MARDI Res. Bull. (1986) 14(3), (2t9-230)

AGGREGATE STABILITY AND SUSCEPTIBILITY TO EROSION OFSOME HIGHLY WEATHERED SOILS IN PENINSULAR MALAYSIA

H. GHULAM MOHAMMED*

Keywords: Aggregate, Structure, Erosion, Dispersion.

RINGKASAN

Walaupun tanah terluluhawa di rantau tropika lazimnya dianggap sebagai tanah yang beragregatbaik, terdapat bcberapa perbezaan penting dalam ciri-ciri agregat tanah tersebut. Sifat-sifat dankestabi lan agrcgat mempengaruhi kcupayaan tanah untuk menahan hakisan. Keputusan kaj ianmengutarakan dua faktor pcnt ing untuk kestabi lan agrcgat . ia i tu kandungan l iat yang mcncukupi dannisbah kandungan bahan organik dan kandungan l iat vang t inggi .

Empat s i r i tanah diu j i dengan hujan t i ruan selama satu jam dcngan intensi t i 35 mm scjam.Semuanya mengalami hakisan pcrc ik. Tetapi hakisan tersebut lebih t inggi bagi tanah yang kurang stabi l ,ia i tu s i r i Munchong I dan s i r i Bungor. Hujan t i ruern bcrturut- turut mcningkatk i in hakisan perc ik bagitanah vang kurang stabi l , tetapi t idak member i scbarang kesan pada tanah yang stabi l , sepert i s i r iRengam dan s i r i Kuantan. Tanah vang stabi l in i t idak mengalami perceraian agregat scmasa kaj ian in id i ia lankan.

INTRODUCTION

The soil erosion process is made up ofthree important phases: detachment,transportation and deposition (Er-l lsoN,l94l). The init ial and essential step,detachment, depends greatly on theinherent properties of the soil aggregates,specifically on their stabil ity against theforce of raindrops and of running water(GneeNrnNo, 1977). In addi t ion, theproperty of the soil to swell and slake whenwetted also affects its abil ity to resistdetachment by raindrops or by runoff. Thisis especially important when the soil issubjected to several rainstorms within a fewdays.

The stabil ity of soils againstdetachment is particularly relevant to theMalaysian agricultural scene. This isbecause of the continuing soil erosionproblems being exper ienced. Erosion inPeninsular Malaysia is attributed to twomain factors. Firstly, the tropical rainfallregime with high annual total, high intensityand frequent occurrence of wet spells(Dare, 1960; NreuwoI-r , 1981) is h ighlyerosive. Secondly. eros ion is accelerated bythe continuing replacement of the forest

vegetation with agricultural crops and byrapid urban development.

The most serious erosion affects thesedentary soils which are mostly found inthe inland areas, where slopes are ac()mmon topographic fealure. They arehighly weathered and leached, and belongto the Ul t iso l and Oxisol orders of Soi lTaxonomy. The highly weathered soils ofthe tropics have often been described ashaving very stable aggregation associatedwith high contents of clay and the hydrousoxides of iron and aluminium. However. amore detailed check wil l reveal that, amongthis group of soils, there are some which areunstable. This may be due to one or more ofthc binding substances being present insmall amounts only. A study of theaggregate stabil ity of nine soil types bySoouc (1913) revealed that some topsoils,for example the topsoi l (0-15 cm) of theSerdang series containing l1ok clay, aresignificantly less stable than others such asthe topsoils of Kuantan, Segamat and Prangseries with clay contents ranging from 447cto 64 per cent. Subsoils (15 - 30 cm) werefound to be less stable than topsoils (0 - 15cm) because of lower organic matter andhvdrous oxide contents. Areas where the

Ccntral Rcscarch l -a i rorator ics Div is ion. N{ARDI. Scrc l i rnq. Selangor. Malavsia

219

subsoil is exposed due to cultivation andother causes, are observed to suffer severesoil erosion.

The aim of this paper is to providefurther information on the aggregation ofsome sedentary soils in Peninsular Malaysia.Although it is generally known that soils oflower stabil ity are relatively more erodible,certain characteristics of soil aggregates andsoil moisture content modify this relationship.Some of the main points arising from alaboratory study on aggregate stability andsoil erosion are discussed.

SOILS

The soils of Peninsular Malaysia can bebroadly classified into three groups r.e.,sedentary, reworked and alluvial. The alluvialsoils occupy the coastal plains and river floodplains, whereas the sedentary soils aregenerally found in the inland areas. Sincethere are genetic differences between thesegroups, their physical characteristics alsodiffer. This paper focusses on a number of thesedentary soils.

