridging

a. For simple tillage and planting of row crops on the contour, use P factor\ in Table 75 1- 1 according to the slope. If contouring and ridges were established with side slopes, enter Table 75 1-2 for correction.

Example:

For a contoured \lope of 14 % with a side \lope of 3 C/c a P factor of 0.7 was chosen from Table 751-1. The side rlope effect is con\idered by entering Table 75 1-2 for a 12 C/c and a 16 (;/c dope (P corrected = 0.85 and 0.83, re\pectively) and interpolating the two values to a 14% slope ( P corrected = 0.84).

b. For ridges with a height of more than 1 0 cm, choose a P filctor according to slope and minimum ridge height l'rom Figure 75 1 - 1 . Correct it for thc effects of an eventual side slopc as cxpluined in a..

c. If ridges do not persist during the entire year but are rnounted, for examplc, during the growing pesiod of a crop and levelled during harvest. the P factor can not be fully credited. In this case only the soil loss ratios of those crop stages are rnultipliecl with the P factor for ridges for which the ridges are intact. For the crop stage periods without ridges P equals 1 .

Example:

M a i ~ e is planted on level ground on a 10% slope. When canopy covcr reaches I O C/(, 15 cm high ridges are mounted with a 1 (k \idc \lope. The term C x P f'actor is calculated like in Table 75 1-3.

7.5 71'lie effect ol'pro[ective method - Support practicc factor ( P )

The erosivity ratios were taken from Figure 74-1 from the 'north' curve a4suming that crop stage SB - 10 4arted on the 130 day. The duration of the crop stage periods and the soil loss ratios were taken from Table 74-2 (column 2 and 5 ) . A P factor- of 0.36 corresponds to 15 cm ridges on a 10 C/c \lope (Figure 75 1 - 1 ). The corrected P factor i \ interpolated fi-om Table 75 1-2 ((0.59 + 0.54)/2 = 0.57). The I-esulting C x P factor is 0.075 which compares to C x P = 0.063 if the ridge4 would be credited for during the entir-e cropping cycle.

Tied contour ridges

6 C x P

column 2 x 3 x 5 0.0 1 1

0.020 0.01 1

0.0 15

0.0 1 8 0.075

Values in literature for soil loss with tied contour ridges range between 0.21 and 0.035 o n slopes between 4.5 and 7 % (Table 41-2Annex). Ties between the ridges have no effect i f the ridges are perfectly on contour. If the ridges have a side-slope, ties will stop or reduce the sideways evacuation of runoff'. A reduction of the ridge efficiency due to side slope will therefore be much less. For practical purposes it is proposed to choose a P fhctor as for contour ridging from Figure 75 1 - 1 and to dismiss the correction for the side slope.

7.5.2 Bufferstrips

4 3 1

Bufferstrips are < I to several m large strips within fields mostly composed of quick growing species or natural vegetation. They are laid out on contour- in orcler to decrease runoff velocity thereby causing deposition of suspended sediment. The efficiency of bufferstrips depends on the quality of the strip (stsip widths. vegetation density). ils age and its position on the slope. The runoff which asrives at the upper bufferstrip end has a certain transport

5 2 crop stage

SR - 10

10-50 50-75 75 - H H - SR

SB - SR

erosivity ratio 0.02 0.07 U.06

0.51 0.36 1.00

corrected P factor

1 .OO 0.57 0.57 0.57 1 .OO

Total :

soil loss ratio 0.56 0.5 1 0.32 0.05 0.05

P factor

1 .OO

0.16 0.36 0.36 1 .OO

capacity and sediment load. Runoff velocity and tran\port capacity are reduced in the bufferstrip by the higher hydraulic roughness and friction exerted by the vegetation. Additionally, part of the runoff will infiltrate within the strip which has generally a higher infiltration rate than the adjacent cropped soil. If the transport capacity becomes less than the sediment load, soil i5 deposited in the bufferstrip. However, if runoff leaves the strip on the lower end, it rnay regain \peed and pick up new sediment from the cultivated \trip underneath. Thus, the mo\t t'avourable case i \ . i f no runoff leave\ the strip.

Planted bufferstrips generally do not reach their full protection efficiency during the first rainy season 01- the first year while the plants' root and canopy system is still establishing. P factors for the second year are. therefore, often lower than for the first year. There is also some evidence, that the efficiency of bufferstrips may decrease with increasing sedimentation in the strip (Barfield & Albrecht, 1982). This will depend on the growth habit of the strip vegetation (e.g. canopy or twig density close to the ground) and how tist the vegetation can grow up and cope with a heavy sediment load. If large amounts of sediment arrive at the bufferstrip, a small terrace will form within a couple of years.

A special case of bufferstrip are the riparian bufferstrips along rivers which prevent sediment entry. However, they do not prevent soil loss from the slope above. An indication for the effectiveness of riparian bufferstrip5 with increasing strip width is given in Figure 752- 1 .

7.5 The el'1'ec.l of protective method\ - Supporr pl-acrice I.21cro1- ( P )

Determination of P factors for bufferstrips

F~glrr-(1 7-52- I : P flrc*tor\ for ripnr-irrrl l? l i fo~~tr - ip \ o/ I ~ ~ / / C - ' I . C ~ I ~ ~ 11,illt11\ ~ I ~ i . i \ ' ~ ' ~ l f l . o ~ l ( I 8(jc lo/)^ C L ' / ~ / I 1/11 11~7/1~11/1 \~~l i l i lo l l t l o ~ l ~ l of cScl. I t/~tl Oliffor lt>11gtl1 (SCIIIII ICIC~I. c4; A ~ i o r \ ~ t ~ ~ / l ~ l , 1992)

a. Bufferstrips

a. Systc~natic trials for the effect of hufferstrips are still deficient. Somc P factors as calculated by the RUSLE are givcn in Table 752- 1 .

0.95-

b. Further P factors can be taken from the experililerltally determined P factors in Table 4 I-SAnnex for comparable situations.

buffer width [m]

0.75-

z 5 0.70 m 2 0.65

0.60 0.55

0.50

0.45

0.25 y=369*? -304 1

----. \% i - 1 ! -

i i \ r _ _ _ - _ . . - - -

\

1 I I 1 j I l l , l I l l I I

1 1 1 1

0 1 1 1 1 -

5 10 15 20 25

Ttrhlcl 752- I : P Jilc.to~-.s ,fol- hziffi.~-.str-ip.~ crs ( Y I I ( . L I I L [ ~ P C I 17). f / 7 ~ RIJSLE (Fo.vtc11- c)tcrl., 1992)

I percent of slope I position of strip*' I P factor

' ' 1 kw example 0.3-0.5 means f l i ~ ~ t the strip start\ after 4054 of the \lope Icngth doun- \lope and ends after half' 01' the slope lcngth

covered by strip 2054 ~n 2 \ti-lp\ 10% in 2 \trip\

h. Riparian bufferstrips

P factors for riparian bufferstrips can be chosen from Figure 752- 1 .

0.4-0.5 and 0.9- 1 .O 0.35-0.40 and 0.65-0.70

7.5.3 Contour bunds and heaps

0.67 0.7 1

Results from trials indicate the different efficiency of stone-bunds and earthen bunds. Runoff occurring on the uppermost side of LL field picks up e lo city and sedi~nent. Arriving at the first bund it is co~npletely stopped by :in earthen bund or slowed down by a stone bund. The sediment is deposited i l l front of thc bund. Using eat-then bunds, the process is repeated between first and second bund. second and third bund and so on. However, stonc- bunds, which are permeable, allow a part of the writer to pass the bunii. This water regains velocity and transport capacity on the lower side of the bund and entrain new sediment in addition to the runoff' produced o11 the lower side itself.

The ci'ficiency of earthen bunds is thus much higher in the first year compared to stone-bunds (Table 4 1 -4Annex). However, the data indicate that in the second year the effect of the two types becorncs similar. The stone- bunds become less permeable due to sediinent which progressively fills and clogs the inner space of the bunds. Thc earthen bunds apparently became less ef'f'icient duc to holes which occur in the bund or to the lowering of the bund by raindrop innpact or overtopping. The decreasing efficiency of the earthen bunds also make higher rnainte~iance necessary compared to stone-bunds.

7.5 Tlir rf'l.rc,t ol'psotc.ctivt~ metliod~ - Support ~?s;rctic~r t;rctol. (1 ' )

Ilctermination of P factors for contour bunds and heaps

a. Stone-hunds and earthen bunds

Thc P kictors in Table 41-4Annex can be used for first and subsequent years in comparable situations. As a bund can be compared to a ridge a sirnilat- slope influence on the efficiency of bunds is assumed. Therei'ore, i r is ~~roposed to multiply the available P factors with the ratio of the P tictor for a 15 cm ridgc (Figure 751-1) o f a given slope to the P factor on a 3 C/c slope (Table 753- 1 ). If' maintenance is regularly carried out on earthen bunds each year. the P fzlctor for the first year can also be used for subsequent years.

11. Heaps

Not tnany data are available for the specific effect of heaps on soil loss. 7'hc \.cry variable influence is shown by the data in Table 41-3Annex. The influence of' heaps depends on their arrangement on the slope (up and down- slope. on contour. in cluinti~plcs), their sire and height which depend on slope and top soil dcpth (cf. Chapter 4.4). The data in Table 41-3Annex should only bc used if tlie intluence of mounds is not yet included in the C f'actos (e.g. for yam in Table 74-1 8 it is not necessary to use additionally a P factol- for heaps)).

7.5.4 Ditches and terraces

slope

[%I 0

I 7 - 3

4 5 6

7

8

9

10

1 1

12

13 14

15

Hillside and drainage ditches (cl'. Figure 44%- I ) clecrcase soil loss by reducing tlie erosion effective slopc length (L fiictor). Thus, the clown-slope acceleration of runoff and its concentration is controlled. Slope length in the LISLE is defined as that part of a slopc where no major deposition is

In the case of drainage ditches, the sediment charged water \pill\ 11-ecly illto the ditch. The slope length is the distance between the lowcr ude of a ditch t o the ilppcr {ide of the next ditch. For a Fanya Juu type terrace (cf. F~gure 448-1 ). deposition begins in front of the excavated ridge and \lope Icngth is calculated from the lowei- end of the ditch to thc area where dcpo\ition begins in front of the next terrace.

Terraces not only reduce slope length but al\o gradient which i \ con\idered in the LS factor. For \loping bench terraces (cf. Fig~tre 449-2). the

ratio

C-I 2 12

1.67 1.21 1 .OO

0.79

0.70

0.6 1

0.58

0.55

0.55

0.56

0.57

0.58

0.59

0.6 1

0.65

slope [%I 16

17 I 8 19 20 2 1

- 3 - 3 23 24

2 5

2 6

2 7

28

29

30

ratio

[-I 0.70

0.76 0.82 0.9 I

1 .oo 1 . 1 1

1.2 1

1.33

1.45

I .64

1 .X2

I .9X

2.15

2.27

-.- 3 55

width of the bench is considered as slope length for soil loss prediction. The soil eroded frorn the bench reaches the toe drain where it is either deposited or washed off into the waterway anti out of' the field.

The soil deposited either in front of the Fanya JLILI terraces, in the ditches or in the toe drains is not yet lost ttom the field. It can be regarded as di\trihutcd within the field. Excavation of the ditches will partly put it bach on to the field. The amount of soil which i\ actually transported out of the f'icld relative to the amount o f soil erodcci is callcd the sediment delivery ratio. It varies with the 51de-slope ot the ditches (Foster & Highfill. 1983) from 0.1 for level clitchcs to I for ditche\ with a side-\lope of 1 L/c (Table 754- 1 ).

.,. I including tcrr;lces with i~ndergrounti outlet$ ":1 froin Wi\ch~lleier Kr Smith (1978): all other values from

STIR = 0.1 ';:c7.h4g: g = \ide \lope 1'4 I ' ' '3 ner ero\ion may occur in the channel\ depending on flow hydraulich and crodihilily ol' lht.

ch:unl~rl\; if c1i:uinel crohion occur\ SDR > I

terrace grade [%]

clo\ccl ou t l e t '

0 0 . 1

0.2 0.4

0.6

0.8 0.9

> 0.9- 3

Determination of P factors for terraces

sediment delivery ratio 0.05 2

0.10 0 . 1 3

0.17 0.29

0.49

0.83 I .OO -

Calculate soil loss A = R'''K*'L'''S"C'I'P I'or each terrace by using slope lengths as explained above. The gradient is either the slope-gradient in the case of hillside ditches or [he bcnch gradient in the case of bench terraces.

P may be compmed ot contouring factor if tillage and planting are carried out on contour (cf. Chapter 7.5.1 ) which needs to be multiplied by the \eciiment delivery ratio fi-om Table 754- 1 .

Example:

A slope is divided into I0 reverse-sloped ten-aces 10 m wide and 100 m long with a bench gradient of 5% and a side-slope of 0.4%. The benches are cropped to cassava with nlaize arranged on contour. Further data ( R = 500 Nlh; K = 0.15; LS = 0.3 1 (from Figure 73- 1 ); C = 0.2 1 (from Table 74- 12)). P SOI- contouring is 0.5 (from Table 75 1 - 1 ) and the sediment delivery ratio for a 0.4% side-slope is 0.29 (Table 754- 1 ) . Thus soil loss for this situation is A = 500 ;:: 0.15 *: 0.3 1 'I: 0.2 1 'I: 0.5 $: 0.29 = 0.7 1 tlha. Each terrace has 0. 1 ha. All terraces together would thus loose 0.7 1 t.

7.6 Soil loss tolerance limits

Soil loss tolerance limits define the soil loss rates which are tolerable in order to maintain the soil's diverse functions during a specified time. The ef'f'ect of' soil loss depends strongly (311 the type of soil. Soil loss always implies a loss of nutrients and structural components (clay. organic matter) which are enriched in the sediment. The soil profile is shortened. rooting depth and water storage capacity decreased. On very deep, homogeneous soils. the damage will be less than on soils with unfiivourable layers or solid rock close to the surface. Cornpared to less weathered soils of the temperate and semi-arid zones, loss of surface soil is more severe on highly weathered soils whose nutrient storage and availability depends largely on the organic tnatter of the su r f~ce soil while the subsoil fertility is low.

The yield decline associated with erosion depends also on the crop. Mbagwu et 31. (1984) showed that removal of 5 c111 of soil reduced ~nuize yields by 95% on an Ultisol whereas on Alfisols mean yield decline was only 52%. Cowpea yields were only reduced by 63 and 22% on Ultisols and Alfisols. respectively. In own measurements, maize yields on an Ultisol were rero after four years of erosion had stripped off the surfice soil. On an Alfisol, however. which had been exposed for 8 years, a poor yield was still possible.

7.6 Soil lo\\ tolcrancr l i l ~ ~ i t \

Ideally, the soil loss rate should not exceed the soil formation rate of the parent material. Most reported annual weathering rates for tropical climates are below 500 kglha (Table 76-1) which is far below agricultural soil loss rates. The lost productivity is irrecoverable by external inputs. This is even rnore true ['or small scale farmers in developing countries which do not have the necessary inputs to mitigate soil damage.

Tolerance value\ alco depend 011 the intended purpose of (oil con\ervation. In general. the purpose for erosion control will be agricultural production. However, in flood prone areas water retention can be the more important goal whereas for the municipal authorities wdiment damage\ on road ditche\. waterway5 or in the public sewerage 5ystem may be decicive.

In order to formulate tolerance limits, a decision about a seasonable conservation time ~iiust be taken. This is rnore a political and social clucstion than a scientific orit.. Can we tolel-ate a 50 C/c yiclcl decline in 50 years. 100 year-s, - or arc we still responsible h r the well-being of 0111- ancestors in 5 0 0 or I000 years? An answer to this clucstion must be founci in order to calcitlate a mean. annual tolerable soil loss.

country

Cenlral

AI'r~ca

Puerto

KICO

Kcnq a I , Kenya

1 OSA

climate L sub-humicl

humid

humid

1 arid

parent annual literature

rate [kgka]

Toleraiice values. - first developed in the USA. were based on e\timates ol' a sul-face +oil formation of 2.54 cm in 300 to 1000 year\ (Bennett, 1928). This e\timatc wa\ later changed to 2.54 cm in 30 year\ u hich i \ about 1 1 t/ha"'a (Pimcntel et al., 1976). Thi\ wa\ thc b:i\i{ fol- 5ettiny

150-300

< 150

300

Dunne et al. ( 1978

Dunnc el al. t 1078)

KII-hhy i 1980)

maximum annual soil loss rates in the USA to 1 1 tlha. Thus, setting of this and all subsequent values was more based o n expert judgemelit and PI-actical considerations than on scientific data.

Tolerance values for tropical soils have not yet been formulated on a n international level. However, some countries use tolerance values and propositions were made by some ~luthors. A summary of existing values is given in Table 76-2.

literature

Chin 61L Tan ( 1974) Central African Federation after Hudson ( 1986)

, - -

Nyagumbo ( 1992) Hudson ( 1986) La1 ( 1980) Lal ( 1983) Hurni ( 1980)

applied for

tropical \oil\

clayey \oils

5a1idy \ ( M I \ undy \0115/Zimbabwe \hallow. erodible \oil\

\hallow h~ghland mils trop~cal soil\ Eth~op~a

tolerance limit [t/ha*a]

15-25 I I

9 5

2-5 2.5

0.2-2 2

Annex 1 Rainfall and erosivity

Annex 1.1 Erosivity for single sites

Check if your site is included in Table I 1 - 1 Annex. Erosivity was directly calculated for these sites. Also verify if erosivity data are available fl-orn the meteon)lopical services or research stations-';. If your locatiotl is near to one of the sites in Table 1 I-1Annex and has the same annual rain volume. you may as well use the erosivity given in the table. If the rain volurne of your site varies within 10% of a nearby station, you can extrapolate the erosivity value linearly. The error for stations between 400 and 4000 mm in Table I I - 1 Annex is supposed to be less than 6 '/r (Figure 1 1 - 1 Annex).

E.t-c1111/710 ,fi)ta (~.~tl-eri~oll/tiorz: 7/10 ,site ill Tc//)/o 11-1 Ar7r1o.1- t - c ~ ~ 1 i 1 ~ ) . < 1330 rllrll/cr (?/' r.clirl ,\,it11 trri c~t-o,si\~it>' of' 1249 N/lr. Yorrt- o ~ ~ i l .site rlcjtrt-1)). 1ltr.c

1 .(Of1 I I I I I I ~ I . E!.,,,,/~)I- ?,or rt- .site cwrl he1 c~trlc~~1krrpcl1~~ (1249/1440)'!'/ 300 = 1 I28 N/I1.

