177

C.S. Chan and A.W. CheongJ. Trop. Agric. and Fd. Sc. 29(2)(2001): 177–187

Subsurface drainage effect on soil and rice crop(Kesan sistem saliran di bawah permukaan tanah terhadap tanah sawah dan padi)

C.S. Chan* and A.W. Cheong*

Key words: subsurface drainage system, productive rice yield, soil bearing capacity, watertable levels

AbstrakKesan sistem saliran di bawah permukaan tanah terhadap tanah liat lom berpasirdi MARDI, Seberang Perai telah dikaji dari tahun 1995–2000. Paip saliran dibawah permukaan tanah dipasang pada kedalaman 45 cm dengan jarak 2, 4, 8, 16dan 28 m masing-masing untuk menguji hubungan antara keupayaan galas tanahdengan kedalaman tanah untuk sistem saliran tersebut.

Kajian ini juga menilai keupayaan galas tanah dan kedalaman tanah yangberlainan terhadap jarak paip saliran. Aras air tanah yang berubah-ubahdisebabkan oleh jarak paip saliran yang berlainan dan kaitannya dengan hujanjuga dikaji. Keupayaan galas tanah meningkat dengan kedalaman tanah.Kedalaman 15 cm dan seterusnya mencukupi untuk menampung berat jentuai.Keupayaan galas tanah adalah lebih tinggi pada jarak paip saliran yang padatpada waktu hujan yang biasa, tetapi tidak begitu pada musim hujan atau kemarau.Hasil padi pada luar musim dan musim utama dari sistem saliran di bawah tanahtelah dibandingkan dengan sistem yang sedia digunakan di stesen. Jurang hasilantara dua musim tidak nyata sekali dalam sistem saliran di bawah tanah dengancatatan purata hasil luar musim lebih tinggi sedikit daripada musim utama. Puratahasil pada kedua-dua musim dari sistem saliran di bawah tanah umumnya lebihtinggi apabila dibandingkan dengan petak yang tidak mempunyai kemudahansaliran tersebut pada sebarang musim.

AbstractStudies on subsurface drainage were carried out on sandy clay loam soil inMARDI, Seberang Perai Research Station from 1995 to 2000. Subsurface drainsinstalled at a constant depth of 45 cm and variably spaced at 2, 4, 8, 16, and 28m were tested to establish a relationship between soil bearing capacity and soildepths in the installed subsurface drainage systems.

Additionally, soil bearing capacity at different soil depths in relation to thedrain spacing were evaluated. Fluctuation of the water table levels due to theeffect of subsurface drain spacing with reference to rainfall were also studied.Soil bearing capacity increases with soil depth which from 15 cm and beyond isstrong enough to ably support the weight of a large combine harvester. Soilbearing capacity is higher at denser subsurface drainage intervals only when withnormal rainfall, but not as indicative when at its extremes, being with too muchor too little rain instantaneously. Rice yields from the off-season and main seasonbetween the subsurface systems were compared with conventional yields

*MARDI Research Station, Seberang Perai, P.O. Box 203, 13200 Kepala Batas, Pulau Pinang, MalaysiaAuthors’ full names: Chan Chee Sheng and Cheong Ah Wah©Malaysian Agricultural Research and Development Institute 2002

178

Subsurface drainage effect on soil and rice crop

elsewhere in the station. It has been shown that in the subsurface drainage plotsthe yield gap between the two seasons was insignificant, but with the recordedoff-season average yield being slightly higher than the main season. Averageyields obtained under subsurface drainage systems were however noticeablyhigher when compared with harvest from fields without such a drainage facilityirrespective of the season.

IntroductionSubsurface drainage has been widely used indeveloped countries for the last fivedecades, but it was only during mid-1980sthat it began to be emphasized as a solutionto waterlogging and soil salinity problems inirrigated areas of the developing world(Chieng and Visvanatha 1997).

In the late1980s, the Japanesegovernment has encouraged the use ofsubsurface drainage for sustainable farmingof paddy fields. Over half a million ha ofpaddy field have since been installed withvarious types of subsurface drain in the landconsolidation projects (Ogino andMurashima 1992).