The sedentary soils are developed bythe in-siru weathering of a variety of parentmaterials, including various types of igneous,metamorphic and sedimentary rocks. Theintense weathering under the humid tropicalclimate of Peninsular Malaysia results in basesbeing leached, silica solubilised, and iron,aluminium and manganese preferentiallyretained. Thus, the soil profiles are normally{ow in bases and silica, but rich in iron andaluminium. The quantities of these substancespresent depend on the composition of theparent rock. For example, Kuantan seriessoil, derived from basalt, is very rich in bothiron and aluminium, whereas Serdang series,derived from sandstone, has relatively loweramounts of those substances.

Humid tropical weathering conditionsand time have brought about mineralogicalchanges through physical disintegration andchemical reaction resulting in the formation ofclays. Thus, because of the advanced stage of

weathering, the clay content tends to be high.However, this property is sometimes alteredby terrain and the nature of the parentmaterial. Soils derived from sandstone orgranite, for example, tend to contain rela-tively higher amounts of quartz which isnormally of sand size. Those located on steepslopes tend to lose a larger proportion of thefiner material through erosion.

The soils used in this study are somesedentary soils of Peninsular Malaysia such asBungor, Kuantan, Munchong, Rengam andBukit Temiang. These soils show differencesin parent material, texture and pedogeneticfeatures as expressed by their classificationaccording to Soil Taxonomy (Table 1). Thetopsoil layer (0- 15 cm), the layer mostlulnerable to erosion, was investigated on.

METHODS

Mechanical Analysis

The soil sample which had passedthrough a 2-mm sieve, was used for thedetermination of particle size distribution.Organic matter was oxidised by hydrogenperoxide. A mixture of sodium hexameta-phosphate and sodium carbonate was used asdispersant, and the samples were physicallyagitated. Clay and silt were determined by thepipette method (Blecx, EvaNS, ENSIratNGeR,WHnE and Clenx, 1965).

Organic Matter Content and pH

The organic matter contents of the soilswere determined by the Walkley and Blackmethod. The values of soil pH in water weredetermined in a 1:5 soil to water ratio.

Iron and Aluminium Contents

'Total' iron and'total' aluminium wereextracted by aquaregia (CorreNIe, VeRLoo,VElcur, and KIBxr,Ns, 1979).'Free' iron and'free' aluminium were extracted by thedithionite-citrate-bicarbonate method(MEHna and JncrsoN, 1960). Iron andaluminium in solution were determined byatomic absorption spectrophotometry.

220

Table 1. Sampling location, classification, particle size distribution, pH and organic mattercontent of the topsoil layer (0-15 cm)

Soi lseries

ClassificationSampling accordingsrte to Soi l

Taxonomv

Part ic le s ize dist r ibut ion (%)Soi l

Clay Si l t F ine Coarse pH<2pm 2-20 sand sand ( in water)

pm 20-200 >2(npm pm

OM OM:Clavcontent rat lo

(vo)

Bungor

Kuantan

Munchong I+

Munchong I I+

Rengam

Buki t Temiang

Serdang,Selangor

Buki t

Goh ,Pahang

Serdang,Selangor

Serdang,Sclangor

Sg . Bu loh ,

Selangor

ChangkatBruasEstate,Perak

TypicPalaeudul t

Hapl icAcrorthox

TropepticHaplorthox

TropepticHaplorthox

TypicPalaeudul t

OrthoxicTropudul t

2.80 0.066

4.53 0 .071

2.8',7 0.040

3..18 0.052

3.68 0 .074

4 2 . 1 l . 7 3 3 . 3

63.5 25 .3 6 .5

7 1 . 3 6 . 8 1 1 . 1

6 7 . 1 1 3 . 4 1 1 . 5

49.6 6 . I 7 .7

22.9

1 0 . 5

8 . 0

36.6

4 . 8

A <

5 . 6

.{.5

22.3 10 .9 14 .U -52 .0 ,1 .8 1 . 9 2 0 . 0 8 6+Munchong I and Munchong II were collccted from two different sites. Some earth-moving activity appeared to have

taken place at the s i te of Munchong I . I t is expected that Munchong I I would be more l ikely to exhib i t thecharacter is t ics of the topsoi l of a benchmark Munchong.

Aggregate Stability

Aggregate stabil ity was init iallyassessed by subjecting aggregates to pressurebetween the thumb and forefinger. Then, itwas determined by two other methods,namely dry-and-wet sieving (De LeeNHrrnand De Boonr, 1961) and prolongedimmersion in water.