+ Bresch (1 993) --..----. error by 10% less - error by 10% more

annual rain volume [mm]

Erosivity for single site3

Tcrhle 1 I - 1 Atztzc1.r: Erosi\?itj?, ruirl ~lolunze, mea.c.urerner~t prriotl Ji)r ~irzgle sites

country site erosivity measure- mean literature ment annual

period rain [Nlh] la1 [mm 1

Algeria Go~~ra r i 139 3, 555 Marour ( 1997) - - tleriz 53 2 338 - .. - - - Madjoi14 5 0 1 330 - " -

- - Sidi Moharncrl Cherif 53 7 338 - " -

Burkina Faso Bobo-D~oi~litsso 09% 58 1 150 Galabert & Mlllogo

i 1973):"2 - - Dori 468 4 7 530 - " -

- - Fudu- N'Gourma 772 48 X O O - " -

- - Faritho- Ba 84 1 6 10x3 - " -

- - Gaollit 1076 53 I240 - " -

- - Gunab 599 5 709 Roose ( 1975) - - Mogteclo 656 6 754 Galabcrt & Millogo

( 1973 ):,:2 - - Niangoloko 1161 1 3 1330 - " -

- - Oi~agadi)i~gou 763 2 1 8x0 - " -

- - Ouahigouq a 607 49 700 - " -

s. ' . . ,rrl,c (Mctco) 72') 30 840 - " -

Rurundi Mushitsi (C;iheta) 499 2 1 157 Stocking & ~1wcI1

i 1976)

(:ameroon Rnf'ia 818 3 4 2 Bresch ( 1993) - - Bamrnda 1395 6 3315 " -

Biuiganptc 509 I 3 Nill ( 1993) - - Bato~r i 750 I I 1472 Rrcsch ( l993) - - Dihum hit 1627 I 2220 Nill ( 1993) - - Dnuala 323 1 I I 3566 Rresch ( 1093) - - Dschiuip I OX4 4 1070 Seguj ( 197 1 ):'.2

- - Giu-oua 400 X 974 Rre5ch ( 1993

- - Marouu 540 12 751 - " -

Meiganga 858 10 1477 - ' I -

- - Nachtigal l Oh3 3 1330 Nil1 ( 1903

- - Ngaounci6rC 746 14 1485 Brcsch ( 1093)

- - Nlioi~nclja 1015 X 1901 - " -

Perlh~r Mlchcl iBanso~1) 777 4 1560 Nill i 1903) - Pol i 1326 4 l3XX Bresch ( 1993)

- Y aoundP 933 13 1593 - " -

Cameroon Yoko 667 8 1543 - " -

Chad Dell 054 3, 1 1 100 Audr!, i 1973) 1

Annex 1 . I

countr> site erosivity nieasure- niearl literature nient annual

period rain

I N h l [a] [nim 1 Ibor? Coast Ab1tll~111 3 1 8 0 37 3 100 Roo\c ( I075 )

Roosc cY: Rerr~-:uld

( I07 I I"? Roo\e & .ladin

( lo(>o).,:2 Koosc ( 1975 I Wenncl- ( 1977 1 '.3 - -

K o r h o ~ o

Eldorrr

K ~ \ L I I I I L I Ki1;lle

1 -0db iII-

Malindi Momba\a

Ni~irobi (Kahctc) Nakuru

Nanyuhi

Niu-oh

Voi

Bcfandriana

Allokoto l

Calab~u-

I - . ~ L I ~ L I

I hadan I horn

Y \ L I ~ ~ ~ I

O n i t s h ~ ~

0uen . i

Port- Harco~lrt

Llnludi kc Rurare

La1 ( l076h) /\I-mon ( I OX4 ' I

Salako ( I 0 8 8 )

Armon ( 108I). '- 1 Salako ( 19x8

-

- -

- -. G L I ~ L I L I

- Giscnyi

- - tiamemhe - - Kigali (airport)

lii~hcrigcri

Senegal B ~ u ~ l h c y

Erosivity Ihr 4inglc sitcx

countr~ site erosivity measure- mean literature ment annual

period rain

INJhI la1 Imml Zanihia Chlp'itu 6x5 10 1016 Pauuelyn et al ( 1988)

Zimbabwe

Kabompo

Kahn e

Kiil.11~1 Polder

Kawma

Muinilunga

Nclol:~

Seshche

Rci~bl-idgc

Cliil,inp;~

'hi\unjhanjc

I k t t

Enheldoorn Fort Victoria

Ciohnc 1nya11g:i Karol I>upane

Siili\hury T,jolo!jo Tul1

- ' I In: S:iliiko ( 1988)

" 2 i n : Ruosc ( 1975)

3 in: Moore ( 1979)

Annex 1.2

Annex 1.2 Erosivity regressions

See if your own site is listed in Tlible 12-1 Anrzex or if it is close to a site listed there. There is no evidence as to how far these regressions can be used apart from the specific sites for which they were calculated. However, the quality of rain will generally not change in the same geographic area within some kilometers. A qualitatively similar rainfall is presumed which can differ in volume.

If regressions in Tciblt. 12-1 Annex are given for single storms, the mean annual is calculated by summation of the storm erosivity for several years. It is not possible to use these regressions with annual rain volumes.

Tclhle 12-1 Arzizex: Regression,y.fi)r the calculntio~z of erosiliihl (P~,,l,l = nzeu17 aiznu~11 rain volut?ze [mm], ElLjo = lnearz cirznu~il erosivih; El-,o, = erosillihx c?f'ci S ~ O I - I ? ~ i; PI = ruin llolume ( f a .srorm i)

No. country1 area location regression remarks literature

El3,, [ N h l I Hurkina Faso Bobo- El = 0.0 ISX'?Pl - I .24 for i ~ n e l r i tormi: Galahert Kc

D ~ o u l ; ~ \ \ o regrei\~ori\ 101 N o . I and M I I I o ~ o

2 were \erq i ~ m ~ l a r to (19731 111.

eclllatlon No 3 Dclwaullr DcIu; ILII I~

thesefol-e proposed to u\e ( 1973)

No, 3 for the u h c l reglor]

hetueen 440 and 1 160

111111

2 " - Dori icc N o . I - -

3 - " - Ga~npela and Ell,,, = 0.01 58 'PI ,,,I- 1.2 foi- \ ~ n g l e \tor~il\: hoth I ) e l ~ a l ~ l l c .

Gon\C neat I-? > O.UX, n = 7 year\ of \ i te \ Ilad very \i11111~11 I J I I I ( 1077)

0 u g a d o ~ 1 ~ 0 ~ 1 \ ~ r ~ g l e itol-liii clistrihur~o~i and ucl-c c\ ;~lu;rteci together \% 1111

Allohotol N ~ g c r

4 ('amcroon a11 LI,,, = ( I 1+0 012 P;lnn)' I X \t;ltloni \ + ~ t h 3 to I4 131r\cIi

rz = 0 Ol year\ c C ~ c h I 10c)3 I

Erosivity regressions

inland > 1250 Elio = 0.269 'Pann+ 1 13

m altitude

Uganda pla- El,(, = 0 833.'.Pann-396

teau

No. country1 area location regression remarks literature

EI,,, [ N h l 5 Kenya coa\tal 7one El = I I49"'Pann-840 calculated for Lamu. - " -

tn 50 hm Mal~ndi. Mombnsa. Dar ec

inland Salaam and Z a n ~ i bar

6 - ' , - inland < 1250 El;,, = 0 57 1 'Pann-80 calculated for Lodwar. - " -

m altitude Makindi. Voi. I h d o m a ,

Kigo~na . Mwan/a and

Tabora I

7 - , ' - calculared for Eldoret.

Equator. Kitale. Nairob~

Airport. Dagoretti. Kabete

and Wilson. Nakuru.

Lyamungu and Mbeya

" calculated li)r K~surnu.

Entehbe. Foprt Portal.

Ciulu. Ji~?ja. Kahalc.

Kampala. Ka\e\e.

Masind~, Mbarara and

Tororo

0 - , ' - Katumanit Elio, = 0.9;>P,-97.4 regres\ion for \ ~ n g l e Ill\aker &

Mach;rho\ r2 = 0.9; n = 35 \terms hased o n 35 event\ O n \ ~ a d ( 1984)

I0 - " - Kat~~rri~irii / El,,,, = 0.020hPI~o,-3.9 - - - -

Machakos r2 = 0.99. n = 35

I I Niger Allohoto El7(,, = 0.01 58"PI ,,,,- I .:! \ee No. 3 Dclwaulle (1973)

I:! Nigeria Alore El,,,, = (0.12+U.I8P,)' ('or single \torn]\ Nill ( 1993) r' = 0.87: n = 240

13 - " - Ibadan El io, = ( lU48i'P,- 1059)*0.017 fol- single s t o r m : La1 ( 1976b)

r? = 0.67 regression for 5torm5 of 3

years

14 Rwanda a11 El3,, ca. (0.433':'Pann) biisrd on 6 stations Kyumugahc

and I 0 years & Berding (1992)

15 Sahel Ello = 0.87 * Pann Roose ( 1977)

I 6 Tanzania \cc No . 5 to 8 u5e regre.;sion\ No. 5 to 8 \ee No. 5 to 8 Moore ( 1979)

Uganda pla- u\e regre~s ion No. 8 see No. 8 17 Uganda - -

Annex 1.1

No. country1 area location regrcssio~i remarks literature

El.,,, [N/h1 I8 Zambia ;dl t l , , , , = 0 .0230 :PI " 1 tol- \ ~ n g l e \torlns Paun el! n ct

r2 = 0.7 1 . n = 2348 al t IWX)

Kahwr tl ,I,, = 0.0235 P, ' x7 -

r.2 = 0 (17. n = 2h 1

27 Zimbabwe all El?, , = 0.2 l 5 Panrl-5 1 . 1 r 2 = 0 h l . n = O h

78 - " - Ea\rern cl~\trlct\ I I ,,, = 0 I83 ' Parlrl-54.2

r? = iI,X7. 'I = I0

I K e ~ r c \ \ l o ~ l \ 111 hlool-e (1070: T;lhlc 3 ) 101 K I I ~ , 1 25 u c r r r ~ l t c ~ - c d Irlto rryl-c\\iorl. El;,, [ t t Lon\ 1111~1crc a1 =

0.02Y KE>?_5-26. 1 2 = 0.C) which W;I\ ca lc~~l ;~ tcc l 11-o111 1 1 s t ; ~ t ~ o ~ i \ 111 Kcn) :~ (Wrnnc.1-. 10771 ,11111 rnul r~pl~r t l I>!

1 735 In ol-cicr- to tran\l;i~-ln to SI unlrs

Annex 1.3 National iso-erodent maps

Burundi (Simonart ct al., 1993)

Annex 1.3

Cameroon (Bresch, 1993)

GABON CONGO

National i\o-el-odent maps

Marocco (Arnoldus, 1977)

ALGERIA

SPANISH SAHARA

Annex 1.3

South Africa (Smithen 8L Schulze, 1982)

ZIMBAWE

Zambia (Lenvain et al., 1988)

Annex 1.3

Zimbabwe (Stocking & Elwell, 1976)

Annex 1.4 Regional iso-erodent maps

West Africa (Rouse, 1977)

GABON

Annex 1.4

Africa north of the Sahara (Arnoldus. 1978)

National ~rainl'all ( l i \ r 1 . i h u t i ~ ) 1 1 111;qx

Annex 1.5 National rainfall distribution maps

Angola (Grif'f'i th. 1972)

1250

Atlantic Ocean

No information

NAMIBIA

Equatorial Guinea (Griffiths, 1972)

I

GABON

Atlantic Ocean

Ethiopia (Griffiths, 1972)

SOMALIA

KENYA

Ghana (von Gnielinski, 1986)

BURKINA FASO

IVORY

National I-ainfall distribution map4

Ivory Coast (Wiece, 1988)

BURKINA FASO

LIBERIA 2000

Atlantic Ocean

Annex 1.5

Liberia (Griffiths, 1972)

National rainl'all tiislr-ihution maps

Madagascar (Sick, 1979)

Annex 1.5

National I-ainfall cii\tl-ihution map\

Mozambique (Nelson, 1984)

TANZANIA

0 b

ZIMBABWE

Indian Ocean

SOUTH AFRICA

Nigeria (British West African Meteorological Service, 1954)

CAMEROON

Atlantic Oman

Annex 1.5

Sao Tom6 and Principe (Servico Meteorol6gico)

Atlantic Ocean

Atlantic Ocean

V

National rainfiill dix~ribution map\

Sierra Leone (Griffiths, 1972)

UGANDA

National rainfall di\tribution map\

Tunisia (Schlicphake. 1984)

Remada . -

Rail1 volume iund distribution for single sites

P I P - - a r t - r r r - - - a - t r - r t r , - r , z l r , c r ~ ~ X CI - r- Cr -

Cl r , - - P I - - = - C C , ~ ~ ~ J C I - S X X -

P I - PI C I CJ r1 r. -t r.

Anncx 1.6

* 2

5 2 ~ , , , 1 , , , , , 1 , , , , , 1 , ~ ~ , 1 ~ , ~ ~ c : : : : : : : : : : : : : : : : : : : : : : : : :

1 1 1 / , 1 1 / 1 , 1 , 1 , I / / I I I I I I I I

Rain volume and distribution for single sitea

R a i n volurnc and distribution lot- sitlglc h i l t . \

rain

vol

ume

and

dist

ribu

tion

[m

m]

coun

try

site

an

nual

ja

n fe

b m

ar

apr

may

ju

n ju

l au

g se

p

oct

nov

dec

reco

rd

liter

atur

e in

yea

rs

Gab

on

Bita

rn

1746

49

72

18

2 19

1 23

1 10

1 26

37

27

7 29

7 20

8 75

12

Su

chel

(19

72)

- -

Lib

revi

lle

2592

20

6 29

1 26

4 39

5 24

4 40

0

11

106

359

416

260

18

Gri

ffith

s (1

972)

-

- M

akok

ou

1754

83

12

4 25

5 24

5 18

2 51

5

18

132

331

210

118

? -

"-

- -

Med

oune

u 19

67

15.5

16

3 23

5 19

9 16

0 59

4

14

168

337

301

172

? -

"-

- -

Mek

ambo

16

61

77

130

182

174

178

88

28

70

153

268

209

104

? -

"-

- -

Min

voul

15

29

42

79

139

155

194

122

47

55

200

263

175

58

13

Such

el (1

972)

-

- M

itzic

18

42

118

110

226

207

222

46

10

14

150

346

247

146

30

Gri

ffith

s (1

972)

G

hana

A

ccra

78

7 16

37

73

82

14

5 19

3 49

16

40

80

38

18

?

"4

- -

Ada

86

4 7

18

72

94

174

208

57

12

61

99

47

15

? vo

n G

niel

insk

i ( 19

86)

- -

Aku

se

1118

23

46

10

5 12

8 16

3 18

0 65

39

96

13

1 10

2 40

?

-"

-

- -

Axi

rn

2129

51

62

12

9 14

2 42

0 53

5 15

6 54

87

20

5 19

2 96

?

-"

-

- -

Enc

hi

1651

41

65

13

8 16

0 21

0 27

0 14

3 75

16

1 21

0 12

5 53

?

-"

-

- -

Ho

1421

36

78

13

9 14

5 17

7 18

3 11

1 83

14

9 19

1 79

50

?

-"

-

- -

Ket

e K

rach

i 14

04

19

37

82

127

167

192

158

124

223

185

68

22

? - " -

- -

Kin

tam

po

1567

8

40

102

160

186

239

144

117

277

213

68

13

35

Gri

ffit

hs(1

972)

-

- K

umas

i 14

81

17

59

137

145

182

234

126

74

176

202

98

31

? vo

n G

niel

insk

i ( 19

86)

- -

Nav

rong

o 10

73

2 4

19

50

116

126

190

263

231

63

7 2

? -

"-

- -

Tak

orad

i 11

86

33

37

80

95

245

280

89

36

50

120

83

38

? -

"-

- -

Tam

ale

1053

2

9 51

88

12

1 13

2 12

8 18

9 21

7 98

13

5

? -

"-

- -

Wa

1111

5

10

42

75

132

145

144

215

234

85

19

5 ?

- "

-

- -

Wen

chi

1354

8

42

97

150

178

204

93

69

200

219

75

19

1

-"

-

- -

Yen

di

1186

2

7 32

95

13

3 14

6 15

3 18

8 25

9 13

2 21

8

? -

"-

rain

vol

ume

and

dist

ribu

tion

[mm

]

coun

try

site

an

nu

al

jan

feb

mar

ap

r m

ay

jun

jul

aug

sep

oct

nov

dec

reco

rd

lite

ratu

re

in y

ears

G

uine

a B

eyla

40

99

50

147

293

360

631

555

392

490

495

290

274

122

29

Gri

ffit

hs (

1972

) C

onak

ry

Kou

rous

sa

Mam

ou

Abi

djan

B

ouak

e Fe

rkes

sedo

ugou

S

assa

ndra

T

abou

E

quat

or

Gar

issa

K

isum

u M

agad

i M

omba

sa

Moy

ale

Nai

robi

Pe

rker

ra

- -

Port

Vic

tori

a -

- V

oi

Les

otho

M

okho

tlon

g L

iber

ia

Mon

rovi

a L

ybia

B

enga

si

- -

Der

na

- -

Tri

poli

- -

- -

- -

*4

Gri

ffit

hs (

197

2)

- -

- -

*4

Gri

ffit

hs (

1972

) -

-

Sut

herl

and

&

Bry

an

- -

Gri

ftit

hs (

197

2)

- -

"4

- -

- -

- -

Annex 1.6

Annex 1.6

Rain volume and distribution I'or single sites

= ~ c c a 3 m w , m n = t - - m n r 1 ~ 1 1 ~ , - m ~ , c c a * F i r , - m at- m T, X 2

Annex 1.6

Rain volume and distribution for 5ingle sites

Annex 1.7 Estimation of the erosivity of the 10 year storm (EI,,/lO)

Data for the EI,,,/IO will be deficient in many cases. For sorne areas close correlations exist between single storm volume (Pi [mrn]) and single storm el-osivity (EI,,,, [Nlh]). Frorn these correlations an estimate of the El,,,/ I O can be made if the volume of the I O year storm is known (c.f. Figure 75-2). However, the regressions in Figure 17-1 Annex show that these are rather large differences between climatic zones which recommend a cautious use of such regressions. The regressions in Figure 17-IAnnex are mean relationships of several regressions from Cameroon (Bresch, 1993) and Zambia (Pauwclyn el al., 1988):

coast (Cameroon) Douala: EI,oi = ( 1.45 + 0.095'I'P, )2, r2 = 0.75, n = 830

inland (Cameroon) Y oko : El ,,,, = (0.14 + 0.1 39'!'P, )', r' = 0.80. n = 352 Batouri: EI ,,,, = (0.37 + 0. 133'I'P,)', 1-2 = 0.74, n = 324 Yaoundk: El3(), = (0.07 + 0.1 53'!'P, )'. r' = 0.8 1 . n = 553

highland (Cameroon) Ramenda: EI,O, = (-0.08 + 0. I 52'::P, )2, 1-2 = 0.82. n = 423 Nkoundja: El,(), = (0.02 + 0.148'!'PI)', r' = 0.73, n = 469

north (Cameroon) Maroua: EI ,,,, = (0.08 + 0. I56"'PI)2. r' = 0.82, n = 252 Garoua: El ,,,, = (0.13 + 0.1 5Uq'P, )', r' = 0.84, 11 = 132 Poli: EI ,,,, = (0.26 + 0. I 49'I'P, )2 , r' = 0.83, n = 175 Ngaoundh-6: Elqo, = (0.20 + 0.15 1 "'P, )'. r2 = 0.78, n = 573