The main reasons for the installation ofsubsurface drainage in paddy field are toimprove the soil and create more conduciveworking conditions for the use of farmmachinery especially for large scale paddyplot farming as well as non-rice cropfarming on paddy field (Ogino andMurashima 1993). The installed subsurfacedrainage system serves to hasten surfacedrainage of excess rainfall during cropestablishment. This is to prevent totalseedling submergence. Besides, it helps toremove excess topsoil water within theplough layer by subsurface drainage duringcrop harvest.

Mid-season drainage generallyimproves rice plant growth and leadingprobably to crop yield increase subsequently(Matsushima 1970). The implementation ofsubsurface drainage system improves rootzone soil layer conditions, particularly bykeeping soil layer aerobic to activatefunctions of paddy roots which resulted inhigher yields than before, when soil layerconditions were anaerobic (Okamoto 1997).

The trenching type of subsurfacedrainage system is most commonly used. Itconsists of drain pipes and envelopematerials such as gravel and rice husks.Drain spacing varies from 7 m to 15 m, with10 m being a popular distance. Trench depthis about 50 cm and the width ranges from20 cm to 35 cm depending on the trencherused and diameter of the drain pipes(Murashima and Ogino 1994). Deteriorationof drainage capacity has been observed afterthe completion of drainage pipe installation.Research focusing on this problem revealedthe importance of the placement of envelopematerials to sustain the drainage capacity.Vlotman et al. (1997) provided scientificcriteria, from which guidelines are nowavailable to determine the necessity for drainenvelopes and the selection of envelopematerials to protect subsurface drainsinstalled in any soil type.

Stability of the soil can be evaluatedfor the approximation of opening sizesrequired to provide a safe and permanentdrain envelope. Logically, experiences ondrainage in the temperate zone cannot be ofany direct or accurate reference to the needsin the humid tropics (Warin et al. 1997).Prolonged heavy rainfall affects not only thedesign drain flow rates but also soil transfermechanisms. In particular, more surface run-off and interflow (lateral flow due toperched water table) are expected.

A mega subsurface drainage project,the Rajasthan Agricultural DrainageResearch Project (RAJAD) was initiated in1991 to conduct research and technologytransfer in subsurface drainage to controlsalinity and waterlogging problem and tolower the water table for sustainableagricultural production (Rakesh and Mundra

179

C.S. Chan and A.W. Cheong

1997). The RAJAD project scheduled to runfrom 1992 to 1999, covered an initial targetarea of 25 000 ha. The effect of subsurfacedrainage system on both water table depthand crop yield in Mit Kenana pilot area wasevaluated during the period from 1992/3 to1994/5 (Kenawey et al. 1997). Resultsshowed some obvious increase in cropyields of rice, maize and wheat, havingbenefited from the implementation of thesubsurface drainage system.

Economic analysis showed that thesubsurface drainage system had helped inincreasing crop yields as well as improvingsoil productivity and consequently the totaleconomic value of such a productionsystem.

From the economic point of view,material cost of denser spacing is moreexpensive than work cost for installationeven at a deeper level. Selection of themaximum drain depth is influenced to alarge extent by the in situ soil environment.Crop yield increase would be of primaryimpact despite extremely high initial cost ofsuch a drainage system development, whicheventually would be cost effective once landproduction is maximized (El-Hawary et al.1997). A wise and careful selection ofdrainage system design is to be expected inensuring the economic return be maximizedeven from an expensive capital outlay.Since the 1980s the assured and continualusage of fertilisers for both the seasons hasbeen speculated to be more yield enhancingin the main season for it to have overtakenthe normally higher yield in the off-seasonbefore that time. Generally lower yield inthe present off-season has also been said tohave resulted from inconsistent water,mostly being insufficient throughout thecrops but at times with sudden excessrequiring management attentionimmediately.

The purpose of this study is to recordchanges, on-farm water table levelsfluctuation and crop performance from theeffect of subsurface drainage at differentfield densities.

Materials and methodsStudies of the subsurface drainage projectwere conducted from 1997 to 2000 inMARDI Station, Seberang Perai where thesoil texture of the research plot may beclassified as sandy clay loam withproportions of sand, clay and silt at 55, 37and 8% respectively. Over the duration ofthis study, the same research plot has beenupgraded gradually and systematically asshown in the layout plan (Figure 1).