One hundred aggregates in the sizerange of 2 mm to 8 mm were subjected tomaximum pressure between the thumb andforefinger. The number of aggregates brokenwas noted to provide a preliminary assess-ment .

In the dry-and-wet sieving method,1 500 g of air-dried soil (diameter < 8 mm)was sieved to obtain four size fractions. i.e..4.76-8.00, 2.83-4.76, 2.00-2.83 and <2.00 mill imetres. From here the aggregationpercentage, that is the percentage of soilaggregates > 2.00 mm diameter, was thencalculated. For the wet sieving, 100 g of thefract ions of 2.00-2.83, 2.83-4.76 and

4.76-8.00 mm in the same proportion asobtained in dry sieving, were used. Each ofthe three fractions was placed in a nickel cupand was wetted by large water drops fall ingfrom a height of 50 centimetres. Sufficientnumbers of drops were added to bring thesoil moisture content to field capacity. Thewetted soil aggregates were kept in a high-humidity chamber for 24 hours.

The aggregates were then sieved underwater for f ive minutes. Sieves with thefollowing aperture sizes: 4.J6, 2.83, 2.00,1.00, 0.50 and 0.297 mm were used. Theamount of soil retained on each sieve wasdried and weighed.

The mean weight diameter (MWD)was calculated according to the formula ofDE LeBNneER and DE Booor (1967\.

MWD _ t % weig.ht mean diameter of

L/ ot soll x

the size fraction

The instabil ity index is given by thedifference between the MWD values of the

221

dry aggregate distribution and the wet

aggregate d is t r ibut ion.

To complement the results of the dry-

and-wet sieving method in characterizing

soil aggregate stabil ity, the extent of disper-

sion and slaking of the soils was also

assessed.

For each soil, about 300 aggregates in

the size range of 3 mm to 5 mm were

immersed in water for one week. It was

observed earlier that the largest amount of

stable aggregates was in this size range.

Slaking and dipersion in water were allowed

to take place without any disturbance. This

was followed by drying according to theprocedure described by GnpENE-KEt-t-v(re73).

The water was carefully replaced with

a solution of an organic solvent, methanol.The concentration of the methanol solution

used was progressively increased from25Vo,

50%,75Vo to 100 per cent . The solut ion was

changed every24 hours. The 1fi)7o methanol

solution was replaced with ether which was

allowed to evaporate giving d.y soi-

aggregates.

The methanol conferred hardness to

the aggregates, allowing size separation by

sieving. The fractions of sizes obtained were

0 . 3 - 0 . 5 , 0 . 5 - 1 . 0 . 1 . 0 - 2 . 0 , 2 . 0 - 2 . 8 , a n d2.8-4.7 mi l l imetres.

Soil Erosion and Runoff

In order to measure soil loss and

runoff, simulated rainfall was applied to

aggregates <8 in diameter. Four of the soils,namely, Munchong I , Bungor, Kuantan andRengam were selected for this investigation.These soils were chosen to represent a widerange of aggregate stabil ity. The MunchongI was preferred over Munchong II because itis much less stable than Kuantan, Bungorand Rengam. The Bukit Temiang was leftout so that the influence of high amounts ofcoarse sand and gravel in the erosionprocess was excluded in th is s tudy.

The air-dried soil sample was gently

packed into a soil pan (30 x 30 x 5 cm) and

tilted to a slope of 33 per cent. Slopes of

such steepness are considered as seriouslimitations to crop growth and are commonin the inland areas.

Free subsurface drainage from the soilpan was facilitated by a fritted false bottom

underneath the soil sample. The density of

packing was 1.1 g per cubic centimetre.

Simulated rain was aPPlied at anintensity of 35 mm/h for an hour using a

ra infa l l s imulator which had been descr ibedby Gasnle ls , DE Booor and MIN:nuw(1973). Runoff, splash and percolation were

collected at regular intervals.

Runoff was collected at the end of atriangular-shaped runoff concentratorattached to the front of the soil pan. Soilmaterial splashed onto the sides of the pan

was collected by placing metal splash boardsat about 6 cm away from each side of thepan. Percolation was collected throughopenings at the bottom of the Pan.

Rainfall intensity was calculated basedon readings of f ie ld ra in gauges.