I ~ I - Zambia a regr-ession was given by Pauwelyn et al. ( 1988):

The curve f.ot- the coa\l i r ~ Cameroon \hould give I-eawnable e\t~mate\ for \ite\ along the We\t African Coast wrth pronounced influence of the monwon. The inland curve should be applicable for tho Central

E \ t l ~ ~ l a t ~ o n 01 thc I0 ycnl \ t o m clo\ lL l t>

Al'rican Lone between I 000 and 1 500 1nm of rainfall. The regre5sion f'oi- Northern Cameroon can probably extcndcd to I'urther areas of West Africa in the 7une between 600 to I000 mm.

single rain volume [mm]

-t- coast + inland --m- highland

-t- north ++ Zambia

Annex 2 Slope length and gradient

Lkvice for rnea\uring >lope-len~tli and ~riiclic'nt

Annex 2.1 Device for measuring slope length and gradient

10 m transparent hose with - 10 mm internal diametre

Fig1rt.c 2 1- 1 Atl l rr .~: Wcrfut. I o~~c~ l , / i ) t . tllccr.sr~t-c~t~rc~tlt ( ? f ' ~ I o ~ ~ c ~ - l ~ ~ t l g t / ~ ~ I I I ~ ~ I . C I C / I C I I I

- - - - 7

-

The two scaled bars arc placed o n level ground and the hosc is filled with hatel. u p to the zero riiark on the bars. The stoppers need to be taken ol'f' tlie hose ends bc1'ol.e starting the rne;tsut-cment. Make sure thai no air bubbles are in the hose! If the hose c1iamctc1- becotncs too small. i t is dif'f'icult to eLracuate air bubbles from thc hose. A slilall cluai~tity of hoi~sehold detergent may help in this case. For the tneasuremcnt. one person keeps one o f the bars upright while a second person moves down-slope until tlie string bctween the bars is completely streched out. The distance between the two bars should now be 5 111. The vertical distance between the two bars can be read 011 the scale. If. for cxatnplc. the vertical c-list;uice is 40crn. the water in the liose o f the lower bar S I ~ O L I I C I be beside the 20cm 11iai.k w1ic1-eas o n the higher bar i t is 2Ocnn below the zern mark. The gradient ( s ) can be calc~rlated by:

For practical purpose\ i t is easier to double the scale on the bar\ (e.g. 0 .2 In = 0.4 m ) in order to receive the vertical distance right away.

scaled bars 2 m high

5 m string with knots every metre - -- --

- - - - - - -

stopper to close hose for transport

Annex 2.2 Conversion of slope gradient in degrees to percent

gradient

degrees [%I 1 2 2 3 3 5 4 7 5 9 6 1 1 7 12 8 14 9 16

10 18 1 1 19 12 2 1 13 2 3 14 2 5 15 2 7 16 2 9 17 3 1 18 32 19 3 4 20 36 2 1 38 22 40 23 42 24 45 2 5 47 26 49 2 7 5 1 28 53 29 55 30 58 3 1 60 32 62 33 65 34 67 35 70

degrees I% 1 36 73 37 7 5 38 78 39 8 1 40 84 41 87 42 90 4 3 93 44 97 45 100 46 1 04 47 107 48 1 1 1 49 115 50 119 5 1 123 52 128 53 133 54 138 55 143 56 148 57 154 58 1 60 59 166 60 173 6 1 180 62 188 6 3 196 64 205 65 214 66 225 67 236 6 8 248 69 26 1 70 275

Nurnber of day in the yeas and con-espot1d1n.g date

Annex 3 Cover and management factor

Atlrle.k 3 . 1 N~~r?zhc)t- t ! / ' t l t l~ , irz r l l ~ orit it. trrztl c.or.ro.\porztlirzg tklto.

0 I 1 O J 2 0 3 1 1 3 0 4 - 1 1 4 05-Jan 5 0 0 1 1 h 07-Jim 7 O X - I X O O J I 0 0 - I I 0 I I I I I I [?-Jan I I 1 3 . l 1 1 13 14-Jan I 4 15-Jan 15 10-Jan 0 7 - I 7 IX-Jan IX I I I 9 70-Ian 20 11-Jan 21 22-.I;ln 22 23-Jan 23 24-.la11 24 75-Jan 25 26-Jan 2(, 27-J:ln 27 28-.1;111 28 - 1 29 30-.la11 30 S I 31 01-Feh 32 02-I:eh 33 ()3-Fc13 4 04-Feh 35 05-Fcb 36 Oh-Fch 37 07-I'ch 18 ()X-r:yt~ 7') 09-l:ch 40 10-Feh 4 I I 1 -FL>h 42 12-I.ch 4 1 I 7-kt,h 44 t ~ h 45 15-Fch 46 I 6 -F~ th 47 17-Fch 48 I X-l,el> Jc) I 9 ~~h 50 ?()-I.ch 5 l 2 I.Ft.ll 57 72.i.+h 53 23-1.~13 54 24-l.et, 55 75-keh ,jh 70-f.eh 57 27.reh 58 2X-Fch 59

01-Mas (10 O M h l 03-Milr 02 04-Mas 03 05-Mas A4 Oh-M;I~ 6.5 07-Mar 1 0 OX-Mat 67 O M 6X 10-Mar 60 I I -Mi ls 70 1 ?-Mar 71 13-Mar 72 14-Mal- 7.1 15-Mar 74 - M I 75 l 7 1 ; 1 1 7h IS-Mar 77 IC)-Mar 78 O M 70 21-Mas SO 22-Mar XI 23-Mar. 82 24-Mar X3 15-Mar 84 26-Mal- X5 27-Mar Xh 28-Milt- 87 20-Mill- 88 0 - M a r SO 31-M;IS 0 Ol -Apr 1 I 01- iZp~ 92 3 3 04-Apr 4 05-Apl- 05 00-Apr 0 0 7 - A ~ I - 97 ()X-Apt- 08 0 qC) 10-Apl I 0 0 I I -Aps 101 12-Apr 102 I I I 0 14-Apr 0 15-Apr. 105 I (I-Apr 106 17-Apr 107 IX-Apr. 108 LC) ,jpl- 0 ) 20-Apr I I 0 2 I -Apl - 1 1 1 22-,A,pl 1 12 73-Apr 113 24 -Ap1 114 25-Apl- 1 1.5 A I I 2 7 . ~ ~ ~ - 1 17 2X-i\pr 1 1s o . ~ ~ - 1 1 9 i ( ) . ~ p ~ . 2 )

day in the year Ol -Jul 1 X2 02-Jul I X3 03-Jul I 84 O L I 185 05-Jul 1 Xh O J I 1x7 07-Jul I 88 OX-Jul ISC) 09-Jul I 9 0 10-Jul 101 I I-Jul I 92 12-Jtrl I 93 13-Jul I 04 4 J 1 l I95 15-Jul I 00 I h-.l~rl I 97 17-Jul IOX IS-Jul I 09 J I 200 20-Jul 201 I - I 202 22-Jul 203 2 3 - I 204 24-Jul 205 25-Jul 206 I - I 207 27-Jul 208 28-JuI 20C) 20-5111 210 30-Jul 211 31-Jul 712 Ol -Aug 2 13 0 2 - i Z u ~ 2 14 0 - 1 2 5 04-Aug 2 10 05-Auf 217 00-Auf 2 1 X 07-Aug 2 19 OX-Aug 1 2 0 09-Aug 221 O A I 222 I I - A u g 27.1 12-Aug 224 13-Auf 225 I4-Aug 226 A 227 It,-Aug 228 17 -AL I~ 220 18-Aug 2.30 10-Auc 23 1 20-ALI~ 232 21 -Aug 233 22-Aug 234 23-Aug 235 2-1-Aug 236 25-Ally 737 26-.4ug 238 2 7 - A L I ~ 230 28-Aug 240 2 0 - A L I ~ 241 A I 242 3 1 -AL I~ 243

date/ no. of Ol -Ma! 12 1 02-M;I) 122 03-Mi11 123 0 - 1 ) I 05-Ma) 125 Oh-Ma) I O M 127 OX-Ma) 128 09-M;i) 129 10-Ma) I 3 0 1 I -Ma) I 1 2-Ma! I32 13-Ma) 133 14-M;I) I 15-Ma) 135 I h-M;l) 136 17-Ma) 137 I X-Ma) 1.38 IY-Ma) I 20-M<I) 1-10 21-Ma) I ??-Ma) 142 23-Ma) 4 3 ?&Ma) 144 25-Ma) 1-15 26-May 4 27-Ma) 137 28-Mil) I4X 71)-M;1) 140 I ) I50 M I I51 O J L I I 152 02 -J~ I I 153 - I 5 4 0 4 - I 155 05-Jun 156 O J 157 07-Iu11 15s OX-Jull 150 OC)-.lull I 6 0 10-Jun 101 I I - J ~ r n 162 12-JLIII 103 13-Sun It14 I - L I I 165 15-Jun 0 I h-Jurl 1117 17-Jun IOX 18-JUII I 00 - J l 1 170 20-.lulr 17 1 7 1 -Sun 172 22-Jun 17'1 23-Jun 174 24-Jun 175 25-Jun 176 J I I 177 27-Jurl 178 2X-Jull 170 - I 180 ( I 1 I

Ol -Sep 244 02-Sep 2-15 03-Sep 2 4 04-Scp 2-17 05-Sep 248 Oh-Scp 219 07-Sep 250 OX-Scp 25 1 OC)-Scp 252 10-Sop 5 I I-Scp 5 17-Scp 255 13-Sep 250 14-Scp 257 15-Sep 258 16-Sc11 259 17-Sep 260 IS-Sep 201 1')-Scp 262 20-Sep 0 3 21-Scp I

22-Sep 205 23-Sep 260 24-Scp 267 25-Sep 268 20-Sep 26') 27-Sep 270 28-Scp 271 2C)-Scp 7 30-Scp 273 01-0ct 274 02-0c.t 275 03-Oct 276 0 O C t 277 05-0ct 278 06-Ocr 279 07-0c1 280 OX-Oct 2X I 09-Ocl 7x7 10-0c.t 283 I I - 78-1 12-0ct 2115 13-0c.1 280 1-1-0c1 287 15-Oct 2x8 Ih-Oc( 789 17-(Jet lC)0 IX-(-)cl 2'11 1')-Oct 202 20-Ocl 2<)3 2 1 -(-)el 2 4 22-OCt 295 23-Oct ?')(I 24-Oct 207 75-(Jet 298 20-0ct 2')q 27-Ocl 300 28-Ocl .{()I 2 9 - 0 ~ 1 302 70-Ocr 303 3 I-0c.t 304

OILNO\ 305 0 2 0 306 3 - 3o7 O N 308 05-No\ 300 Oh-NO\ 3 10 07-No\ 3 l I OX-NO\ 3 17 00-KO\ 3 1 3 0 - 3 I 4 1 I 3 I 5 12-Yo\ 3 It ] 13-No\ 3 17 14-No\ 318 I ~ - N O \ 3 I 0 I t ) -No\ 320 17-No\ 32 1 IX-No\ 322 10-No\ 373 N O 374 N O 325 22-NO\ 726 23-No\ 327 24-No\ 328 25-N0\ 324 ? ~ - K o \ 1 0 27-K0\ 331 28-KO\ 332 N O 333 30-No\ 334 OlLDec 335 07-1)t.i. i i ( 7

O3-l)ei. .i37 04-LILY 338 05-1)c.c 1 lC) Oh-l)cc 340 07-1)ec 34 1 OX-llci. 342 0') I)cL. 34.3

1 0 - 1 1 ~ ~ 444 I lLl)c.c 3-15 I~-L)cL' 340 13-l)cc 447 14-l)cc 348 15-1)cc 340 10-DCC, 350 17-1)cc 35 1 18-l)cc. i S 1 IO-DCL' 351 70-l)ec 75-1 2 1 - 355 27-I)CC .;SO ~ ~ - D c c 357 24-l)cc 35X 75Dec 35') ~ ( ~ - I ) L . c 300 27-L)t.i 7f1 l ?X-L)ec. 302 ?')-[kc 363 30-Drc 1(34 '1 I - I )CC 305

Ficlds rnethocls for- the ~ncnsur-rmt.n~\ ...

Annex 3.2 Field methods for the measurement of mulch cover and canopy cover

I . Mulch cover measurement by the meterstick method

Put a meter\tick on thc ground and count on one side all the cln- marks which are in contact with mulch tilaterial (Figul-e 32-]Annex). The mulo11 cover (MC) i \ given by:

number of cm mark\ in contact MC (%) =- letiglI3 of nletcl-\tic]\ (cm) 1 00

Ex:irnple: 38 cm-rnarks are i n contact with one side of a 2 111 long rneterstich M C is 38/200 = 19 (2.

This method 1s well sulted for \~iiall plots. Thc nur-nbcr ot measul-ernents depend\ on the unifonnnc\s of the cover. In own mea\urenicnt\ 12 rcpllcation\ with a 2 m long metel-\tick were taken o n 500 m' plot\ equivalent to a random 23 rn transsect. It 15 iniportant that the {tick i \ randomly placed in the plot. Random placement can be a\\ured by throwlng the meterstick into the plot.

Annex 3.2

2. Mulch cover measurement by the cord and knot method

This method is similar to the meterstick method. Marks or knot5 are attached to a 10 to 20n1 long cord which is streched out on a field (Figure 32- 2Annex). The number of knots in contact with mulch material ('I: 100) divided by the total number of knots on the cord gives the percent nlulch cover.

Field method$ for [he me;ljLIrcmcnt

3. Mulch and canopy cover measurement by visual estimation

Form a quadrat of i x I m using 4 wooden sticks o f I m length or two f'oldable 2 rn-sticks to mark out a 1 m' area in the field (Figure 32-3Annex). If the observed area is well delimitated, cover is easier to estimate than on an undefined area. The size of the area can be srnaller than 1 m' but should not be larger because visual estimation becomes more difficult with increasing size of the area. Cover is estimated visually i n the delimitated area. Calibration of the eye can be facilitated by the examples ill Figure 32- 4Annex. Estimations become also easier if wires are streched at regular distances on the wooden frame. Estimations are reasonably precise after some routine.

Fig~lsc] 32-3 AIIII (~ .Y: Mtrrkirlg o ~ l t r l r l trr-err wyitl~ cliflorclzt cle\~ic.e.c ( t r . cc'ooclcil fj-il111e w9it11 c~,it-o-ilet, I?. t ~ ~ > i g . $ , c. tn70 2 r~~-.\ti( ,h,$)

4. Measurement of canopy and mulch cover by a sighting frame

This method was proposed by Elwell 81. Wendelaar (1977). Ten hollow pipe\ ale attached to a frame (Figure 32-SAnnex). By peering through onc of the pipes a ~ ~ n a l l area can be observed. Mulch or ciinopy cover in this area i \ either rated as 'yes' or 'no' ol- rated o n a scale bctwccn 1 and 10. In thc first case. the number of points observed with cover (= yes) divided by the n~lmbel- of all points ob\crvcd gives the coverage.

Annex 3.2

Figure 32-4 Annex: Selected coverages for the calibration of the eye (the stick in the pictures is 1 m long)

Field rllelhod\ 1.01- rhe n ~ e a ~ u r - c r n c . ~ ~ ~

1- 1 OOcm -I - 3.92cm

F i ~ l / s ( ~ -32-6 A I I I ~ O ~ : 0/1\01-\,(itio11 O ~ I . O I - ( / \ ~ I ~ ~ / L / ( ~ I I ( ~ o ~ / 1719 /)/(111t /loigll/

60-

50- - 8 U

C

L

canopy height [cm]

FieIci methods f o r (he nlca\u~-zmenr ...

In the second ca\e the \urn of rattngs for all ten pipes gives the coverage for the 1 rn transsect. 100 obcervations in a regular cropstand were pt-ecle within 2 C/r coverage (Cackett. 1964) whereas 300 observations here necescary for a 5 % accuracy in an irregular croprand (Elwell & Gat-dt~es. 1975). Quansah et al. (1990) used an average of 5.4 observationslm'. In own nleasurcment\ 180 polnts on a 500m2 plot (0.3 observat1on\/m2) proced adequate (Nill. 1993). This version of sighting frame can be u\ed for mulch and durlng the f'irst 2-4 weeks of plant growth while the plants are st111 \mall becau\e the observation errol- increases rapidly with increa\ing plant he~ght (Figure 32-6Atitiex).

A modified version of the sighting frame avoids this obsel-vation error- by sliding a graduated stick through the pipes (Figure 32-7Annex). In this case the nurnber of contacts of the stick with leaves or mulch at-c counted. This version allows at thc same time to measure thc mean canopy height above gt-oi~r~cl.

An alternative in tall crops is the use of a min-or on tlie \~ghtitls frame which allows an observation of' thc canopy cover outlined again\t the \hy (P~gilre 32-8Annex). The sighting I1-:~tiie\ are ea\y and cheap to construct.

Cursor

gunsight

End Elevation on Cursor Showing the Gurml@t

CONSTRYCTION DETAILS FOR SIGHTING FRAME ,,,_.,, -- -

Fielcl merhods I'or the mea~ur ' t 'ment ...

Stop for Cursor bar Support arms

Isometric View on Mirror

'C bracket and

horizontal

Sectional Elevation A - A

CONSTRUCTION DETAILS FOR MIRROR

Annex 3.3 Growth curves for mono- and mixed crops

Growth curves for the following mono- and mix-crops are given for:

Balmbara nut canavalia cassava cotton cowpea groundnut m a z e maizelcassava mixcrop pigeon pea rice sorghum soya suntlower tea tobacco

Figure 33- 1 An1ze.r Figure 33- 1 Aiznex Figure 33-2 Annex Figure 33-3 Annex Figure 33- 1 A1111e.r Figure 33-4 An1ze.u Figure 33-5 A I I I I ~ . ~ Figure 33-2 Anne. .u Figure 33-2 An1ze.r Figure 33-5 Anr~ex Figure 33-5 A1ule.r Figure 33-4 Ar1ne.x Figure 33-3 Arlrle-x Figure 33-6 Arzrze-x Figure 33-3 A1zize.r

Growth curies for mono- ancl rn ixed cl-op4

1 00

90

80

- 70 ae Y ,- 60 > 00 so > a g 40 m

30

20

10

0 1 0 20 40 60 80 100 120 140 160

days after emergence

L

100

90

80 ~

& 60- >

a o 40--

- -- 'i

- -. 1 l l 1 l l l l l 1 l l r l l l l 1 l 1 1 1 1 >

0 20 40 60 80 100 120 140 160 days from emergence

100

go--- -- - -- - -- - -------- -

- - - - -- - - - - ---- - -- -

--- - --

0 l l l l l l l l l l r l l l l l l l l l l l l l i l l l l l l l

0 20 40 60 80 100 120 140 160 days from emergence

Growth curve$ lo r mono- ancl rn~xed crop\

100 -

_ __L_

o l l l - l l l l r l l l l l l l l l l l l l l l l l l l l l l l l '

0 20 40 60 80 100 120 140 160 days from emergence

8 0 ~

70---

60 - 8

SO-.