Perforated 3-inch high densitypolythene (HDPE) corrugated pipes wrappedwith a polyweave net were placed intrenches measuring 30 cm wide and 60 cmdeep. The pipes were then encased in amixture of sand and gravel to acircumference thickness of 10–15 cm(Figure 2). Finally, the trenches were filledwith dug earth to the original level. Theinstalled pipes, at a constant depth of 45 cm,were variably spaced at 2, 4, 8, 16 and 28 mapart. These pipes eventually drain throughthe wall into the open perimeter ditch. Thepurpose of using polyweave and envelopewas to prevent soil from entering into thepipe thereby averting sedimentation andblockage. A three-sided u-shaped concreteperimeter ditch with dimensions of one-meterwidth and depth was constructed forcollecting drainage water from thesubsurface drains.

In-field improvement done to the plotwere the installation of two water meters atthe inlet, proper land levelling and theconstruction of two drain outlet structuresthat can control field as well as perimeterditch’s water levels at different heights.

For this study, transplanting machinewas used to establish the crop in order toachieve crop uniformity. Three parameters,essentially to determine crop yieldperformance in general and detailed effect ofsubsurface drainage spacing on on-farmwater levels and machine trafficability wererecorded. The experiment was conducted asa randomised complete block design withthree replicates.

180

Subsurface drainage effect on soil and rice crop

Figure 1. Layout of subsurface drainage project plot in MARDI, Seberang Perai

A hand held penetrometer was used tomeasure soil bearing capacity after drainageat 10, 15, 20 and 25 cm below groundsurface. The measurement was first madewhen the field was completely drained (nostanding water) at about 2 weeks before thepaddy crop is due for harvesting.

During the same period, when water inthe field has been completely drained out,small tube wells were installed to measurefield water table levels. Each tube well was

made from a 50-mm polyvinylchloride(PVC) pipe of 150 cm length. Holes weredrilled around the pipe at regular intervals toabout 50 cm from one end. These holesshould enable water to flow in and out ofthe tube freely. The perforated end of thepipe was wrapped with plastic netting toprevent these holes from being blocked byloose soil and sedimentation. A hole with thesame diameter size was dug to a depth of120 cm in the middle of each treatment plot

181

C.S. Chan and A.W. Cheong

for the wrapped tube well to be inserted intoit. A water table indicator was used tomeasure the water levels which had earlierbeen allowed to stabilise.

Crop yield of each treatment wasestimated using crop cutting test (CCT)method with three samples for eachtreatment. The CCT plot was 3 x 3 m2 or itsequivalent of 9 m2. The harvested grainswere dried, winnowed and weighed, thenconverted to per unit area crop yield.

Direct seeded yields from cropsestablished differently outside the mainexperimental area of similar soil type butnot subjected to subsurface drainage maynot be analysed statistically together withyields resulting from subsurface drainagetreatments proper.

This may at least be compared broadlybased on collective information over a totalof nine seasons’ crops concurrently grownfrom within and outside the experimentalarea to highlight the probable effect ofsubsurface drainage.

Results and discussionEffect of different subsurface drain spacingtreatments on rice yieldThe results of rice yields under subsurfacedrain spacing of 2, 4, 8, 16 and 28 m areshown in Table 1. Analysis of varianceshowed no significant difference at 5% levelexcept for 97 off-season when comparingthe crop cutting yields within each season.Two cropping seasons, 98 off-season and98/99 main season, recorded much loweryields possibly due to the adverse effect ofAl Nino and La Nina respectively.

When compared with the stationaverage seasonal yield from main season1995/96 to off-season 2000, yields fromsubsurface drainage plot were higher exceptin 98/99 main season. This could be due tothe controlled flooding condition effect onthe subsurface drainage plot against the adhoc drainage as practised in the other fieldswithin the station. Comparing the overallcrop yield, the subsurface drainagemechanically transplanted trials yielded4.57 t/ha while only 3.58 t/ha was harvestedat the station direct-seeded row sown cropwithout any subsurface drainage.