RESULTS AND DISCUSSION

General Characteristics of the Soils

Particle size analysis, pH, organicmatter content and other information aregiven in Table I. Contents of iron and

af uminium are given in Table 2. Brief notes

on the soils are given in Appendix L

Characteristics of Aggregate Stability

The values of mean weight diameter(MWD), instabil ity index and aggregationpercentage are shown in Table 3. Thenarrow range in MWD (dry) indicates a

close similarity in the distribution of soil

aggregates in the air-dried state. A rela-

tively large proportion of these aggregates,ranging from 34.9Vo to 54.IVo are larger

222

Table 2. Iron and aluminium contents of the soils under study

Soil series

Iron content (7o)

Aquaregia extracl DCB extract

Aluminium content (70)

Aquaregia extract DCB extract

BungorKuantanMunchong IMunchong IIRengamBuki t Temiang

3 . 1 019.265 . 5 84.791 . 6 10.69

3 . 1 811.245.003.931 . 5 70.62

2.65t0.246 . 1 25. ' .713.2'73.5'7

0 .873. i l1 . 3 8t . 7 10.500 . 1 9

DCB : Dithionite-citrate-bicarbonate

Table 3. Aggregate stabil ity data obtained by dry-and-wet sieving

Soil series Aggregation (7o) Mean wt. diameterInstability

indexDry We1

BungorKuantan

Munchong IMunchong I IRengamBuki t Temiang

45.4334.9040. l052 .7049.1054. l0

4 . 4 14.294.40/ ^ 1

4 . 5 14.60

2.603.071 .802.78J . L J

2.82

1 . 8 11 . 2 22.601 . 6 41 . 2 81 . 7 8

Soi l ser ies

Table 4. Distribution of 3-5 mm aggregates after immersion in water for one week

Distribution (7o wlw) of 6 aggregate sizes (mm)

2.8-4 . '7 2 .0-2 .8 1 .0 -2 .0 0 . 5 - 1 . 0 0.3-0.5 <0.3

BungorKuantanMunchong IMunchong IIRengamBuki t Temiang

48.368.530.856.750.51 8 . 4

1 5 . 510.21 1 . 510.3I l . )

7 . 9

1 0 . 86 . 8

15.412.114.124.0

9 . 12. '76 .89 . 25 . 4

l 0 . l

8 . 76 . 1

1 9 . 86 .1

l 1 . tl - ) - - )

'7.6

5 . 7l f . /

5 .3o . 4

1 6 . 3

than two mill imetres. This is a favourablesoil structural characteristic.

The stability of the soils against theforces employed in the sieving method, thatis water drop impact and agitation in water,is shown by the values of MWD (wet) andinstabil ity index. These values show morevariation than the values of MWD (dry).Kuantan series was shown to be very stable,whereas Munchong I series was relativelyunstable.

However, in interpreting these results,one must also consider the different texturalcomposition of these soils. High gravel andcoarse sand content, as in Bukit Temiangseries, tends to increase the values of MWDas well as aggregation percentage. Thus, the

high MWD values and low instabil ity indexobtained for Bukit Temiang, contradict thefindings during an init ial assessment ofstability by pressing air-dried aggregatesbetween the thumb and forefinger(Appendix 2). Therefore, a more completeassessment of the aggregate stability has tobe made by also taking into account theextent of aggregate dispersion that occurredafter the air-dried aggregates wereimmersed in water for a week.

Table 4 shows the percentages of 3-5mm aggregates which have broken downinto the various size fractions. Soils withhigher percentages of smaller-sized aggre-gates are considered to be less stable thanthose with lower percentages of such aggre-gates. A careful examination of the data

223

reveals that Bukit Temiang and Munchong I

are relatively less stable, whereas Kuantanis the most stable (Appendix 2).

A categorization of these soilsaccording to aggregate stability is carriedout. This categorization takes note of boththe results of dry-and-wet sieving andprolonged immersion in water of 3-5 mmaggregates, the init ial stabil ity assessmentby pressing air-dried aggregates as well asthe textural differences of the soils. Ascoring scheme was devised for this purpose(Appendix 2). The aggregate stabilitycategories were:

Very stable - KuantanStable - Munchong II, Bungor,

RengamModerately - Munchong I

stableUnstable - Bukit Temiang

The process of aggregation depends onthe existence of clays. If the bathing l iquidhas a sufficiently high ionic concentrationand a large proportion of divalent ions suchas Ca*r, clays tend to flocculate (Russall,1973). The aggregation process is init iatedwhen the clays flocculate and cluster to formdomains. Substances which are said to binddomains together include organic matter

and sesquioxides. Thus, the high stabil ity ofKuantan series may be attributed to the highcontent of these substances (Tables I and 2).These constituents interact with each otherand are strongly held together by a com-bination of electrostatic, Van der Waals andhydrogen bonds, as suggested by Et--Swntrv( 1980).