-.- -----..-"- -.

-. --,

-

1 r 8 1 1 1 1 1 1

800 1000 1200 1400 1600 days after emergence

Annex 3.4

Annex 3.4 Detailed C factors

T~thlt. 34-1 Anrze.~: Detciilc-'d C ji~c.torLs,for,fc)rrrst, hush and gr0.s.s vegetcrtiotl

no.

6 7 8 9 10

1 1

12

13

13

land

-use

de

scri

ptio

n C

fac

tor

mea

sure

- co

untr

y lo

cati

on

liter

atur

e m

ent

peri

od

mea

n ex

trem

es

[a]

crop

fa

llow

wit

h m

aize

res

idue

s 0.

09

- M

alay

sia

Sul

aim

an e

t al.

( 19

83)

resi

dues

"3

-"

-

fall

ow w

ith

mai

ze

0.0

1 -

- -

-

resi

dues

and

mul

ch

-"

-

fall

ow w

ith

mun

g be

an

0.25

-

- -

-

resi

dues

-

"-

fa

llow

wit

h gr

ound

nut

0.28

-

- -

-

resi

dues

-

'*-

fa

llow

wit

h co

wpe

a 0.

03

- -

- -

resi

dues

resi

dual

fal

low

eff

ects

*'

bush

fal

low

w

ell

deve

lope

d: 1

st ye

ar

0.8

0.78

to 0

.81

1 C

amer

oon

Yao

und

Nill

(19

93. p

. 16

3)

afte

r cl

eari

ng

--

-

2nd

year

s af

ter

clea

ring

"2

0.9"

2 -

- -

- -

-

gras

s fa

llow

Im

pera

ta g

rass

; 1s

t yea

r 0.

3 0.

33 to

0.4

9 1

- -

Nac

htig

al

- " -

afte

r cl

eari

ng

-"-

2nd

year

aft

er c

lear

ing*

2 0.

7 1

- -

- -

- -

land

-use

de

scri

ptio

n C

fact

or

mea

sure

- co

untr

y lo

cati

on

liter

atur

e m

ent

peri

od

mea

n ex

trem

es

[a]

Pan

icum

ri

ght

afte

r cu

ttin

g. 8

- 0.

058

3 -

"-

-

- -

-

max

imum

14

% c

anop

y co

ver

Dig

itar

ia

slop

e 12

%; p

lant

ed 3

0.

007

3 B

razi

l A

lago

inha

L

epru

n et

al.

( 19

86.

decu

mbe

ns

year

s ol

d pa

stur

e (C

=

p. 2

26)

mea

n of

3 y

ears

) B

rach

aria

I s

t yea

r 0.

387

- In

done

sia

- A

bdur

achm

an e

t al

. de

cum

bens

( 19

84)

- -

fully

est

abli

shed

0.

002

- -

-

- -

- -

plan

ted

past

ure

(C =

mea

n 0.

003

6 B

razi

l B

rasi

lia

Lep

run

et a

l. (

1986

.

1 I

of s

ix y

ears

): 5

.5%

slo

pe

p. 2

28)

"1

Sur

pris

ingl

y, s

avan

na f

allo

w h

ad h

ighe

r re

sidu

al e

ffec

t w

hich

mig

jt be

due

to g

ener

ally

low

er o

rgan

ic m

atte

r le

vels

and

slo

wer

tu

rnov

er

"2

resi

dual

eff

ects

for

2nd

yea

r as

sum

ed a

s 50

% o

f th

e I s

t yea

r *3

O

nly

the

crop

resi

dues

are

left

in t

he f

ield

and

the

spon

tane

ousl

y gr

owin

g ve

geta

tion

afte

r ha

rves

t

mea

n ex

trem

es

Tab

le 3

4-2

Ann

ex: D

etai

led

C f

acto

rs f

or b

anan

a

mon

ocro

p 8

7~

sl

ope.

l s

t ye

ar:

leav

es

0.56

0.

14 to

1.0

8 pl

aced

aro

und

trun

ks;

2nd

year

: le

aves

pla

ced

on

cont

our;

spa

cing

5 x

3 m

.

Alt

erna

tive

ly f

or a

you

ng

plan

tatio

n (

1 st

year

) -

- as

abo

ve b

ut s

paci

ng 2

x 3

m

0.16

0.

04 to

0.3

-

- as

abo

ve b

ut s

paci

ng 3

x 3

m

0.3

0.08

to 0

.58

- -

as a

bove

but

spa

cing

4 x

3 m

0.

42

0.1

to 0

.83

- -

with

mul

ch

0.00

029

0.00

029

to

0.00

036

no.

[a1 2

Bur

undi

M

ashi

tsi

Ris

hiru

muh

irw

a ( 19

92. p

. 90)

land

-use

de

scri

ptio

n C

fac

tor

mea

sure

- co

untr

y lo

cati

on

liter

atur

e m

ent

peri

od

2 -

"-

-

-

2 -

"-

-

-

2 -

"-

-

-

Ivor

y A

diop

odou

m

Roo

se (

197

5,

Coa

st

p. 3

013

1 )

6 7 8 9

-L

.-

as

no.

3. s

paci

ng 3

x 3

m

0.00

09

0 to

0.0

07

2 B

urun

di

- " -

-

-

plus

com

plet

e m

ulch

cov

er

-.*-

0.

04

Lew

is (

198

6 in

: Y

oung

. 19

89. p

. 33)

inte

rcro

pped

w

ith b

eans

0.

1 -

-

-.

+-

w

ith s

orgh

um

0.14

-

-

Detailed C factor\

I I

mea

n ex

trem

es

[a]

no.

land

-use

de

scri

ptio

n C

fact

or

mea

sure

- co

untr

y lo

cati

on

liter

atur

e m

ent

peri

od

16

* 1

egus

i m

elon

- C

o1oc

~j~n

thu.

s c,it

r~~

llu,

s L

. *2

10

-12

are

rath

er s

mal

l va

lues

: ho

wev

er, i

t was

not

spe

cifi

ed i

f th

e re

sidu

es w

ere

kept

wit

hin

the

fiel

d "3

th

ere

was

a s

light

tre

nd o

bser

vabl

e th

at th

e su

bfac

tor

for

inte

rcro

ppin

g in

crea

ses

with

inc

reas

ing

pine

appl

e de

nsit

y. T

he r

atio

s in

- te

rcro

p/m

onoc

rop

wer

e 0.

77, 0

.94

and

0.79

for c

owpe

a in

terc

ropp

ed w

ith 4

5.00

0. 6

0.00

0 an

d 70

.000

pin

eapp

le p

lant

s. W

ith m

e-

lon

the

rati

os w

ere

0.43

, 0.

53 a

nd 0

.54.

res

pect

ivel

y.

subf

acto

rs

- -

~nc

rea\

ing d

en\~

ty by

0.75

0.

7 1

to 0

78

1 -

- -

-

1500

0 pl

ants

: ca

lcul

ated

1 7*3

18*3

from

den

siti

es a

bove

-

- 0.

84

0.79

to 0

.93

1 -

- in

terc

ropp

ing

inte

rcro

ppin

g w

ith

- -

cow

pea

(200

00 p

lant

slha

) -

- -

- -

"-

in

terc

ropp

ing

with

egu

si

0.5

0.44

to 0

.53

1 -

-

mel

on*

1 (5

000

plan

tslh

a)

Annex 1.3

T(lhltj 14-4 Aizr1e.r: Detailed C,firctor* ,for cn.s.sm)n

Detailed C' factor\

mea

n

extr

emes

[a

] no

till

spac

ing

I x

l rn; o

nly

0.0

15

0.0

15

--

-

-

- -

- -

till

age

of p

lant

ing

hole

s: 7

to

20%

slo

pes

no,

inte

rcro

p ca

ssav

a in

terc

ropp

ed w

ith

0.08

0.

109

to

3 G

hana

K

wad

oso

Bon

su &

Obe

ng

mai

ze (

75

x 25

cm

) and

0.

023

(197

9.p.

515

) co

coya

m: p

low

ed a

nd

harr

owed

; 19

. 3,

15 k

g/ha

N

PK

; 7.

5% s

lope

-

- as

abo

ve b

ut 3

% sl

ope

0.06

6 T

I -

"-

E

jura

-

- -

- in

terc

ropp

ed w

ith

mai

ze;

0.24

0.

43 to

0.0

5 1

Kig

eria

Ib

adan

A

ina

et a

l. (

1979

)

1st s

easo

n C

= 0

.33;

2nd

se

ason

C =

0.0

5

cass

ava

des

crip

tion

C

fac

tor

mea

sure

- co

untr

y lo

cati

on

liter

atur

e m

ent

peri

od

inte

rcro

pped

wit

h m

aize

; 0.

007

1 -'

.-

Alo

re

Sabe

l-K

osch

ella

re

sidu

es i

ncor

pora

ted;

on

60

( 19

88)

I I

cm h

eaps

ori

ente

d ac

ross

6%

slo

pe; d

ista

nce

betw

een

heap

tops

0.7

x I

.4 m

; 3

repe

titi

ons

no.

30

71

77

- -

33

33

5 26

cass

ava

desc

ript

ion

C fa

ctor

m

easu

re-

coun

try

loca

tion

lit

erat

ure

men

t pe

riod

m

ean

extr

emes

[a

] - - -

~nt

ercr

oppi

ng m

aize

+ ric

e 0.

588

- In

done

sia

- A

bdur

achm

an e

t al

. +

C;lS

SaV

a (

1984

) -

--

in

terc

ropp

ed ~

vit

h \o

ya

0.18

1 -

- -

- _

.._

w

ith g

roun

dnut

0.

195

- -

- -

- -

as n

o. 9

but

int

ercr

oppe

d 0.

02

I U

SA1H

;lwai

i M

olok

ai

El-

Sw

ait)

et

a], (

198

8,

with

sty

losa

nthe

s an

d p-

9)

Des

mod

ium

tri

tlor

um

- ..

- ca

ssav

a on

lev

el g

roun

d:

0.17

0.

15 to

0. I

9 -

- - .. -

-

- -

-

spac

ing

1.7

x 0.

6 m

acr

oss

slop

e: 3

row

s of

Pha

seoi

us

or i

owpe

a be

twee

n ca

ssav

a;

spac

ing

0.35

bet

wee

n ro

us

- -.

- in

terc

ropp

ing

of m

aize

+

0.35

7 -

Indo

nesi

a -

Abd

urac

hman

et

a1.

rice

+ ca

ssac

.a w

ith r

esid

ue

( 19

84)

mul

ch

inte

rcro

p +

inte

rcro

pped

wit

h ri

ce +

0,

079

- -

- -

lnul

ch

mai

ze p

lus

6 tl

ha o

f ri

ce

\tra

w n

~ul

ch

2 3

\ ? A m

2. - m :< n

T $.

5. - - - Q

Annex 3.4

T~rblr 34-4 Arznrx, c.ontivzuc,

- 0

C 3

3 L bCI

k i' %

8 4

2 G a 4

. TJ - 4

a" 2 2 2 ri 0, - P a

QI

a u

L

u .- -

e .- CI

u - +a

u

1 8

I

E u a z g . z a I 2 E & -

b">

L.

u U 2

C 0 .- - .% L b! 4

s

6 C

- - - "8 00 d " o 2 E 6

- % 0 - 0 2 , z3. d -. e -

cf - C . f - - Z - p, - I-- 'U Y a p - c f '" c

3 - . - u z - c f - J2 5 2 2 2 s g .go 2 2 ~ 2 2 . g z g 2 % k s a c Q - J z 5 a m 0 a &i L ~ K < = Z E S ~

E 3 3

W . - a d

cf &

. - -a e d 4

s , $ 4 k i

- z

.. . - a m u LC V:

. - X 'U J;

X

2 2 o

cf

a 4 2 , % 2 2 I t:

: : s - - , , , ?

C

I-- - v l ? ? - i ; ; I I I ,

c4 7 r! 3 3

"! C1 CI w Tt

g s a 0 7 5 2 2 3 0

- y 3 O C *

# -g - 3 s 3

2 .. h, 3 ~ z ~ G .s C k

0

0 cf

L go? KO) P L ; ~ z e $ , x g E - = g ~ ~ $ w , * - - O 2 x 8; -2 .E z : * o m

g c ~ d F ...; e b .s .= 6 S 5 .$ o I- b ~ Q.s

- ~ ~ ' U C g $ g ~ c . - e, g 2 i 5 c z 2 g e z,o " * W O 3 . . , 3 - a , 0 , , J ; v : c r

a 2 0 C I

3 - I I , I . * 4 -

E - 1

. . . . , I I , I

- P4 F, -T m at--

Annex 3.4

Tab

le 3

4-6

Ann

ex:

Det

aile

d C

fac

tors

for

mai

ze

mea

n ex

trem

es

[a]

plow

*2

spac

ing

0.9

x 0.

5 m

; sl

ope

0.16

0.

02 to

0.1

2 3'

" 1

Zim

babw

e D

ombo

shaw

a V

og

el( 1

992.

. p.

13)

no.

3.5 7~;

corr

ecte

d fo

r co

ntou

ring

(0.

5)

.. $:

? -

- sp

acin

g 0.

9 x

0.62

m:

slop

e 0.

18

4.5

%..

corr

ecte

d fo

r co

ntou

ring

(0.

5)

- -

0.51

trea

tmen

t de

scri

ptio

n C

fac

tor

mea

sure

- co

untr

y lo

cati

on

liter

atur

e m

ent

peri

od

0.0

4to

0.1

7

3*1

- -

Mak

ahol

i

0.5

USA

/ M

olok

ai

El-

Sw

aify

et

Haw

aii

al. (

198

8.. p

. 4)

Indo

nesi

a -

Abd

urac

hman

et

al.

( 19

84)

Mal

aysi

a S

erda

ng

Mok

htar

uddi

n &

M

aene

( 19

79)

- -

Sula

iman

et

al. (

1983

)

12 %

slo

pe; a

long

slo

pe

0.1

1 -

3 B

razi

l A

lago

inha

L

epru

n et

al.

(198

6., p

. 22

6)

6 -

- 5.

5 %

slo

pe:

hand

hoe

.

-

till

age;

0.5

6 -

- sp

acin

g: 0

.6 m

bet

wee

n 0.

82

- 3

row

s; 2

0000

to 5

100

0 pl

ants

lha

depe

ndin

g on

rai

n vo

lum

e; 3

0 -

60 k

g/ha

N..

30

Bra

silia

-

-

Ken

ya

Kat

uman

i U

lsak

er &

Kil

ewe

( 19

84..

p. 2

33)

Drlailed C factors

no.

29

30

31

32

33

34

35

36

37

38

trea

tmen

t de

scri

ptio

n C

fact

or

mea

sure

- co

untr

y lo

cati

on

liter

atur

e m

en t

peri

od

mea

n ex

trem

es

[a]

- .b

-:5

4 sp

acin

g 0.

9 x

0.5;

with

0.

012

- 3

Zim

babw

e D

ombo

shaw

a V

ogeI

( 199

2,, p

. 1 3

) m

aize

res

idue

s; c

orre

cted

fo

r co

ntou

r pla

ntin

g -

- -*

4 sp

acin

g 0.

9 x

0.6

m; w

ith

0.01

1 0.

016

to

3 -

- M

akah

oli

- '*

-

mai

ze r

esid

ues:

cor

rect

ed

0.0

17

for c

onto

ur p

lant

ing

-"

-

7.5

Ye s

lope

; spa

cing

75

x 0.

03 1

0.00

63 to

3

Gha

na

Kw

ados

o B

onsu

& O

beng

25

cm

,. fe

rtil

izer

19.

. 3 a

nd

0.07

7 (1

979.

. p. 5

15)

15 k

glha

of

NP

K

I -

--

as

abo

ve; b

ut 3

C/c

slop

e 0.

035

-*5

- -

Ej u

ra

- -

redu

ced

spac

ing:

0.7

5 x

0.25

m:

7.5

0.02

0.

038

to

3 -

- K

wad

oso

- " -

tilla

ge

5% s

lope

; 19

., 3

and

15

0.00

57

kglh

a of

N,,

P a

nd K

: pl

owed

wit

hout

har

row

ing:

tw

o ha

ndw

eedi

ngs

0.04

1 -*

5 1

- -

Eju

ra

as a

bove

; 3 5%

slop

e - " -

- -

spac

ing

0.75

x 0

.25

m;

0.00

29

0.00

563

to

3 -

- m

ulch

K

wad

oso

-"

-

fert

iliz

er 1

9.. 3

and

15

kglh

a 0.

00 13

of

N,,

P an

d K

; plo

wed

and

ha

rrow

ed:

mul

ch q

uant

ity'

? as

mul

ch p

lot

abov

e; 3

%

0.00

4 -*

5 1

- -

Eju

ra

-"

-

- -

slop

e: m

ulch

qua

ntit

y?

-"

-

with

im

pera

ta m

ulch

0.

02

- M

alay

sia

- Su

laim

an e

t al.

(198

3)

inte

rcro

pped

1 w

ith g

roun

dnut

plu

s 3

t/ha

0.12

8 In

done

sia

- A

bdur

achm

an e

t al.

mul

ch

stra

w m

ulch

(1

984)

men

t pe

riod

no.

mea

n ex

trem

es

[a]

ridg

e sp

acin

g: 0

.75

x 0.

25 m

: 3

'3-

0.02

6 0.

057

to

3 G

hana

K

wad

oso

Bon

su &

Obe

ng (

197

9,,

slop

e; 1

9.. 3

and

15

kglh

a of

0.

0087

p.

515

) N

., P

and

K:

plow

ed a

nd

trea

tmen

t de

scri

ptio

n C

fact

or

mea

sure

- co

untr

y lo

cati

on

liter

atur

e

harr

owed

; ri

dges

acr

oss

slop

e -

- ri

dge

on c

onto

ur;

7.5

5% sl

ope

0.05

4 -*

5 I

- -

Eju

ra

- -

- -

spac

ing:

I00

x 2

0 c

m. o

n 0.

67

0.25

to 0

.95

Ivor

y A

diop

odou

m

Roo

se (

1975

,,

ridg

es a

long

slo

pe

Coa

st

p. 3

013

I)

tied

-rid

ging

- sp

acin

g 0.