Figure 2. Detail cross section of subsurface corrugated draincoil installation in MARDI, Seberang Perai

Soil surface

Backfill

Gravel & sand mixture

100-mm diameter corrugateddraincoil wrapped withpolyweave net

Dimension in mm100 100 100

150

150

100

50

182

Subsurface drainage effect on soil and rice crop

Comparison of off-season and main seasonyields between subsurface drainage trials toaverage station yieldsSeasonal average of mechanicallytransplanted crop yields from the subsurfacedrainage trials were 4.50 t/ha and4.63 t/ha respectively for the main seasonsand off-seasons extending from main season1995/96 to off-season 2000. For the sameperiod, the station’s normally direct-seededrow sown average yields were 4.04 and 3.01t/ha for the main and off-season croprespectively. Except for the main season98/99, seasonal yields from the station werehigher than yields from the subsurfacetreatments (Table 1). Indirect though it mayhave been, such difference may largely bedue to timely subsurface drainageparticularly for the off-season crop, as yieldpotentials between mere transplanting anddirect seeding seldom differ too extremely(Cheong 1995). Yield difference between themain and off-season were so much narrowerfrom the subsurface drainage treatment ascompared with the station’s average wherethere has been a marked drop in yield forthe off-seasons’ (Figure 3). Thus, the resultshave shown that subsurface drainage seemedto have benefited the off-season crops more

Table 1. Effect of different subsurface drain spacing treatments on rice yield

Treatment Seasonal yield (t/ha)spacing(m) 95/96 96 96/97 97* 97/98 98 98/99 99 99/00

(Main) (Off) (Main) (Off) (Main) (Off) (Main) (Off) (Main)

2 2.58 3.59 – 4.00a 6.70 4.38 3.07 5.91 4.814 3.50 3.62 – 4.67ab 6.21 3.72 3.26 5.17 4.618 3.91 4.11 – 4.76ab 6.60 4.10 3.06 5.59 4.71

16 4.17 4.17 – 5.54bc 6.35 4.01 2.97 5.77 4.8528 3.82 4.00 – 5.80c 6.64 4.19 3.21 5.57 4.89AVG 3.60 3.90 – 4.95 6.50 4.08 3.11 5.60 4.79AVG St 3.63 2.79 5.22 1.93 3.2 3.39 4.66 3.94 3.50Remark Normal Normal Ins Inf Normal Normal ElNino LaNina Normal E. Sub

*Means followed by a common letter are not significantly different at the 5% level by DMRTAVG = AverageAVG St = Station’s average from direct seeded row sown crops without subsurface drainage treatments(not analysed statistically together with subsurface drainage treatments) for general comparison onlyIns Inf = Install infrastructureE. Sub = Early submergence

0.00.51.01.52.02.53.03.54.04.55.0

Average station yields

Subsurface drainage yieldsOff-seasonMain season

Yie

ld (

t/ha)

Figure 3. Comparison of main season and off-season average yield from surface drainagetreatments against untreated composite yieldelsewhere in the station

than the main season crops in terms of yield,at least when sudden excess water may havebeen drained in time effectively throughsubsurface drainage.

Relationship between soil bearing capacityand soil depths under subsurface drainagesystemResults of soil bearing capacity for differentsoil depths at different subsurface drain

183

C.S. Chan and A.W. Cheong

spacings are as shown in Table 2. Soilbearing capacity generally increasesexponentially with increasing soil depth. Tosupport combine harvesting without causingsignificant damage to the soil surface ofpaddy field, the required soil bearingcapacity has to be 350 kPa. Recordedaverage values of soil bearing capacity weremuch higher than this required value whenthe soil depth is at 15 cm or more(Figure 4).

From the analysis of variance, the kParesults showed no significant difference at5% level due to the drain spacing at eachrespective depths. However, the surface

Table 2. Results of average soil bearing capacity (kPa) for different soildepths at different subsurface drainage spacings at 5 days after drainage(DAD) from 96 off-season to 99 main season at MARDI, Seberang Perai

96 off-season 97 off-season 98 off-season 99 main season

DAD1 5 5 5 53rain2 55.1 0 16.6 16.3

Soil bearing capacity at 5 cm depth2 m 237 474 64 584 m 216 416 55 608 m 183 426 61 72

16 m 187 364 68 8228 m 131 338 80 94

Soil bearing capacity at 10 cm depth2 m 419 760 486 1524 m 363 703 315 1078 m 314 801 445 154