The other soils, namely Rengam,Munchong II and Bungor, are also stablebut relatively less so than Kuantan. Thebinding of soil contituents in these soils isprobably less intense because of the rela-tively lower contents of the mentionedconstituents (Table 2).

However, high amounts of clay andsesquioxides alone do not necessarily ensurehigh aggregate stabil ity. The relative ease of

dispersion of Munchong I seems to point toa low organic matter: clay ratio (unpublishedinformation). It appears that its high claycontent (717a) requires a relatively highorganic matter content to aid in the bindingof all the clay domains. This indicates thatthe role of organic matter is very importantin stabil ising aggregates.

However, it is evident from the l itera-ture (Bnven, GanoNr,R and GanoNeR,1972; Russen, 7973) and from this studythat a prerequisite for the formation ofstable aggregates is the presence of clayparticles. Their active surfaces enable themto react with ions and the binding agents aswell as with each other. Thus, one finds thesoil with low clay content, Bukit Temiang,exhibit ing the lowest aggregate stabil ity.

However, in the case of BukitTemiang series, a small proportion of itslarge aggregates was found to exhibit astabil ity similar to that of the other soils,when pressed between the fingers. Whenthese large and stable aggregates were

specially selected and analysed for texturalcomposition, the clay content was found tobe reasonably high compared with that ofthe whole sample. Table 5 gives a com-parison between the clay contents ofselected 5-8 mm aggregates and those of

the whole samples. This substantiates theabove comment that a considerably high

clay content is a prerequisite for stableaggregation.

Erosion and Runoff Resulting fromConsecutive Rainstorms

Table 6 gives the erosion and runoffdata obtained after applying simulated rainof intensity 35 mm/h and of one-hourduration for three consecutive days oninit ially air-dried soil samples. During eachstorm, soil splash occurred almoslimmediately. The measured side splasherosion (SSE) generally increased as a rain-storm progressed (Figure I). Runoffoccurred only in the Munchong I andBungor series soils, starting on the first day

224

Table 5. Particle size distribution of 5-8 mm aggregates and clay content of whole samplesPart ic le s ize dist r ibut ion (% wlw)

Soil series 5-8 mm aggregates Whole sample

Clay< 2 p m

S i l t2-20 pm

Fine sand20-20O pm

Coarse sand> 200 pm

Clay< 2 p m

BungorKuantanMunchong IMunchong IIRengamBuki t Temiang

+ J . 4

63.270.769.164.943.1

6 . 426.86 . 5

12.48 . 2

12.6

30.5) . 1

12.01 1 . 89 . 0

12.9

19.74 . 9

1 0 . 86 . 4

1 1 . 93 l . l

4 2 . 163.51 1 . 367.149.6l l . - )

Table 6. Erosion and runoff from three rainstorms applied on consecutive daysVariable Bungor Kuantan Munchong I Rengam

Side splash erosion (g)

I st rainstorm2nd rainstorm3rd rainstormTotal

Soi l wash (g)

1st rainstorm2nd rainstorm3rd rainstormTotal

Runoff (cmr)

l st rainstorm2nd rainstorm3rd rainstormTotal

3 . 1 67.30

12.7223 .18

t . ,

0 . 1 00 . 7 80.88

0.791 . 1 9I .843.82

1.041'7.782 2 . 1 746.99

1 . 1 32 . 5 63 . 1 86.87

0 . l 60 .220.2-s0 .63

U40

558598

36804

1 2782 1 1 t t

000U

0t,,00

3rd darl 4

1 2

I (.)

o

6

3 1' a

e . . .

f ; ' so 1 6

' a 1 a

6= , ^

.? 108

6

1

2

Figure L Increase in side splash erosion withthe progress of rainstorms for Bungor and

Munchons I series soils.

for Munchong I and on the second day forBungor.

' l 'he time to incipient runoff

decreased on each successive day. ForMunchong I, the times taken were 48, 5 and3 minutes for the first, second and third dayrespectively. For Bungor, the durationswere 50 and 5 minutes for the second andth i rd day respecl . ive ly .

Under natural conditions in manyareas in the humid tropics, the peak inten-sity of rainstorms may greatly exceed 35mm/hour. For example, 28Vo of the rain-storms recorded between July 1978 andAugust 1980 in Serdang, Selangor hadmaximum 30-minute intensity (136) values of40 mm/h or more (GHur-ev MonavlrEoand ZAKr,1981). The peak in tensi ty va lueswould be much higher. The rainfall recordsalso show that, during rainy periods, stormsoccur on a number of days successively.Thus, it is common for high-intensity rain tofall on wet ground and, increase runoff anderosion on soils of relativelv low stabil itv.