9 x

0.5:

rid

ges

on

0.01

8 0.

006

to

3 Z

imba

bwe

Dom

bosh

awa

Vog

eI( 1

992.

, p.

13)

/not

ill*

5 co

ntou

r; 2

5 cm

hig

h co

ntou

r 0.

026

ridg

es w

ith 1

% l

ater

al s

lope

; tie

s at

int

erva

ls o

f 1.

5 m

and

ap

p. 1

5 cm

hig

h -

- "

*5

spac

ing

0.9

x 0.

6; r

idge

s on

0.

004

0.00

3 to

3

- -

Mak

ahol

i -

-

cont

our,

25

cm h

igh

cont

our

0.0

6

ridg

es w

ith 1

'3%

late

ral

slop

e;

ties

at i

nter

vals

of

1.5

m a

nd

app.

15

cm h

igh

* 1

Onl

y da

ta o

f th

e 2n

d to

4th

exp

erim

enta

l ye

ar w

ere

cons

ider

ed. R

esul

ts o

f 1 s

t yea

r sk

ippe

d du

e to

res

idua

l ef

fect

s of

nat

ural

fal

low

"2

P

low

ing

wit

h ox

en d

raw

n si

ngle

fur

row

mou

ld-b

oard

plo

w r

ight

aft

er h

arve

st t

o 20

- 2

5 cm

dep

th

$3

Rip

ping

bet

wee

n fo

rmer

mai

ze r

ows

wit

h an

oxe

n dr

awn

sing

le r

ippe

r tin

e to

20

- 25

cm

dep

th

:k4

As

"-3 b

ut r

esid

ues

of f

orm

er m

aize

cro

p le

ft

"5

Soil

was

plo

wed

to

20 -

25 c

m d

epth

at

begi

nnin

g of

exp

erim

ents

. 25

cm

hig

h co

ntou

r ri

dges

(I

% s

ide

slop

e) w

ere

form

ed w

ith o

xen

draw

n m

ould

boar

d ri

dger

. In

sub

sequ

ent y

ears

.. no

plo

win

g w

as c

arri

ed o

ut b

ut r

idge

s w

ere

refo

rmed

aft

er h

arve

st a

nd 6

wee

ks a

fter

pla

ntin

g. 1

5 cm

hig

h cr

oss

ties

wer

e al

so fo

rmed

tw

ice

a ye

ar a

t 1 m

inte

rval

s in

the

fur

row

s. P

lant

ing

was

on

top

of t

he r

idge

s.

*6 P

lant

ing

hole

s w

ere

open

ed w

ith h

and-

hoe

(bad

za h

olin

g ou

t in

Zim

babw

e)

P

\O

Annex 3.4

Detailed C I'aclon

a

1 cI

L

* .- -

G .- Y

2 - +. u G 3 8

I

2 , - 2 E . E 7 Z E g -

r/l

L L

u C,

C J

e U

E .- u a

.C

L

2 -e

a L

6

- F,

g - - .' -

.y 0

E . - rJ - 3

" . - x , a' -

S i 6 30

? y - "

6\ 2 ;

Eo : . - Z 8

- P I

30

p! C

9 2 - - -

rc,

6

- - 2 g - X

9 $ 0 - - - - -

- a - a * p z Q ar C Z S -.

- - F, 2 - G 2 - 0

Lg 2 = - 2 . - $ 2 h , 1 2 . z 3 -

u p , : 0 7 C ! 3 m - ~ ~ c l , gvl a 0 s L

". r . - . - C

J.

E 2 , I - , < ., 1 , 2

-S 3 . - . - J. - a, = 5, .- a s= - - a , 5 5 s

' 'C: 3

-

0 !

C I P -

rc, !

C

LC

b

E l ,

I-. 3\ 5- - x g

-22

c 2 2 2 f: d

3 . - -

I

z, : ,

r, Cc' '

I I I

.-

g % s g G G

5 s " '3

. 3 - ; C !

z 5 - I

5 :, :,

I 0 0 0 - - -

?-. - -

- 3 c

a + 5 6

, g ,I g + * g - , Q g

cr, - 2

. . blJ -

+ 5 6 a ' 3 $ 2 , J. r

'" 'c -

no.

17

18

19

20

21

22

23

24

crop

de

scri

ptio

n C

fac

tor

mea

sure

- co

untr

y lo

cati

on

liter

atur

e m

ent

peri

od

mea

n ex

trem

es

[a]

--

-

plow

ed:

mai

ze r

esid

ues:

6%

0.

006

0.00

2 to

2

-"

-

- -

Nil1

(19

93)

slop

e; s

paci

ng 7

5 x

25 c

m;

0.0

14

2 re

peti

tion

s -

a-

no

till:

mai

ze r

esid

ues;

0.

01

0.00

04 to

2

- " -

Ib

adan

La

1 ( 1

976a

, p. 3

1 )

plan

ted

alon

g sl

ope

0.02

--'-

no

till:

mai

ze r

esid

ues:

6%

0.

004

0.00

2 to

2

-"

-

Alo

re

Sabe

l-K

osch

ella

sl

ope

0.00

6 ( 1

988)

- .'

- no

till:

mai

ze r

esid

ues;

6%

0.

0003

0.

0002

to

2 - " -

- -

Nil1

(19

93)

slop

e: s

paci

ng 7

5 * 2

5 cm

0.

0004

ac

ross

slo

pe

Iris

h po

tato

es

- 0.

22

- R

wan

da

- L

ewis

( I9

86 in

: Y

oung

198

9, p

. 33

) le

mon

gra

ss

0.43

4 -

Indo

nesi

a -

Abd

urac

hman

et

al.

( 1 9

84)

papa

ya

with

out

cove

r cr

op

2.1

- -

Abd

ul R

ashi

d (I

98

1 in

: Sul

aim

an e

t al

. 19

83)

soya

0.

399

- In

done

sia

- A

bdur

achm

an e

t al

. (1

984)

notil

l (pr

obab

ly w

ithou

t 0.

103

resi

dues

left

bec

ause

not

ill-

subf

acto

r =

0.6

7): s

lope

2 3- 7

'4

P

desc

ript

ion

C fa

ctor

m

easu

re-

coun

try

loca

tion

lit

erat

ure

-

25

26

27

no.

crop

m

ent

peri

od

mea

n ex

trem

es

[a]

-"

-

0.11

0.

02 to

0.1

9 1

Nig

eria

Ib

adan

A

ina

et a

l. ( 1

979)

-‘.-

pl

ante

d on

con

tour

on

4 ou

t 0.

38

1 In

done

sia

- K

eers

ebilc

k ( 1

990.

p.

570)

of

8 d

iffe

rent

Ind

ones

ian

soil

type

s -

'b -

hand

tilla

ge;

slop

e 5.

5%;

0.15

4 B

rasi

lia

6 B

razi

l L

epru

n et

al.

( 198

6. p

. 22

8)

29

30

3 1

U

?

4

b -2

n 0 5. %

5.5

6

0.23

-

Rw

anda

-

Lew

is (

I986

in:

swee

t Y

oung

198

9, p

. 33)

po

tato

es

0.5

- -

toba

cco

2nd

cycl

e R

oose

(19

75, p

. 40)

0.45

-

Rw

anda

-

Lew

is (

1986

in:

toba

cco

You

ng 1

989.

p. 3

3)

.

Annex 3.4

Detailed C factors

no.

6 7 8 9 10

11

12

noti

llage

de

scri

ptio

n C

fac

tor

mea

sure

- co

untr

y lo

cati

on

liter

atur

e m

ent

peri

od

mea

n ex

trem

es

[a]

-"

-

seed

sli

t rip

ped

betw

een

0.8

1 0.

6 to

1.1

8 4

Zim

babw

e D

ombo

shaw

a V

ogel

( 1

992,

p. 1

3)

form

er m

aize

stu

bble

(a

cros

s sl

ope

of 4

.5%

) -".

as a

bove

0.

51

0.37

to 1

.0

4 - " -

M

akah

oli

- " -

with

res

idue

s ca

lcul

ated

fro

m c

assa

va;

0.17

0.

l I

to 0

.37

2 ye

ars

Col

ombi

a S

anta

nder

de

Rei

ning

(19

9 I,

p.

1 I 1

) sl

opes

7 -

13%

w

ith 3

Q

uili

chao

re

ps

-"

-

as a

bove

but

slo

pes

13 -

0.10

4 0.

06 t

o 0.

27

2 ye

ars

- " -

Mon

dom

o "

20%

w

ith 2

re

ps

-"

-

mai

ze r

esid

ues

left

in f

ield

; 0.

4 1

0.3

to 1

.18

4 Z

imba

bwe

Dom

bosh

awa

Vog

el (

199

2, p

. 13

) sl

ope

7 -

13%

-

+&

-

as a

bove

but

slo

pe 1

3 -

20%

0.

29

0.1

I to

0.7

7 4

- "

- M

akah

oli

- " -

-"

-

calc

ulat

ed f

rom

whe

at-

0.14

6

Bra

zil

Gua

iba

Lep

run

et a

l. (

1986

. p.

mai

ze r

otat

ion

plan

ted

on

230)

12

% sl

ope

Annex 4 Protection and management

Annex 4.1 Detailed support and management (P) factors

Detailed \upport and management (P ) tactor5

Table 41-2Annex: Detailed P factors for ridges

Tab

le 4

1-3A

nnex

: Det

aile

d P

fact

ors

for

mou

nds

no.

desc

ript

ion

P f

acto

r*'

mea

sure

- co

untr

y lo

cati

on

liter

atur

e m

ent

peri

od

mea

n

extr

emes

[a

] I

cass

ava

1.13

i

Ivor

y C

oast

A

diop

odou

m R

oose

( 1

975.

p. 3

9)

mou

nds

2 pi

neap

ple

0.29

-

- -

-

mou

nds

Tab

le 4

1-4A

nnex

: Det

aile

d P

fact

ors

for

bund

s

no.

met

hod

desc

ript

ion

P fa

ctor

*'

mea

sure

- co

untr

y lo

cati

on

liter

atur

e m

ent

peri

od

mea

n ex

trem

es

[a]

I st

one

bund

s 3%

slo

pe; 3

0 cm

hig

h st

one

0.27

1

Nig

er

Allo

kote

R

ose

&

bund

s ev

ery

80 c

m v

erti

cal

Ber

tran

d di

stan

ce, o

n co

ntou

r; I

st y

ear

(197

1. p

. 12

76)

3 -

--

as

abo

ve b

ut 2

nd a

nd 3

rd y

ear

0.05

-

-

3 ea

rthe

n bu

nds

like

sto

ne b

unds

no.

I bu

t w

ith

0.00

22

- -

less

sto

nes

whi

ch c

over

an

eart

hen

core

veg

etat

ed b

y na

tura

l he

rbes

4

-..-

as

abo

ve b

ut 2

nd y

ear

0.04

1

-"

-

- -

- -

1)e~ailrd support and rn:uiagcment (1') [actor-\

T(rhlu 4 1-5 Alrilc..~: Dcr(ril~d P,/irc~ot.sfi)r . bidfor- srr'iljs

Some usefill species for soil and water conservation

Annex 4.2 Some useful species for soil and water conservation

Tuhle 42- 1 Aizrze.~: U.st$ul sperie .~ , f i r ero.sioi~ c.ontt-01, greerz manurirzg, rnulcl? or- (.over c-rap

Annex 4.2

bota

nica

l co

mm

on

rem

arks

lo

ngev

ity

ecol

ogy*

' na

me

nam

e

biot

em-

rain

fall

pH

drou

ght

wat

erlo

g-

salt

sh

ade

liter

a-

pera

ture

[m

m]

ging

tu

re

Muc

una

ca-

velv

et b

ean

legm

inou

s co

ver

and

a '?

h '?

7 7

7 3

1 pi

tata

gr

een

man

urin

g cr

op

Pan

icum

G

uine

a gr

ass

good

on

rise

rs;

fodd

er

p >

20

600-

4.

5-6.

0 1

max

imum

40

00

Pas

palu

m

Bah

ia g

rass

co

ver

crop

and

ero

sion

p

>I2

80

0-

5.1-

>8.

1 1

nota

tum

co

ntro

l 1

00

0

Pen

nise

tum

K

ikuy

u gr

ass

eros

ion

cont

rol

in

P 8-

24

600-

5.

1-6.

0 1

clan

dest

i-

high

land

s 40

00

num

Pen

nise

tum

el

epha

nt

fodd

er a

nd m

ulch

P

> 8

60

0-

4.5-

5.5

1 pu

rpur

eum

gr

ass

4000

Styl

osan

- st

ylo

legm

inou

s co

ver

and

p >

30

1000

- 4.

5-5.

0 -

1 th

es g

uia-

gr

een

man

urin

g cr

op

4000

ne

nsis

Vig

na u

n-

cow

pea

legu

me

a >

13

60

0-

4.5-

>8.

1 1

guic

ulat

a 30

00

Eup

horb

ia

live

hedg

e. p

ropa

gate

d p

1 ba

lsam

ifer

a by

cut

ting

s or

see

ds

Jatr

opha

po

urgh

re

live

hedg

e se

eded

: sl

ow

p 7

curc

as

esta

blis

hmen

t: c

uttin

gs

seve

rely

atta

cked

by

term

ites

bota

nica

l co

mm

on

rem

arks

na

me

nam

e

Aga

ve s

isa-

si

sal

live

hedg

e: s

omet

imes

la

na

not

acce

pted

clo

se t

o ho

mes

tead

s (h

ide-

away

fo

r sn

akes

) A

ndro

po-

Gam

ba g

rass

ero

sion

con

trol

; fod

der

gon

gaya

nus

long

evit

y ec

olog

y"

biot

em-

rain

fall

pH

drou

ght

wat

er-

salt

sh

ade

liter

a-

pera

ture

[m

m]

logg

ing

ture

P

> 2

3 <

300-

5.

6->

8.1

+

3 -

Bra

char

ia

Con

go g

rass

er

osio

n co

ntro

l: f

odde

r;

p >

I 7

,? - 3

ruzi

zien

sis

quic

k es

tabl

ishm

ent.

regr

owth

ver

y ka

riab

le

Styl

osan

- er

osio

n co

ntro

l: r

apid

p

7 >

? ?

1 ,?

'? 2

thes

ham

ata

regr

owth

Tep

hros

ia

gree

n m

anur

e on

cla

qe)

p =. 21

1000

- 1.

5-7.

5 -

3 vo

gelii

so

ils

1000

Pha

celia

ta-

ph

acel

ia

fodd

er.

co\e

r an

d gr

ee

a >

7 )

I

9

,? '?

3 na

ceti

folia

m

anur

ing

crop

Bot

hrio

- T

urke

stan

re

cult

ivat

ion

of e

rode

d p

8- I0

10

00-

6.1-

7.0

+ 3

chlo

a bl

uest

em

past

ures

1 1

00

1 iseha

emum

I

Pue

rari

a ku

dzu

legu

min

ous,

cre

epin

g p

> 2

2 10

00-

4.5-

7.0

- +

+ 3

phas

eolo

ides

co

\er

and

gree

n 30

00

man

urin

g cr

op. f

odde

r

Some u\et '~~I \pccics fo r soil and warcr conscr-vation

lJ\el'ul trees

Further information on suitable spec ies can be found in:

200 to 500

Acacia albida

Acacia radiana

Acacia senegal

Annona senegalensis

Balanites aegyptiaca

Roscia salicifolia Commiphora africana

Conocarpi~s lancifolius Dobera glabra Eupliorbia balsamifera Mael-va crassili)lia Parkinsonia aculeata I'rosopis juliflora Z i ~ i p h u s spp.

Young ( 1 989) Hudson ( 1975) von Maydell ( 1983) Merlier & Montegut ( 1982) ICRAFIGTZ Multipurpose TI-ce & Shrub Database

annual precipitation [mm] 500 to 900

Adansonia digitata Anacardiuln digitata

Azadil-achta indica

Bauhinia spp.

Cassia siamea

Combrctum spp. Eucalyptus camaldulensis Ficus syco~norus Haxoxylon persicurn Parkia biglobosa Salvadora persica Sclerocarya birrea Tamarix articulata Terminalia spp.

900 to 1200

A l b i ~ i a lebbeck

Anocgeissus

leiocarpus

Borassus aethiopum

Butyrospermum parki i

Casuarina

equisetifolia Cordia abyssinica Dalbergia

melanoxylon Erythrina abyssinica Mat-khamia spp. Tamarindus indica

Literature

Literature

Abdul Rashid, F. D. (1975): Studies on the correlation between the erodibility of a standard sand and five Malaysian soils. Unpublished R . Agr. Sc. Project Paper, University of Malaya, Kuala Lumpur, Malaysia.

Abdurachman, A.; Abujamin, S. & Kurnia, U. (1984): Soil and crop management practices for erosion control. Pernbl-. Pen. Tanah Dan Pupuk 3: 7- I?.

Aina, P.0; Lal, R. & Taylor, G.S. (1979): Effects of vegetal cover on soil erosion on an Alfisol. In: L A , R . & Greenland, D.J. (eds.): Soil physical properties and crop production in the tropics. John Wiley & Sons, Chichester: 50 1-508.

Allison, F.E. (1973): Soil organic matter- and its role in crop production. Develop-nients in Soil Science 3, Elsevier, Amsterdam: 637 pp.

Ambassa-Kiki, R. bi Lal, R. (1992): Surface clod size distribution as factor influencing soil erosion under silnulated rain. Soil Tillage Res. 22: 3 1 1-32?.

Armon, M.N. (1984): Soil crosion and degradation in south-eastern Nigeria in relation to biophysical and socio-economic factors. PhD-thesis, Depart~i~ent of Agronomy, University of Ibadan, Nigeria.

Arnoldus, H.M.J. (1977): Methodology used to dctermine the maximum potential average annual soil loss due to sheet and rill erosion in Marocco. In: Assessing soil degradation. F A 0 Soils Bull. 34: 39-48.

Arnoldus, H.M.J. (1978): An approxiniation of the rainfall factor in the Universal Soil Loss Equation. In: DeBoodt, M. & Gabriels, D. (eds.): Assessment of erosion, John Wilcy, New York: 127-1 32.

Asoegwu, S.N. & Obiefuna, J.C. (1990): Preliminary evaluation of pineapple mixed cropping systems for protecting reclaimed gullys in the tropics: an experiment in south-east Nigeria. Land Degradation & Rehabilitat~on 2 , 3: 237-241.

Auerswald, K. & Eicher, A. (1992): Comparison of German and Swiss rainfall simulators: Accuracy of mcasurement and effect of rainiall sequence on runoft' and soil loss rates. Z. Ptlanzenernihr. Bodcnk. 155: 191-195.

Auerswald, K.; Kainz, M.; Schriider, 1). & Martin, W. (1992a): Cornpal-ison of German and Swiss rainfall simulators - Expcril-nent:~l setup. Z. Pflanzcnerniihr. Bodenk. 155: 1-5.

Literature

Auerswald, K.; Kainz, M.; Wolfgarten, H.J. & Botschek, J. (1992b): Comparison of German and Swiss rainfall simulators - Influence of plot dimensions. Z. Pflanzenernahr. Bodenk. 155: 493-497.

Auerswald, K.; Sinowski, W. & Hausler, W. (1992~): Gew" asserver- sauerung und Gewiissereutrophieri~ng - iikologische Folgen lateraler Stofffliisse in Biiden. Landwirtschaftliches Jahrbuch 2: 87-96.

Barfield, B.J. & Albrecht, S.C. (1982): Use of a vegetative filter zone to control fine- grained sediments f'rom sul-face mines. Symposium on Surfaces Mining Hydrology, Sedimentology and Reclamation, Univ. of Kentucky, Lexington, Kentucky, Dec. 5- 10, 1982; 48 1-490.

Becher, H.H. (1990): Cornparison of German and Swiss rainfall simulators - Relative drop energies and their distribution in time and space for simulated rains. Z. Pflanzenerniihr. Bodenk. 153: 409-4 14.

Bennett, H.H. & Chapline, W.R. (1928): Soil erosion a national menace. US Dep. Agric., Circ. 33.

Rlaikie, P.M. (1985): The political economy of soil erosion in developing countries. Longman, London.

Bonsu, M. (1980): Assessment of erosion under different cultural practices on a savanna soil in the northern region of Ghana. In: Morgan, R.P.C. (ed.): Soil conservation, problems and prospects. Proc. Int. Conf. Soil Conservation, Silsoe, U.K., John Wiley 8r Sons, New York: 247-253.

Bonsu, M. & Obeng, H.B. (1979): Effects of cultural practices on soil erosion and maize production in the serni-deciduous rainforest and forest-savanna transitional zones of Ghana. In: Lal, R. 8r Greenland, D.J. (eds.): Soil physical properties and crop production in the tropics. John Wiley & Sons, New York: 509-5 19.

Bouwman, A.F. (1989): Modelling soil organic matter decomposition and rainfall erosion in two tropical soils after forest clearing for permanent agriculture. Land Degradation & Rehabilitation, 1 : 125- 140.

Bresch, J. (1993): Die Erosivitiit der Niederschliige Kameruns. M .Sc. thesis. Lehrstuhl fiir Bodcnkunde, Technische Universitat Miinchen, Gerl~~any: 92 pp.

British Westafrican Meteorological Services (1954): Preliminary note on the climate of Nigeria. Manuscript: 4 pp.

Bundesamt fur Seeschiffahrt und Hydrographie (1991 a): Handbuch der Westkiiste Afrikas. NI-. 2040. Hamburg, Germany: 520 pp.

Bundesamt fur Seeschiffahrt und Hydrographie (1991 b): Handbuch der Siid- ilnd Ostkiiste Afrikas. Nr. 2046. Hamburg. Germany: 344 pp.

Rundesamt fur Seeschiffahrt und Hydrographie (1992):

Literature

Mittelmeerhandbuch. 111. Teil (1989). Nachtrag Nr. I . Nr. 2029. Hamburg, Germany: 363 pp.

Cackett, K.E. (1964): A simple device for measuring canopy cover. Rhod. J. Agric. Res. 2: 56-57.

Caine, N. (1980): The rainfall intensity-duration control of shallow landslides and debris flow. Geografiska Annaler. 62A: 23-27.

Carrara, P.E. & Carroll, T.R. (1979): The determination of erosion rates from exposed tree roots in the Piceance Basin, Colorado. Earth Surface Processes 4: 307-3 17.

Carter, C.E.; Greer, J.D.; Braud, H.J. & Floyd, J.M. (1974): Raindrop characteristics in south central United States. Transactions of' the ASAE 17, 6: 1033- 1037.

Charreau, C. & Nicou, R. (1971): L'amelioration du profil cultural dans les sols sableux de la zone tropicale skche Ouest-Africaine et ses incidences agronorniques. Agron. Trop., 26. 9: 903-978.

Chauvel, A.; Pedro, G. & Tessier, D. (1976): R6le du fer dans I'organisation des matkriaux kaoliniques. Science du Sol 2: 10 1 - 1 13.

Chin & Tan (1974): Effects of simulated erosion on the performance of maize (Zea mays) and groundnut (Arachis hypogeae): a preliminary assessment. Syrnp. on Classification and Management of Malaysian Soils, Kota Kinebalu, Malaysia.

Chorley, R.J. (1978): The hillslope hydrological cycle. In: Kirkby, M.J. (ed.): Hillslope hydrology. John Wiley & Sons, New York: 1-41.

Chromec, F.W.; El-Swaify, S.A. & Lo, A.K.F. (1989): Erosion proble~ns and research in Hawaii. In: Crouch, R.J. & Collison, J.E. (1989): An air injected, single nozzle rainfall simulator, designed for use on gi~lly sides. Australian J. Soil Water Cons. 2, 3: 37-39.

CIEH (1985): Isohyktes des pluies journalikres de frequence dkcennale. Cornit6 Interafricain d'etudes hydrauliques, B.P. 369, Ouagadougou, Burkina Faso.

CTFT (1973): Difense et ristauration des sols. Gampela 1967-72. CTFT (Ouagadougou), Burkina Faso.

Croouch, R.J. & Collison, J.E.: An air injectcd. single nozzle rainfall simulator, designed thr use on gully sides. Austr. J. Soil Water Cons.. 2. 3: 37-39.

Dangler, E.W. & El-Swaify, S.A. (1976): Erosion of selectcd Hawaii soils by sirnu-lated rainfall. Soil Sci. Soc. Am. J. 40: 769-773.

&lwaulle, J.C. (1973): Resultats de six ans d'obversations s i~ r I'krosion uu Niger. Bois et ForCts des Tropiclues 150: 15-37.

Ilissmeyer, G.E. & Foster, G.R. (1980): A guide for predicting shcet and

rill erosionon forest land. U.S. Dept. Agric., Forest Serv. Tech. Publ. SA TP- I I : 40pp.

Duke, J.A.; Terrell, E.E. (1974): Crop diversification matrix: Introduction. Taxon, 23, 5/6: 759-799.

Dumas, J. (1965): Relation entre 1'6rodibilitk des sols et leurs caractkristiques analytiques. Cah. Orstom, Ser. Pedal. 3, 4: 307-333.

Dunne, T. (1975): Sediment yields of Kenyan rivers. Unpublished report. Dunne, T. (1978): Field studies of hillslope processes. In: Kirkby, M.J. (ed.):

Hillslope hydrology. John Wiley & Sons Ltd.: 227-293. Dunne, T.; Dietrich, W.E. & Brunengo, M.J. (1978): Recent and past

erosion rates in semi-arid Kenya. Z. Geomorph. N. F., Supp1.-Bd. 29: 130- 140.

Durand, P. (1983): Lutte contre l'krosion. Synthese des essais apres la catnpagne 1983 5 Rushubi. Manuscript: '?pp.

Ekwue, E.I. (1991): The effects of soil organic matter content, raintidl duration and aggregate size on soil detachment. Soil Technology 4: 1 97-207.

El-Swaify, S.A. (1990): Research needs and applications to reduce erosion and sedimentation in the tropics. Proceed. Fiji Syrnp., June 1990, IAHS-AISH Publ. 192: 3- 13.

El-Swaify, S.A. & Dangler, E.W. (1977): Erodibility of selected tropical soils in relation t o structural and hydrologic parameters. In: Soil erosion: Prediction and control. Soil Cons. Soc. Am. Spec. Publ. 2 1 : 105- 1 14.

El-Swaify, S.A. & Dangler, E.W. (1982): Rainfall erosion in the tropics: A state of the art. In: Soil erosion and conservation in the tropics. Soil Sci. Soc. Am. Spec. Publ. 43: 1-25.

El-Swaify, S.A.; Lo, A.; Joy, R.; Shinshiro, L. & Yost, R.S. (1988): Achieving conservation-effectiveness in the tropics using legume intercrops. Soil Technology I: 1 - 12.

Elwell, H.A. (1980a): A soil loss estimation technique for Southern Africa. In: Morgan, R.P.C. (ed.): Soil conservation, problems and prospects. Proc. Int. Conf. Soil Conservation, Silsoe, U.K., John Wiley & Sons, New Y ork: pp. 28 1-292.

Elwell, H.A. (1980b): Design of safe rotational systerns. Department of Conservation and Extension, Zimbabwe, Manuscript: 50 pp.

Elwell, H.A. (1984): Soil loss estimation: a inodelling technique. In: Hadley, R.F. & Walling, D.E. (eds.): Erosion and sediment yield - Some methods of measurement and modelling. University Press, Cambridge, pp 15-30.

Elwell, H.A.; (1993): Summary of crop covcr mesurements for erosion and

runoff prediction and control 1959 to 1993 and 1984 analysis results. The Institi~te of Agricultural Engineering, Harare, Zimbabwe: 32 pp.

Elwell, H.A. & Gardener, S. (1975): Comparison of two techniques for measuring percent canopy cover of row crops in erosion research programmes. Department of Conservation and Extension, Salisbury, Zimbabwe, Research Bulletin 19.

Elwell, H.A. & Stocking, M.A. (1976): Vegetal cover to estimate \oil ero\ion hazard in Rhodesia. Geoderrna 15: 6 1-70.

Elwell, H.A. & Wendelaar, F.E. (1977): To initiate a vegetal cover data bank for \oil loss e\tirnation. Dcpart~iient of Conservation and Exten\ion, Salisbury, Zimbabwe, Ke\ea~-ch Bulletin 23: 42 pp.

Engelund, F. & Hansen, E. (1967): A nomograph on \ediment tran\port In alluvial streams. Teknick Verlag, Copenhagen, Denmark.

Falayi, 0. & Lal, R. (1979): Efl'ect of aggregate size and mulching on cl-odibility, crusting, and crop ernergence. In: Lal, R. Nr Greenland, D.J. (cds.): Soil physical properties and crop production in the tropics. John Wiley & Sons, Ncw York: 87-93.

F A 0 (1990): The conservation and rehabilitation of African lands. Koline, Italy: 38 pp.

F A 0 (1991): Network on erosion-induced loss in soil productivity. Report of a Workshop, Centre for Soil & Agroclimate Research Bogor, Indonesia, March 199 1 : 52 pp.

F A 0 (1993): The conservation and rehabilitation of African lands. FAO. Rorne, ARCl9014: 38 pp.

Firth, C.R. & Whitlow, R. (1991): Patterns of gullying in Zimbabwe. Geojournal 23, 1 : 59- 67.

Flanagan, D.C.; Foster, G.R. & Moldenhauer, W.C. (1988): Storm pattern effect on infiltration, riunoff, and erosion. Transactions of the ASAE 3 1.2: 414-420.

Fones-Sundell, M. (1992): Survival for whom? Farmer, government and donor perspectives on land degradation in sub-saharan Afl-ica. In: Tato, K . & Hurni, H. (eds.): Soil conservation for survival. Soil Water Cons. Soc.: 1 18- 124.

Foster, G.R. (1982): Modeling the erosion process. In: Haan. C.T.: Johnson. H.P. & Rrakcnsiek, D.L. (eds.): Hydrologic rnodeling of small watersheds. ASAE Monograph 5 : 297-380.

Foster, G.H. & Highfill, R.E. (1983): Effect of ter-races o n soil loss: USLE P factor valucs lijr terraces. J . Soil and Water Cons. 38. 1 : 48-5 1 .

Foster, G.H.; I,ombardi, F. & Moldenhauer, W.C. (1982): Evaluation of'

rainfall-runoff erosivity factors for individual storms. Transactions of' the ASAE 25, 1 : 124- 129.

Foster, G.R. & Meyer, L.D. (1972): Mathematical simulation of upland erosion by fundamental erosion mechanics. Present and Prospective Techn. for Predicting Sediment Yields and Sources. Proc. Sediment Yield Workshop, USDA Sediment Lab., Oxford Miss., Nov. 28-30. 1972, ARS-S-40: 190-204.

Foster, G.K.; Moldenhauer, W.C. & Wischmeier, W.H. (1982): Transferability of U.S. technology for prediction and control of erosion in the tropics. Am. Soc. of Agronomy and Soil Sci. Soc. Am. (eds.): Soil erosion and conservation in the tropics: 135- 149.

Foster, G.R.; Weesies, G.A.; Renard, K.G.; Yoder, D.C. & Porter, J.P. (1992): Support practice factor (P). In: Renard, K.G.; Foster, G.R.; Weesies, G.A.; McCool, D.K. & Yoder, D.C. (eds.): The Revised Universal Soil Loss Equation. Draft Version 1992. US Dept. of Agricul ture.

Fournier, F. (1962): Carte du danger d'erosion en Afrique au sud du Sahara. Cornmunaut6 Economique Europienne avec Commission de Coopiration Technique en Afrique, Paris.

Fournier, F. (1967): La recherche en krosion et conservation des sols dans le continent afi-icain. Sols Africains 12, 1 : 5-5 1 .

Galabert, J. & Millogo, E. (1973): Indice d'6rosion de la pluie en Haute Volta. CTFT, (Ouagadougou), Burkina Faso: 34 pp.

Gasser, W. & Ziibisch, M.A. (1988): Erdrutschungen und Mafinahmen der Hang-sicherung. Selbstverlag des Verbandes der Tropenlandwirte, Witzenhausen, Germany: 165 pp.

Ghadiri, H. & Payne, D. (1981): Raindrop impact stress. J . Soil Sci. 3, 2:41- 49.

Griffiths, J.F. (1972): In: Landsberg, H.E. (ed.): World Survey and Climatology. Elsevier, Amsterdam 10: 197- 192.

Griggs, G.B. & Gilchhurst, J.A. (1977): The earth and landuse planning. University of California, Santa Cruz, Duxbury Press.

Gunn, R. & Kinzer, G.D. (1949): The terminal velocity of fall for water droplets in stagnant air. J. of Meterology 6: 243-248.

Hettler, J. (1994): Hiichste Gefahrenstufe fiir die Wiilder im Wasser. Frankfurter Rundschau, 19. 07. 1 994: 6.

Herkendell, J. & Koch, E. (1991): Bodenzerstiirung in den Tropen. Beck' sche Reihe 436, Germany: 184 pp.

Hijkoop, J.; Van der Poel, P. & Kaya, B. (1991): Une lutte de longue

haleine. Amenagements anti-erosifs et gestion de terroir. Systemes de production rurale au Mali: Volume 2. Institut d'Economie Rurale, Bamako, Mali et Institilt Royal des Tropiques, Amsterdam, Pays-Bas: 154 pp.

Hoeblich, J.M. (1992): Le lavaka malgache: Une fhl-me d'krosion parfois utilisable. Reseau Erosion Bulletin 12: 255-268.

Horton, R.E.; Leach, H.R. & Van Vliet, R. (1934): Laminar sheet-flow. Amer. Geophys. Union Trans. 2: 393-404.

Hudson, N.W. (1961): Raindrop size distribution, kinetic energy and intensity of' sub- tropical rainfall: 127- 133.

Hudson, N.W. (1963): Raindrop size distribution in high intensity storms. Rhod. J. Agric. Rcs. 1 : 6- 1 1 .

Hudson, N.W. (1975): Field engineering for agricultural development. Clarendon Press. Oxford. UK: 225 pp.

Hudson, N.W. (1986): Soil conservation. Batsford Lirni ted, London, UK: 324 pp.

Hudson, N.W. (1991): A study of the reasons for success or failure of' soil conservation projects. F A 0 Soils Bull. 64: 65 pp.

Hudson, N.W. & Jackson, D.C. (1959): Results achieved in the measurement of erosion and runoff in southern Rhodesia. Third Inter- African Soils Conference, Dalaba, Africa Soils Bureau 2: 575-584.

Hurni, H. (1980): A nomograph for the design of' labour-intensive soil conservation measures in rain-fed cultivations. In: Morgan, R.P.C. (ed.): Soil conservation, problems and prospects. Proc. Int. Conf. Soil Conservation, Silsoe, U.K., John Wiley & Sons, New York: 185-210.

ICRAFIGTZ (1991): The multi-purpose tree and shrub database. Version 1 . Kainz, M. (1989): Runoff, erosion and sugarbeet yields in conventional and

mulched cultivation. Results of the 1988 experiment. Soil Technology Serie 1 : 103- 1 14.

Kainz, M.; Auerswald, K. & Viihringer, R. (1992): Comparison of German and Swiss rainfall sirnulators - Utility, labour demands and costs. Z. Pflanzenerniihr. Bodenk. 155: 7- 1 1 .

Kamphorst, A. (1987): A small rainfall si~nulator for the determination of soil erodibility. Netherlands Journal of Agricultural Sciences 35: 407- 4 15.

Kaye, C.A. (1959): Shoreline features and quaternary shoreline changes. Puerto Rico. U.S. Geological Survey Professional Papers No. 3 17-B: 49- 1 40.

Keersebilck, N.C. (1990): Erosion control in the tropics. In: De Boodt. M.F.;

Hayes, M.H.B. & Herbillon, A. (eds.): Soil colloids and their associations in aggregates. Plenum Press, New York: 567-576.

Kilewe, A.M. & Mbuvi, J.P. (1987): Evaluation of soil erodibility factors using natural runoff plots. East Afr. Agric. For. J. 53, 2: 57-63.

Kirkby, M..J. (1980) (ed.): Hillslope hydrology. John Wiley & Sons, New York: 43-72.

Kittler, G.A. (1962): BodenflulIS. Eine von der Agrarriiorphologie vernachlassigte Erscheinung. Bundcsanstalt fiir Landeskunde und Raumforschung, Selbstverlag, Bad Godesberg, Germany.

Kollmannsperger, F. (1979): Erosion - eine globale Gefahr. Deutsche Gesellschaft fiir Technische Zusammenarbeit, Eschborn: 63 pp.

Kanig, D. (1992): Thc potential of agroforestry methods for erosion control in Rwanda. Soil Technology 5 : 167- 176.

Kowal, J.M. & Kassam, A.H. (1977): Energy load and instantaneous intensity of rainstorms at Sa~naru. northern Nigeria. In: Greenland. D.J. & LaI, K. (eds.): Soil conservation and management in the humid tropics. John Wiley & Sons, New York: 57- 70.

La Marche, V.C., Jr. (1968): Rates of slope degradation as determined from botanicril evidence. White Mountains, Califoi-nia. US Geological Survey Protssional Paper 352, 1 : 34 1-377.

Laflen, J.M.; Foster, G.R. & Onstad, C.A. (1985): Simulation of individual storm soil loss for modeling the impact of soil erosion on crop productivity. In: El-Swaify. S.A.: Moldenhauer, W.C. & Lo, A. (eds.): Soil erosion and conservation. Soil Cons. Soc. Am.. Arikcny, Iowa.: 285-295.