0 l 0 2 0 3 0 4 0 5 0 6 0

Time (min)

225

-. 4i)

c -

! l o= - v

- 1 0

The results also show that soil splashtakes place on all four soils, indicating thatunder natural conditions this phenomenonis experienced on soils over a wide range ofaggregate stability, if they are not protectedby vegetation or mulch.

The Effect of Initial Moisture Content onRunoff and Erosion

The effect of initial moisture contenton splash erosion is shown in Figure 2. Theinit ial moisture content increases on eachsuccessive day because of the moisturereceived from earlier rainstorms. Table 6shows the increases in runoff and erosion oneach successive day.

. - ls t dav

r - 2nd day

o 3rd day

B BungorK - Kuantan

MI MunchongR - Rengam

20 30

Side splash erosion (g)

Figure 2. Effects of initial moisture contenton side splash erosion.

When init ially dry aggregates aresuddenly wetted, water enters the pores andcauses 'air explosion' (Bolr and KoeNlcs,1972) due to trapped air escaping from thepores. This results in either weakening orslaking of the aggregates. The extent ofweakening and the possibil i ty of slakingtaking place, depend on how strongly thesoil particles are held together by theaggregating agents. In soils which aresusceptible to this weakening process,erosion wil l take place more readily whenthey are further subjected to the impact ofraindrops. This behaviour is shown byMunchong I and Bungor.

Besides the increase in erosionrecorded on each successive day (Figure 2and Table 4), the phenomenon is also shownin Figure 1 and Table 7 where increases inerosion and runoff are observed during theprogress of each individual storm.

Table 7. Side splash erosion durirtg theprogress of rainstorms for the relatively

more stable soils, Kuantan andRengam ser ies

Time af ter Cumulat ive s ide splash erosion (g)

i n i t i a t i u n o f -ls t dav 2nd day 3rd daY

ralnstorm I mtn I

Kuantan

l 02030405060

Rengam

l 02030405060

0.040.070 . 1 30.360 .510.79

0.020.040.070 . 1 00 . 1 30 . l 5

0.250.420.600.'720.891 . 1 9

0.030.070 .120 . 150 . 1 90.22

0.370.'t40.98t . L - )

1 A 1

1 .84

0.020.050.090 .160-220.25

The Influence of Aggregate Stability onErodibility

The two soils with relatively loweraggregate stabil ity, Bungor and MunchongI, suffered runoff and soil wash (sheeterosion) in addition to splash erosion,whereas the other two soils. Kuantan andRengam suffered only minimal splash erosion(Table 6). The difference in soil erosion ofthese two pairs of soils is much larger thantheir differences in instabil ity index wouldsuggest. Thus, soils which deteriorate uponbeing subjected to rain, rapidly lose theirstabil ity when further stresses are applied.Those which are init ially resistant, maintaintheir stabil ity, as exemplif ied by Kuantanand Rengam.

A plot of instabil ity index againstsplash erosion suggests a l inear relationship(Figure 3). However, an anomaly occurswith respect to Rengam. This could be

226

o - Total for 3 stormsr - lst sJorm

B - BungorK - KuantanMI - Munchong IR - Rengam

0 2 0 4 0 6 0Side splash erosion (g)

Figure 3. Relationship between instabilityindex and side solash erosion.

explained by the generally larger aggregatesof Rengam when compared with those ofKuantan (Table 8). Larger aggregates areless easily splashed.

The splashed Kuantan soil are fineaggregates and not dispersed material. Theamount of splash increases slightly withhigher moisture content (Table 7) becausethe finer aggregates are splashed togetherwith the puddled water.

It is clear that aggregate dispersionoccurred only in Munchong I and Bungor.This dispersion led to crust formation andthe onset of runoff.

CONCLUSIONS

As many factors influence soilaggregation and aggregate stability, no onesingle method can completely compare thesoils according to their aggregate charac-teristics. Many factors have to be taken intoaccount when the relative stabilities of thesoils are studied. The factors include texture,instability index obtained by dry-and-wetsieving, and aggregate stability undersubmerged conditions. It was found that,when the clay content was sufficiently high,aggregate stability was influenced by theratio of organic matter content to claycontent.

Aggregate stabil ity is related toerosion resistance, with soils of low stabil itysuffering greater erosion. However, in thehigher range of stabil ity, aggregate size alsoplays a role, there being less splash erosionwhere aggregates are larger.