I,al, R. (1976a): Soil erosion proble~ns on an Alfisol in westel-n Nigeria anti their control: 1ntcrn;itional Institute for Tropical Agriculture. Monograph No. 1 . Ibadan, Nigeria: 208 pp.

Lal, K. (1976b): Soil erosion on AISisols in Western Nigeria, 111. Effects of rainfall characteristics. Geodel-ma 16: 389-40 1 .

I,al, R. ( 1 980): Soi I conservation: Preventive and control measures. In : Morgan. K.P.C. (cd.): Soil conservation. prohlerns and prospects. Proc. Int. Conl'. Soil Conservation, Silsoe, 1J.K.. John Wiley & Sons, New York: 175- 1 8 1 .

Lal, R. (1983): Soil erosion and its relation to pr-oductivity in tropical soils. In: El-Swaify, S.A.; Moldenhauer, W.C. &L Andrew. L.O. (edh.1: Proceedings oT the International Conference of Soil Erosion and Conservation, Hawaii: 237-247.

Ial, R. (1990): Soil erosion in the tropics. Mc Graw Hill, lnc.. New York:

580 pp. Lal, R. (1902): Restoring land degraded by gully erosion in the tropics. Adv.

Soil Sci., 17: 123- 152. Lal, R.; Lawson, T.L. & Anastase, A.H. (1980): Erosivity of tropical rains.

In : DeBoodt, M. & Gabriels. D. (eds.): Assessr-nent of erosion. John Wiley. New York: 143- 12 1 .

Larsen, M.C. & Simon, A. (1993): A rainfiill intensity duration thr-eshold for landslides in a hutnicl-tropical environment. Puerto-Rico. Geografiska Annaler. 75 A, 1-2: 13-23.

Laws, J.O. (1941): Measurements of the hll-velocity of water-drops and I-aindrops. Transacl. Am. Gcophys. Union: 709-722.

Laws, J.O. & Parsons, [).A. (1943): The relation of raindrop-size to inten\ily. Transact. Am. Geopllys. llnion 24: 452-460.

[>envain, J.S.; Sakala, W.K. & Pauwelyn, P.L.L. (1988): Isoerodent map of' Zambia Part 11: Erosivity prediction and rnapping. Soil Technology 1 : 25 1-262.

Leprun, J.C.; Silveira, C.O. da & Sobral Filho, R.M. (1986): Efficacitk des praticlues culturales ;untikrosives testkes sous diff61-cnts climats brdsiliens. Cah. Orstom, Ser. P@dol. 22.2: 223-233.

Levy, G.J. & Van der Watt, H.H. (1988): Effects of clay mineralogy and soil sodicity on soil infiltration rate. S.-Afr. Tyclskt-. Plant Grond 5. 2: 92-96.

Lewis, L.A. (1986): Predicting soil loss in Rwanda. In: Beek. K..I.: Burrough. P.A. & Mc Cormack. D.E. (cds.): Quantified land e\.aluation procedures. Pi-oc. of the Int. Workshop. Washington DC.

27 April - 2 May 1986: 137- 139. Luk, S. (1983): Effect of aggregate size and microtopogsaphy on rainwash

kind rainsplash erosion. Z. Geomorph. N.F.. 27, 3: 283-295.

Lyon, T.C. & Buckman, H.O. (1922): The nature and properties of soils. The Mac-millan Co., New York.

Maene, I,.M. & Chong, C.P. (1979): Drop size distribution and erosivity of' tropical rainstorms under tlic oil palm canopy. Laporan Penyelidikan. Jabatan Sains Tanah, University Pcrtanian. Malaysia, Serdang. Sclangor: 8 1-93.

Mazour, M. (1992): Lcs Sac~eurs cle riscl~~c dc I'irosion en nappe dans Ic bassin versant d'Issel- Tlerncen Algcric. Reseau Erosion Bulletin 12: 300-3 1 3.

Mbagwu, J.S.C.; Lal, R. & Scott, T.W. (1984): Effects of desurftlcing Of

All'i\ol\ and Ult~\ol\ In wuthern Niger~a: I . Crop pel-formance. So11 Sci. Soc. Am. J . 48: 828-833.

McCool, D.K.;Krown, I,.C.;Foster, G.R.;Mutchler, C.K.& Meyer, L.1). (1987): R e v i d \lope stecpnes\ factor l-'o~ the Univer\al So11 Lo\\ Ecli~ation Am. Soc. Agr. E,ng..Mlch~gan: 1387- 1396.

McCool, D.K.; Foster, G.K.: Mutchler, C.K. 8r Meyer, L.D. (1980): Kev~\ed \lope length factor 1'01- the Unlver\al Soil Lo\\ Equation. Tran\act~on\ of the ASAk 32, 5: 157 1 - 1576.

McCool, D.K.; Foster, G.R. & Weesies, G.A. (1992): Slope length and \tccpne\\filctors (1,s) In : Kcnard, K.G.: Foster, G.R.; Wee \~e \ , G.A.; McCool, D.K.&L Yoder, 1>.C.(ed\.): Predict~ng \oil ero\ron by water: A guide to conscrvation planning with the Revised Universal Soil Loss Equation (RUSLE), Draftversion 1992.

Merlier, H.; Montegut, J. (1982): Adventices tropicale\: Flore aux \tade\ plantale el adulte de 123 esp$ce\ africaines ou pantropicales. Ministkre des Relations ExtCricure(;, Coopel-ation ct D6vcloppenient. France.

Meyer, L.D. (1988): Rainfill1 si111uIato1-s f'oi- soil el-osion research. In: Lal. R. (ed.): Soil erosion research methods. Soil Water COIIS. Soc. Am., Ankeny, Iowa: 75-95.

Meyer, I,.D. & Harmon, W.C. (1985): Scdi~iient losscs from cropland furrows of diffel-cnt gradient\. Transactions of the ASAE 28, 2: 348-353.

Moeyersons, J . (1989): La nature de 1'6rosion des versants au Rwanda. Institut Nat. de Recherche Scicntifiquc, Hutare, Republique Rwandaise. Pub. 33: 370 pp.

Mokhtaruddin, A.M. & Maene, I,.M. (1979): Soil erosion under ciifl'el-ent crop\ and management practices. In: Proc. Intern. Conf. on Agric. Engineel-ing in National Devcloprncnt, September 1979, Scrdang. Malay\ia, lJniver\ity of Pertania, Malaysia.

Moore, T.R. (1979): Rainl-all ermivity in East Africa. Creogl-afisha Annalel- 6 1 A. 3-41 147- 156.

Morgan, K.P.C. (1986): Soil ero\ion and conservation. Longman, UK: 298

PP- Mostafa, K.T. & Osama, A.E.K. (1992) (eds.): The world environriient

1972- 1992. TWO decades of challenge. Chapman &L Hall, London, England.

Musgrave, G.W. (1947): The yuantitative evaluation of factor\ in watcr el-o\iol~, a first approximation. J . Soil Water Cons. 2, 3: 133- 138.

Mtakwa, P.W.; I,al, R. & Sharma, R.H. (1987): An evaluation o f the

Universal Soil Loss Equation and field techniques for assessing soil erosion in a tropical Aif'isol in Western Nigeria. Hydrological Processes I: 199-209.

Muller, J.P. (1977): Ida rnicrolyse plasniique et la differenciation des kpipkdons dans Ies sols I~I-rallitiq~~es rouges du Centl-e-Ca~neroun. Cali. Orstom, Ser-. Pkdol. 15. 4: 345-359.

Murphee, C.E. & Mutchler, C.K. (1981): Veril'ication of' the slope factor in the Universal Soil Loss Equation 1'0s low slopes. J . Soil Water Cons. 36, 5 : 300-302.

Mutchler, C.K. & Greer, J.D. (1980): Effect of slope length on erosion from low slopes. Transactions of the ASAE 23, 4: 866-869, 876.

Mutchler, C.K. & McCregor, K.C. (1983): Erosion from low slopes. Water Resources Research, 19, 5: 1323- 1326.

Mutchler, C.K. & Young, K.A. (1975): Soil detachment by raindrops. U.S. Agric. Res. Serv. Kep. ARS-S-40: 1 13.

Nelson, H.D. (1984) (ed.): Mo~ambiquc - a country study. Headquartel-s. i1epa1-trnent of'the Army, DA Par11 550-64, USA: 341 pp.

Ngatunga, K.I,.N.; ],a], R. & Uriyo, A.P. (1984): Effects of surfitce management 011 r~~nof f and soil erosion from some plots at Mlingano. Tanzania. Creodcri~~a 33: 1 - 12.

Nill, D. (1993): Soil erosion from natural and si~nulated rain in forest-, savannah- and highland areas of humid to sub-humid West Africa and influence of management. PhD thesis, Lehrstuhl fiir Bodenkunde, Technische Universitiit Miinchcn, Germany: 270 pp.

Nyagumbo, I. (1992): The influence of socio-econo~~lic fiictors o n potential adoption of' no- till tied-ridging in four coriir-nunal areas of Zi~nbabwe. In: Vogcl, H.: Conservation tillage Ihr sus1ainable crop production syste~~is. Pro.ject Report n o . 5 , P.O.Box BW 41 5, Bol-rowdale, Harare. Zimbabwe: 19-32.

Nyamulinda, V. (1991): Erosion sous cultures de versants et transports solides dans les bassins de drainage des hautes teri-es de Ruhengeri au Rwanda. Reseau Erosion Bulletin 1 1 : 38-63.

O'Laughlin, C.L. (1981): Tree I-oots and slope stability. Forest Research Institute. No. 104, Christchurch, New Zealand.

Oostwoud Wijdunes, D.J. & Bryan, R.B. (1991): Gully devclopmcnt on the Nje~nps Flats. Baringo, Kenya. Catena Supplc~nent 19: 7 1-90.

Othieno, C.O. (1975): Surface runoff and soil erosion 011 fields of young tea. Trop. Agriculture 52, 4: 209-308.

Owens, I,.B. (1974): Rates ot' weathering ancl soil formation on granite.

Literature

Unpubl. report, Dep. Agric. Univ. o f Rhodesia. Patil, P.Y. & Bangal, G.B. (1991): Effects of land slope and cultivation

practices on soil loss and runoff due to rainfall. J . of Maharashtra Agric~lltural Universities 16, 2: 155- 158.

Pauwelyn, P.L.L.; Lenvain, J.S. & Sakala, W.K. (1988): Iso-erodent map o f Zambia part I: The calculation of erosivity indices frorn a rainfall data bank. Soil Technology I: 235-250.

Petri, R. (1992): Erosionsdisposition (C-Faktoren) traditioneller Anbausysterne Kamesuns. M.Sc. thesis, Lehrstuhl fiir Bodenkunde, Technische Universitat Miinchen, Germany: 80 pp.

Pierce, F.J.; I,arson, W.E.; Dowdy, R.H. & Graham, W.A.P. (1983): Productivity of soils: A\se\\ing long-term changes due to erosion. J. Soil Water Cons.: 39-44.

Pimentel, D.; Terhune, E.C.; Dyson-Hudson, R. & Rocherau, S. (1976): Land degradation: Effects on food and energy resources. Science 94: 140- 155.

Poesen, J. (1984): The influence of slope angle on infiltration rate and Hoi-tonian overland flow volume. Z. Geomorph. N . F.. Suppl.-Bd. 49: 117-131.

Pontanier, R.; Moukouri Kouh, H.; Sayol, R.; Seiny-Boukar, I,. & Thebe, B. (1984): Comportemcnt hydriclue et \ensibilitk 5 IqCro\ion de quelques 501s du Nol-d Cameroun. Ministhe de I'Enseignement Supkrieure et de la Recherche Scientifique. Carneroun: 76 pp.

Quansah, C.; Raffoe-Bonnie, E. & Agyei, F. (1990): Runoff and soil lo\\ under four legume\. In: Ztibisch, M.A. (ed.): Land use and the environment. Proc. I Ith Ann. Meeting of the Soil Science Soc. o f Ghana: 67-75.

Raussen, T. (1990): Planung von Windschut~-Ptlan7unger1 in aridcn und semiariden Gebieten. Der Tropenlandwirt, Beihef't 43: 232 pp.

Rehm, S. & Espig, C;. (1976): Die Kulturpllan7en der Tropen und Subtropen. Eugen Ulmer, Stuttgart, Germany: 496 pp.

Reining, L. (1991): Charakterr\ierung und Verminderung der Bodenero5ion durch Washer in kleinbuuerlichen Maniokanbausy\temen. Beitrage /us Agrarwi\senschaften 4, Verlag w. Mahle, Wittenchlick. Bonn. Germany: 266 pp,

Renard, K.G.; Foster, G.R.; Weesies, G.A. & Porter, J.P. (1991): RUSLE- Reviscd U~liversal Soil Losg Equation. J. Soil Water Con\. 46, 1 : 30-33.

Rhoton, F.E.; Meyer, L.D. & McChesney, D.S. (1991): Depth-of-crogion

assessment using iron-manganese nodule concentrations in surface horirons. Soil Sci., 152, 5 : 389-394.

Rishirumuhirwa, T. (1992): Ruissellernent et erosion sous bananier au Burundi. Reseau Erosion Bulletin 12: 83-93.

Romkens, M.J.M.; Roth, C.R. & Nelson, D.W. (1977): El-odibility of selected clay subsoils in relation to physical and chemical properties. Soil Sci. Soc. Am. J. 41 : 954-960.

Roose, E.J. (1975): Erosion et ruissellement en Afrique de I'Ouest. Vingt annke de mesures en pitites parcelles experirnentales. Orstom, B.P. 20, Abidjan, Ivory Coast: 72 pp.

Roose, E.J. (1977): Use of the Universal Soil Loss Eq~~ation to predict erosion in West- Africa. In: Soil erosion, prediction and control. Soil Conserv. Soc. Am., Spec. Pub. 21, Ankeny, Iowa: 60-74.

Roose, E.J. & Bertrand, R. (1971): Contribution i I'ktude de la methode des bandes d'al-ret pour lutter contre I'erosion hydrique eri Afrique de I'Ouest. Agron. Trop. 26, 1 1 : 1270- 1283.

Roose, E. & Jadin, P. (1969): Erosion, ruisellement et drainage oblique sur un sol :I cacao de nioyenne CGte d'Ivoire. ORSTOMIIFCC., Abidjan, Ivory Coast: 77 pp.

Roose, E. & Piot, J. (1984): Runoff, erosion and soil fertility restoration on the Mossi Plateau (central Upper Volta). Challenges in African Hydrology and Water Resources, Proc. of the Harare Symposium, July 1984, IAHS Publ. 144: 485-498.

Roose, E. & Sarrailh, J.M. (1989): Erodibilite de quelques sols tropicaux. Vingt annkes de mesure en parcelles d'erosion sous pluies naturelles. Cah. Orstom, Sel-. Pkdol. 25, 1-2: 7-30.

Roth, C.B.; Darrell, W.N. & Riimkens, M.J.M. (1974): Prediction of subsoil erodibility using chemical ~nineralogical and physical parameters. Report for Office of Research and Development. U.S. Environ. Protect. Agency, Project No. 15030 HIX, Element I BB042, Manuscript: I 1 1 pp.

Ryumugabe, J.B. & Berding, F.R. (1992): Variabititk de l'indice d'agressivitk des pluies au Rwanda. Rescau Erosion Bulletin 12: 1 13- 1 19.

Sabel-Koschella, U. (1988): Field studies on soil erosion in the southern Guinea Savanna of Western Nigeria. PhD thesis, Lehrstuhl f u r Bodenkunde, Technische Universitat Miinchen, Germany: 180 pp.

Salako, F.K. (1988): Determination of potential sheet and rill erosion losses in southeastern Nigeria as a guide for soil conservation planning. Beitrage trop. Landwirtsch. Veterinarmed. 26, 2: 1 17-125.

Salako, F.K.; Obi, M.E. & Lal, R. (1991): Comparative assessment of

several rainSall erosivity indices in southern Nigeria. Soil Technology 4: 93-97.

Schauder, H. & Auerswald, K. (1992): Long-term trapping efficiency of a vegetated filtcr strip under agricultural use. Z. Ptlanzenerniil~r. Bodenk. 155: 489-492.

Schauer, T.H. (1975): Die Blaikcnbildung in den Alpen. Schriftenreihe des Bayer. Landesamtes fiir Wasserwirtschaft, Miinchen, Germany, 8. 29.

Schliephake, K. (1984): Tunesicn. Thiemernann, Stuttgart, Germany: 600 pp. Schuster, R.L. (1978): Landslides, analysis and control. Nat. Acad. Sci..

Washington, D.C., Special Report 176. Schwertmann, U.; Vogl, W. & Kainz, M. (1987): Bodenerosion di~rch

Wasser. Eugen Ulmer Vel-lag, Stuttgart, Germany: 62 pp. Segalen, Y. (1967): I.,es sols et la geornorphologie du Carneroun. Cuh.

Orstom. Ser. Pedol. 5, 2: 137- 187. Servico MeteorolGgico (1972): Cartas das isoietas midias mensais e annais

das ilhas de S. Tom6 c Principe. S. Tomt5: 59 pp. Shainberg, I. (1985): The effcct of exchangeable sodiuin and electrolyte

concentration on crust formation. Advances in Soil Science 1 : 10 1 - 122. Shainberg, I.; Gal, M.; Ferreira, A.G. & Goldstein, D. (1991): El'fkct of

water quality and amcndrnents on thc hydraulic properties and erosion fro111 several niecliterranean soils. Soil Technology 4: 135- 146.

Sheng, T.C. (1989): Soil conservation for small f'arrners in the humid tropics. F A 0 Soils Bull. 60: 104 pp.

Sheng, T.C. (1990): Runol'l' plots and erosion phenor-ncna on trol3ical steeplands. Research needs and applications to reduce erosion and sediinentation in tropical steeplands, Proc. Fiji Synlp.. June 19cNl. IAHS-AISH Publ. 192: 154- 10 1.

Sick, W.D. (1979): Madagaskar. Tropischcs Entwicklungsland zwischcn den Kontinenten. Wissenschaf liclle Ruchgcsellschaft Darmstadt. Germany: 32 1 pp.

Sinlonart, T.; Duchaufour, H.; Bizimana, M. & Mikokoro, C. (1993): La co~~servation dcs sols en milieu paysan burundais. Etude et h i e r a r c l i t i o ~ des stratt5gies anti-erosives. Reseau Erosion, Bulletin No . 13: 72-83.

Singer, M.J.; Rlackard, J. & Janitzky, P. (1980): Dithionite iron and soil cation content as Sactors in soil erodibility. In: Dc Boodt, M. 81 Gabl-iels, D. (eds.): Assessment of erosion. John Wiley. New York: 259-267.

Smith, D.D. (1941): Interpretation of soil conservation data lhr field use.

Agr. Eng., 22: 173- 175. Smithen, A.A. & Schulze, R.E. (1982): The \patla1 di\tl-ibution In southern

Africa ot rainh~ll ei-o\~\ ~ t y Ihr u\e In tlie Un~ver\al Soil Loss Equat~on. Watcr SA 8, 2: 74-78.

Stocking, M.A. (1981): Caujes and prcdictlon of tlie advance of gull~e\ . Proc. South-Ea\t Asian Rcg~onal Sylnp. on Problem\ of \ o ~ l ero\lon and \eci~mcntation, Banpkoh 27-29 Jan.: 37-47.

Stocking, M.A. (1988): Asscss111g vegetative coccr and management cfl'cct\. In: Lal. R. (ed.): Soil erovon rcscarch method\: 163- 185.

Stocking, M.A. & Elwell, H.A. (1976): Ka~ntall e ro \ i \~ ty o \er Rhoclt.\~a. Tranj. In\[. Br. Geogr. 1 , 2: 23 1-245.

Suchel, J.R. (1972): L a I-bpart~t~on dcs plules et le\ rkg~mc\ pluviorndt~-~qi~e\ ail Camcroi~n. Tracaux et Document\ de C;bogr,tphie Troptcale, n o . 5 CEGbT, T~tlence. France: 287 pp.

Sulaiman, W.; Maene, I,.M. & Mokhtaruddin, A.M. (1981): Runoff, \ o ~ l and nutl-ier~t lo\\c\ Ir-oln an liltisol under dif1i.1-ent legu~nes. SouthEa\t A\ian Regional Symp. on Pro171~1i1s ot Soil Ero\i011 and Secl~mentation, Bangkok, Tlia~land, 27-2C) Jan. 198 1 : 275- 286.

Sulaiman, W.H.W.; Mok, C.K.; Maesschalck, G. & Jamal, 'I'. (1983): Atlc;lncc\ 11) \ n ~ l ~uid w,lter conwrvatlon ic\e:~rch In Malay\la. Pel-tan~ha 6 (Kc\ . si~ppl. ) : 1 15- 132

Sutherlancl, K.A. & Bryan, R.B. (1990): Runotf and el-o\lon from a \mall semlal id catchment. Baringo I>i\trtct, Kenya. Appl~ed Geography 10. 91 - 109.

Suaziland (1990): Annual \tatist~cal bullctln. Central Stati\tical Off~co. P.O. Box 336. Mbabane. Swa71land. 148 pp.

'l'ato, K. & Hurni, H. (1992) (eds.): So11 conser\:~tlon Ihr \urv~val. So11 ~ulcl Water Con\. Soc.. Ankeny. Iowa: 420 pp.

l'enlple, P.H. (1972): Me:t\uremcnt\ ot runol'f and soil eroslon at an erosloll plot x a l e with particular I-cferencc to Tanranla. Geografi\ha Alinaler 54A: 203-220.

Temple, P.H. & Rapp, A. (1972): Land\lides In tho Mgeta area. We\tern Ulugun~ Mountains, Tanzania. C;cografi\ha Annaler, 54A.13-4: 157- 193.

'I'hornbury, W.D. (1985): Pnnc~ples of geomol-phology. John W~ley & Sons. Ncw York: 594 pp.

Torrent, J. & Rarrhn, V. (1993): Laboratory riiea\urc~nent ol' \oil colour: Theory and practice. In: Ripham, J.M. & Ciolho\z. E.I. (eds 1. So11 colou~.. Soil SCI. Soc. Anl. Spec. Pub. 3 1 : 2 1-33.

Ulsaker, I,.(;. & Kilewe, A.M. (1984): Kilnol'f and so11 erosion lor an Alii\ol

Literature

in Kenya. East African Agric. Forestry J. (Special Issue) 44: 2 10-23 1 . USDA (1972): SCS National Engineeing Handbook. US Dept. Agric., Soil

Cons. Serv., Section 4, Hydrology. USDA (1979): Proceedings of the rainfall simulator workshop, Tucson,

Arizona. ARM-W- 10. Northern Plains Soil and Water Research Center, Sydney, Montana: 185 pp.

Valentin, C. & Janeau, J.L. (1989): Les risques de degradation structurale de la surface des sols en savane hurnide (C6te d' Ivoire). Cah. 01-storn, Ser. Pkdol. 25. 1-2: 4 1-52.

Van Liew, M.W. & Saxton, K.E. (1983): Slope steepness and incorporated residue effects on rill erosion. Transactions of the ASAE 83: 1738- 1743.

Vanelslande, A.; Rousseau, P.; Lal, R.; Gabriels, D. & Ghuman, R.S. (1984): Testing the applicability of a soil erodibility nomogram for some tropical soils. Challenges in African Hydrology and Water Resources, Proceedings 01' the Harare Symposium, July 1984, IAHS Publ. 144: 463-473.

Vogel, H. (1988): Deterioration of a mountainous agro-ecosystem in the third world due to emigration of rural labour. Mountain Research and Development 8 ,4 : 32 1-329.

Vogel, H. (1992): An evaluation of five tillage systems for smallholder agriculture i l l Zimbabwe. Projcct Report no. 5, P.O.Box BW 415, Borrowdale, Harare, Zimbabwe: 1 - 18.

Von Gnielinski, S. (1986): Ghana. Tropisches Entwicklungsland an der Oberguineakuste. Wissenschaftliche Buchgesellschafi Dal-mstadt, Germany: 278 pp.

Von Maydell, H.J. (1983): Arbres et arbustes du Sahel. Schriftenreihe der Gesellschaft fur Technische Zusamrnenarbeit No. 147: 53 1 pp.

Walling, D.E. (1984): The sedirnent yields of African rivers. Challenges in African Hydrology and Water Resources, Proceedings of the Harare Symposium, July 1984, IAHS Publ. 144: 265-283.

Walling, I1.E. (1988): Measuring scdiinent yield from river basins. In: Lal, R. (ed.): Soil erosion research r-nethods. Soil Water Cons. Soc. Am.. Ankeny, Iowa: 39-73.

Weber, F.K. & Stoney, C. (1986): Reforestation in arid lands. Volunteers in Technical Assistance, 18 15 North Lyun Street, Suite 2000, Arlington, Virginia 22209, USA: 167 pp.

Wenner, C.G. (1977): Soil conservation in Kenya. Land and Management Division, Ministry of Agriculture, Nairobi, Kenya: 205 pp.

Wenner, C.G. (1989): Soil and water conservation in the farming areas of Lesotho: A review and some proposals. Topics in Applied Resource Management 1 : 57-86.

Wiese, B. (1988): Elfenbeinkiiste. Erfolge und Problcme eines Ent-

wicklungslandes in den westafrikanisclien Tropen. Wissenschaf'tliche Buchgesellschaft, Ilarrnstadt, Germany: 303 pp.

Wilkinson, G.E. (1975): Rainfall characteristics and soil erosion in the rainforest area of Western Nigeria. Expl. Agric. 1 1 : 247-255.

Wilson, R.C.; Torikai, J.D. & Ellen, S.D. (1992): Development of' 1-ainf'ull warning thresholds for debris flow in the Honolulu district, Oahu. U.S. Geol. Survey Openf'ilc Report 92-52 1 : 35pp.

Wischmeier, W.H. (1966): Relati011 of field-plot runoff to management and physical fllctors. Soil Sci. Soc. Am. Proc. 30: 272-277.

Wischmeier, W.H. (1975): Estimating the soil loss equations cover and n~anagcliient factor I'or undisturbed areas. U.S. Agr. Res. Ser. (Rep.), ARS-S-40: 285 pp.

Wischmeier, W.H. (1976): Use and rnisuse of the Universal Soil Loss Equation. J. Soil Water Cons. 3 1 , 1 : 5-9.

Wischmeier, W.H. & Mannering, J . V . (1969): Relation of soil properties to its erodibility. Soil Sci. Soc. Am. Proc. 33: 13 1 - 137.

Wischmeier, W.H. & Smith, D.D. (1958): Rainfall energy and its relationship to soil loss. Transactions, American Geophysical Union 39, 3: 285-29 1 .

Wischmeier, W.H. & Smith, D.D. (1978): Predicting rainfall erosion losses - a guide to conservation planning. IJ.S. Depar t l~~et~t of Agricultiirc, Agricilltilre Handbook No. 537: 58 pp.

Wischmeier, W.H.; Johnson, C.B. & Cross, B.V. (1971): A soil erodibility no~i~ograph for fart-nland and construction sites. J. Soil Water Cons. 26. 5: 189- 193.

Yoder, D.C.; Porter, J.P.; Laflen, J.M.; Simanton, J.R.; Renard, K.G.; McCool, D.K. & Foster, G.R. (1992): Cover management factor (C). In : Renard. K.G.; Foster. G.K.; Weesies, G.A.; McCool, D.K. & Yoder. D.C. (coord.): Predicting soil erosion by water. A guide to conservation planning with the Revised Univel-sal Soil Loss Equation (RUSLE). Draft version 1992. US Dept. Agriculture.

Young, A. (1989): Agroforestry for soil conservation. CAB International, Wallingford, UK: 276 pp.

Zingg, A.W. (1940): Dcgrcc and Icngth of land slope as it affets soil loss in runoff. Agric. Eng., 2 1. 2: 59-64.

Zobisch, M.A. (1986): Erfassung und Bewertung von Bodenerosions- prozessen auf Weideflichen im Machakos-Distl-ikt von Kenia. Der Tropenlandwirt, Beiheft Nr. 27: 220 pp.

I nde \

Index

10 year. rtorm 45, 92. 146, 2 1 0 Aggregate 47, 104 Aggregate \i7e 37 Aggregation 36. 38, 4 1 Alficol 66. 108, 147, 160 Algeria 165. 194 Alurniniunn 38 Andi\ol 60 Angola 179. I94 Antecedent rno~\ture 38, 62. 70 Arid 37 Aridisol 66. 137 Arsersment 69 Bambara nut 250 Banana 233 Harcl'allow 28, 35. 75 Hare Icvel 6 1 E3eanr 250 Bed loaci 25 Rcdd~ng 33 Bedload 22. 83 Rcnch terrace 53 Heni n I 95 Brological actrvity 10. 3 1 . 42 Botswana I95 Brick 50 Buffer strip 26 1 Buffcl-\trip 153 Bulk dcnrity 104. 106 Rund 41.48.51. 156 Bunds 260 Burkina Faso 165. 168, I95 Burilncli 105, 17 1 . 195 c' f~lctor I I7 C'abhagc 250 Ca~netwon 100, 122, 165, 168, 172,

195. 210

Canary Island\ 197 Canopy 2 1 7 Canopy cover 124 Canopy height 123 Cassava 238 Cation exchange capacity 47 C;ii~\e

-Illiteracy I4 -Povc1-ty 12

Causes -physical I I

Central Africa I98 Channel crorion 20 Chili 25 1 Clay 35, 36. 37. 47. 10 1 Climate 62 Compaction 20 Congo I99 Conservation I I Conservation bench terrace 53 Contour ridging 33 Contour-ridging 145 Contour-ir~g 32. 135, 258 Coral riff 2 1 Coghocton wliccl 75 Cotton 25 1 Cover 40.72, 1 17.2 17 Cowpea 25 1 Crop stage 1 17 Cropping cycle 37 Dam 20 Damage 45 Damage\ 15

-off-site 15. 20 -on-site 15

Deforestation 1 2, 20 Dcposition 31. 35, 47. 50

I>eprc\siion storage 22 Detachment 22 Detention \torage 22 Digue\ dever\ante\ 50 Digue\ filtrante\ 50 Disperjion 38 Ditch 158 Drainage ditch 5 1 Drop 69 Dunnc tlow 26 Dyhe 50 ti,a\t AI'rica 5 1 Economy 19 Egypt I 00 Encap\illated ail 38 Energy 32, 80 Equatorial Guinca 180. 199 f'ritrcu 199 E r o d ~ b ~ l ~ t y 30, 37. 66. 70, 75, 87, 99 I-,ro\ion 16

-geologic 1 I -\electrvt. I h

Eso\lon cycle 58 L ~ - o \ ~ o n nail\ 77 E r o \ ~ \ ity 32. 92. 1 17. 168 b,r-osi\ ~ t y ~ndice\ 33 Ethiopia 181. 199 Exchangeable catloti\ 37 Fallow 37, 125. 230 Fa l lo*~ 37 Fany~i Juu 5 1 . 158 Fel-tility 20 Fertili~er I9 Filter-strip 4 1 Filter-\trips 47 Flood 20. 50 Flow depth 23 F l o ~ belocity 23, 25 Flume 72 Fodder crop 232

Forest 1 I , 12. 125. 230 Foul-nier index 33 Furrow 45, 46. 146 Gabon 200 Geologic erosion 58 Geology 62 Ghana 182.200 Gradient 29, 43, 2 13 Gravel 43, 10 1 Gro~~ndnut 120, 343 C;roundwater 20 Growth curve 1 19, 226 Growth \tagc 72 Guinca 20 1 Gully 19, 58 Hail 93 Handtillage 42, 145 Hard wtting 28 Heaping 33. 45 Heap\ 156 Hematite 66 Hlghland 66 tIill\ide ditch 5 1 Hill\idc-ditch 41 Horton tlow 26 H u d w n index 33 Incepti\ol 06. 108. 147 Indicatol- 67 Infiltl-ation 22, 26, 36. 32. 48 Inwlberg 58 I~lctitution 13 Intensity 3 1 , 32. 33, 89 Intermittent terrace 55 Interrill 22, 39 Interterrace 53 Irish potatoes 352 Iso-el-odent n ~ a p 33, 34 I\o-erodent map5 9 1 Ivory Coast 166, 183, 20 1 Kaolinitc 35

Kenya 166, 169.20 1 Laboratory te\ts 70 Land tenure 14 Landscape 58 Land3lide 62 Leaching 18 Lemon grass 252 Lesotho 20 1 Level bench terrace 53 Liberia 184, 20 1 1,owland 66 LS factor 40, 1 1 1 Lybia 201 Madagascar 67, 166. 1 85, 202 Madeira 202 Magnesium 37 Maize 120, 245 Malawi 186, 203 Mali 203 Management 37.40 Manning 23 Marocco 173, 203 Mauritania 203 Mauritius 204 Migration 12 Mincral 35 Minerals 43 Mocarnbique 204 Mollisol 66 Mounds 260 Mo~arnbique 187 Mudtlow 62 Mulch 42,43, 1 17, 124, 2 17 Mu\grave equation 86 Namibia 204 Niger 166, 169. 204 Nigeria 166, 169, 188, 204 Not~llage 255 Orchard terrace 55 Organic carbon 17

Organic rnatter 16, 37, 47, 66, 10 1 Overgrazing 12, 63 Overland flow 23 Overpopulation 13 Oxide 35, 108 Oxides 70 Oxisol 35, 66, 147 P factor 145 Papaya 252 Parent material 15. 64, 108 Pediplain 58 Peneplain 58 Permeability 102 Pertneability class 102 Pineapple 235 Plastic foil 43 Ponding 22 Poverty 12 Pressures 3 1 Principe 190 Quintuples 46 R factor 32, 89 Rain drop 3 1 Rainfall 3 1

-Energy 3 1 Rain Fall sirllulator 69, 72 Refugees 13 Relative erosivity 1 19 Remobilization 84 Residual effect 125 Residue 42 Residues 23 1 Retention storage 22 Revised Universal Soil Loss Equation 39, 146 Ridge 259 Rill 19, 22, 29, 39, 45 Riparian filter strip 154 River 20, 67, 84 Root 37, 68, 8 1 , 160

Runoff 15, 28, 42, 45.48. 70, 1 14 Runoff coefficient 28 Runoff plots 75 Runoff velocity 40. 48. 114. 154 RUSLE 40 Rwanda 166, 169, 189 Sabre growth 63 Sahel 169 Saline 38 Sand 36.47 Sao Tom6 190, 205 Schist 62 Sealing 22, 26, 28, 29, 36 Seal 37 Season 3 1 . 38 Sediment delivery ratio 84, 159 Sedin~ent cnrichrnent ratio 17 Sediment traps 70 Seclilnont yield 84 Sei\rn~c activity 62 Selective r-enlovul 43 Serni-arid 37 Senegal 166. 205 Sheet erosion 22 Sheet flow 22 Sick \lope 43, 53 Side-slope 159 Sierra Leone 19 1 , 2C)5 Sighting frame 2 19 Silt 36 SI,EMSA X X Slope 38. 62

-concave 1 14 -convex 1 14 -uniform 1 14

Slope length exponent 40 Slope length 29. 39, l l I . 158, 2 13 S~nectite 36 Socio-economy I I Sod iu~~i 37. 42. 6 I

Soil -bulk density 15 -depth 15 -fertility 18 -formation I 1 , 16 1 -function 15 -functions 160 -organic matter 37 -productivity 15 -properties 36 -texture 15

Soil classification 66 Soil colour 66 Soil loss 37, 87. 1 18 Soil loss ratio 1 18 Soil solution 38 Soil weathering 16 1 Somalia 206 Soudan 192. 206 South Africa 174. 206 Soya 252 Splash 22. 72 Splash erosion 22. 26. 37 Stl-eatn bank ct-osion 20 Streambank erosion 84 S tr11ct 11re 1 5 Structure class 10 1 Subk~ctor 123 Subsoil 10 1 Sub~urface flow 6 1 Surface roughness 22 Surface storagc 42 Suspended loaci 2. 25 Swa~iland 207 Sweet potatoes 253 Tanzania 169, 207 Tchad 165 Temperature 38, 42, 66. 70 Terminal velocity 3 1 Terrace 41, 5 1 . 53. 158

Texture 36 Tied r~dging 43 T~ed-ridge 259 Tied-ridging 145 Tillage 30, 4 1 Tobacco 253 Tolerance 18. 160 Topography 38 Tran\port capacity 20, 23, 154 Ti~nl\la 193, 207 Tul-bulence 25 Uganda 169, 208 Ulti\ol 66, 108, 147, 160 Ulti\ol\ 28 U n i t plot 75, 86 Llniver\al Soil Lo\\ Equation 86 Vegetation 28. 35, 37. 62 Verti\ol 36, 66. 147 Vertiwls 28 Visco\~ty 38 Volcanic a\h \oil\ 106 Water capacity 37, 160 Water layer 3 1 Water mulch 38-42 Water retention 16 1 Waterlevel 2 11 Waterlogging 48 Watershed 20, 26, 28, 62, 67, 84 Waterways 53 Weir 50 Wo\t Al'rica 40. 177 Wind 3?

Yam 253 Y ~ e l d 160 Zaire 208 Zambia 167. 170, 175. 208, 2 10 Zimbabwe 167, 170, 176

Acknowledgments

The author\ thank the Deut\uhe Gcsellschaft fiir Techni\che Zusam~nenarbeit GmbH (GTZ) fol- \ponsci-ing the booh. We a150 ~ipprcciated very much MI-\. Schuhba~~cr 's tedious worh on the irlcluded maps.