Aggregates with a high init ial moisturecontent are more easily eroded because theyare less cohesive when wet. However, this isonly true for the relatively less stable soils.The more stable ones have greater cohesionand suffer minimal splash erosion regardlessof the moisture content.

For a high resistance to erosion, thisstudy showed that a prerequisite was a

Table 8. Field description and laboratory data on aggregate size

Structural units described in the field Distribution+ (Vo wlw) of 4 sizes (mm)Soi lseries Type & class Size (mm) Grade 4.76-8.00 2.83-4.76 2.ff i -2.83 < 2.00

Bungor

Kuantan

Munchong I

Rengam

Medium &coarse SAB

Fine &mediumgranules

Fine &mediumSAB

Mediumgranules &fine SAB

10-2020- 50

2-5

5 - 1 010-20

2 -5

5 - 10

Moderate

Strong

Moderate

Strong

15.89

12.08

14.82

20.43

19.86

13.95

1 5 . 1 6

r7 .39

9.68

9.02

t0 .04

1 1 . 0 6

54.58

64.95

59.98

5 1 . 1 3

+Laboratory dataSAB : Subangular blocks

227

sufficiently high clay content. But this aloneis no guarantee. A wide organic matter: clay

AvlNuootN, Y. and EswnuN, H. (1973). Pedologicalstudies on Kuantan series - a representative soilo f Ma lays i a . MARDI Rep . No .11 , l 1pp .Serdang: MARDI Press.

BAvER, L .D . , GnnoNen , W.H . and GenoNEn , W.R .(1972). Soil Physlcs New York: John Wiley andSons.

BLAcK, C.A.. EvANs, D.D., ENsvtNcen, J.L. , WHrre,J.L. and Cr.qnr, F.E. (1965). Methods of Soi lAnalysis, Part I. Physical and mineralogicalproperties, including statistics of measurementand sampling Madison, Wisconsin: Am. Soc.Agron. Inc.

Bor-r , G.H. and KopNtcs, F.F.R. (1972). Physical andchemical aspects of the stability of soilaggregates. In Proc. Symp. Fundamentals of SoilCondi t ioning (ed. De Booor, M), Gent, 1972,pp .955 -73 .

CorrsNm, A., Vnnloo, M., Vet.cue, G. and KtrrrNs,L. (1979). Analytical Methods for Plants andSor is. Labo. Voor Analyt ischc en Agrochemie,Ri jksunivers i te i t Gent.

Dnle, W.L. (1960). The ra infa l l of Malaya, Part I I . / .Trop. Geog.14,11-28.

De LeeNHeen, L. and DE Booor, M. (1967). Aggregatestability by the change in mean weight diameter.ln West European Methods for Soil StructureDetermination Vol. 6, p. 28. RijksuniversiteitGent: Fak - Landbouw.

Et- t - tsoN, W.D. (1947). Soi l erosion studies. Parts I , I Iand Y. Agric. Engineering 28, 1 45 -6, 197 -201,

349-51.

El-Swntpv, S.A. (1980). Physical and mechanicalproperties of Oxisols. In Solls with VariableCha rge (ed . THENc , B .K .G . ) , pp303 -24 .N .2 . :Soc. Soil Science.

Accepted for publication on 25 September 1986

ratio and largenecessary.

aggregates are also

Gngnlnls. D. , DE Boonr, M. and MtNtnuw, W.(1973). Description of a rainfall simulator forsoil erosion studies. Mededelingan Fak.Landbouw, Rijksuniversiteit Gent 38 (2),294-303.

Gnuuv MoHAMMFTD, H. and Zert , G. (1981). Erosioncontrol by mulching during periods of erosiverainstorms. Proc. Conf. on Management oJTrop. Soi ls , Kuala Lumpur, 1981, pp. 25-37.Avai lablc at M's ian Soc. Soi l Science.

GnEpNE-Ksl lv , R. (1973). The preparat ion of c laysoils for the determination of structure. J. SoilSc i . 24 .277 -83 .

GpeENI-nNo, D.J. (1977). Soil structure and erosionhazard. In Soil Conservation and Manngement inthe Humid T' ropics (ed. GRF.FNDI.AND. D.J.and Lrr , R.) , pp. 17-23. New York: JohnWiley and Sons.

MeHne, O.P. and Je.cxsoN, M.L. (1960). I ron oxideremoval from soils and clays by a dithionite-citrate system bufferred with sodium bicarbonate.Proc. 7th Nat. Conf. Clays 1959, pp. 317-27.

NrEtrwolr, S. (1981). Agricultural droughts in PeninsularMalaysia. MARDI Special Report.

Russert-, E.W. (1973). Soil Conditions and PlantGrowth 10th ed. London: Longman.

SooNc, N.K. (1973). Soi l aggregat ion in PeninsularMalaysian soils and its influence on growth ofPueraria phaseoloides. Paper presented at Conf.on Fert. and Chemistry of Tropical Solls, KualaLumpur. Avai lable f rom Mal. Soc. of Soi l Sci .

WoNc, C.B. (1976). A study of a deep weather ingprof i le and a toposequence on grani te inPeninsular Malaysia. M. Sc. Thesis,Rijksuniversiteit Gent.

ABSTRACT

Although the highly weathered soils of the humid tropics are generally thought of as well-

aggregated soils, there are impertant differences in aggregate characteristics among them. These

influence their susceptibility to erosion. In the group of soils under study, the results highlighted two

factors which are essential for high stability: a sufficiently high clay content and a relatively high organic

matter to clay ratio.

Simulated rainfall of one-hour duration at 35 mm/h caused splash erosion on all series of soils, but

to a much greater degree on the less stable ones, Munchong I and Bungor. Successive rainstorms

increased soil splash considerably on the less stable soils, but produced negligible increases on the more

stable ones, Kuantan and Rengam series. The stable soils did not suffer dispersion under the conditions

of this experiment.

REFERENCES

228

Appendix 1. Description of the various soil series

Soil series Description

BukitTemiang Coarse-grained granite derivative.Sand and gravel contents are relatively high.Clay fraction comprises kaolinite halloysite, gibbsite and traces of mica.Silt fraction comprises kaolinite, gibbsite, quartz and mica (WoNc, 1976).Shallow (< 100 cm) and situated on upper slopes.

Bungor Derived from interbedded sandstones and shales.Sand content is relatively high with a greater proportion of fine sand.Clay fraction comprises mainly kaolinite, gibbsite and goethite.

Kuantan Derived from olivine basalts of Quartenary age, dark brown in colour, verystable microaggregation and very deep profile.Clay fraction comprises kaolinite and halloysite (357o), gibbsite (40%) andgoethite and magnetite (21Eo).Silt fraction comprises i lmenite, magnetite, kaolinite, quartz and gibbsite(ArrarNuoorN and EswnnRN, 1973).Occurs in a gently undulating landscape. (e.g. Bukit Goh Reserve, BukitGoh Felda Scheme, Jabor Valley and Southern Terengganu).

Munchong Derived from iron-rich shale. Occurs in undulating to roll ing terrain. Clayfraction dominated by kaolinite, gibbsite and goethite.

Rengam Derived from granite. Occurs in undulating, roll ing and hil ly terrain withslopes ranging from 10' to 25'. Relatively high sand content. Clay fractioncomprises kaolinite, halloysite, gibbsite, goethite and quartz (WoNc,1976). Silt fraction comprises kaolinite, gibbsite, quartz and mica (WoNc,| 976)

229

Appendix 2a. Scoring scheme for categorizing soils according to aggregate stabil ity

S c o r e 0 l 2 3

No. of aggregates that yield topressurer( ,2) 80- 100 50-80 20 -50 < 20

Instabi l i ty index2

Aggregate > 1 mm3 (o/a wlw) < 50 50-60 60 _75 75-100

Coarse sanda (7c w/w) > 50 ,10-50 20 -40 < 20rAggregate stabi l i t l - asscssed by pressing ai r -dr ied aggrcgatcs bctwccn the thumb and foref inger.-Oh ta i ned

f r om d rv - l nd -we t s r c r i ng .r? ' s table aggregates af ter prolongcd immersion in water.rTextural factor (Thc presence of coarse part ic lcs contr ibutcs to a lower instabi l i ty index. Thus, a lower scorc is g iven

for a h igher content of coarse sand to of fset the error) .

Appendix 2b. Scores for the soils used in the study

Var iable Buki t Temiang Bungor Kuantan Munchong I Munchong I I Rengam

Yield toD r e s s u r e 0 2 3 1 3 3

Instabi l i tyindex 1

Aggregates) 1 m m 1 2 3 1 3 3

Coarse sandc o n t e n t 0 2 3 3 3 2

Total 2 I 1 1 5 l 0 1 0

1 2 0 r 2

230

Dicetak oleh Unit Percetakan MARDI, Serdang, Selangor.Kementerian Pertanian Malaysia.