16 m 310 699 392 18528 m 262 616 424 166

Soil bearing capacity at 15 cm depth2 m 607 945 696 3954 m 572 999 556 2548 m 503 906 729 499

16 m 475 914 713 50628 m 537 858 655 373

Soil bearing capacity at 20 cm depth2 m 835 873 850 6594 m 770 892 825 6448 m 827 958 849 782

16 m 788 957 860 78828 m 802 908 827 7131At least one recording before and several after, but not necessary atintervals with the same number of days were also made for each season2Cumulative of three consecutive days’ rainfall before soil bearing capacityrecording

layer soil dries faster at shallower depths(5 cm and 10 cm) than the deeper layers(15 cm and 20 cm) especially during dryspell as depicted in Figure 5. The rate ofsoil bearing capacity increased at 18 and14 kPa/day for the first five days afterdrainage for 5 cm and 10 cm depthsrespectively whereas 15 cm and 20 cmdepths remain nearly constant during thesame period. These phenomena happenedsimilarly to all respective spacingtreatments. The relatively consistent soilbearing capacity values at the depth around15–20 cm below ground level suggest that

184

Subsurface drainage effect on soil and rice crop

0

100

200

300

400

500

600

700

800

1999 main season1998 off-season1997 off-season1996 off-season

2015105Depth below ground surface (cm)

Soil

bear

ing

capa

city

(kP

a)

the plough pan most probably exists at thislevel.

Soil bearing capacity at different soildepths in relation to subsurface drainagespacingAnalysis of variance showed no significantdifference at 5% level when comparing thesoil bearing capacity at different depths tothe respective drain spacing treatments. Soilbearing capacity was normally higher on theaverage at narrower drain spacing when therainfall was moderate regardless of the soildepths (Figure 6). On the contrary, prolongand heavy rainfalls would always cause thereadings of soil bearing capacity notdifferent within the treatments. Soil bearingcapacity at topsoil layers, 0–10 cm belowground surface, appeared to be quitedependent on the duration and intensity ofrainfall while the lower layer at 10–20 cm,was more sensitive to waterloggedconditions generally.

Figure 4. Relationship of average soil bearingcapacity at different soil depths from differentcropping seasons at MARDI, Seberang Perai

0

2

4

6

8

10

12

14

16

18

20

54321

200

300

400

500

600

700

800

900

1000

days

Cum

ulat

ive

rain

fall

(mm

)

Soil

bear

ing

capa

city

(kPa

)

Cumulative rainfall (3 days)

5 cm depth

10 cm depth15 cm depth20 cm depth

Figure 5. Soil bearing capacity in relation to different soil depths with influence ofrainfall under subsurface drainage system

185

C.S. Chan and A.W. Cheong

Figure 6. Soil bearing capacity at different soil depths in relation to subsurface drainage spacing

0

10

20

30

40

50

60

70

80

90

100

Cumulative rainfall

242322212019181716151413121110987654321

Cum

ulat

ive

rain

fall

(3 d

ays,

mm

)

0

90

180

270

360

450

540

630

720

810

900

28 m drain spacing at 20 cm depth2 m drain spacing at 20 cm depth

28 m drain spacing at 10 cm depth2 m drain spacing at 10 cm depth

Soil

bear

ing

capa

city

(kP

a)

Days after complete field drainage (days)

Effect of subsurface drain spacing onwater table levelsAs illustrated in Figure 7, during the wetseason, the water table levels of the 2 mdrain spacing were always lower whencompared to 28 m drain spacing treatment.As the weather got drier, water table levelsfor both treatments dropped concurrentlyuntil at about 25 cm depth, about theposition of installed drain when water tabledepths of the wider spacing treatment begandropping lower than the 2 m drain spacing.This phenomenon may possibly be due tocontrolled drainage practised in the system,where water levels in the drains were kept atcontrolled depths not less than 30 cm fromthe ground surface except during fallowperiod. Usually, when controlled drainagewas practised, the closer spacing treatmentattained the highest soil moisture contentduring the dry period and the lowest duringthe wet period.

ConclusionEvery season’s rice yields were notsignificantly different between subsurfacedrain spacings of 2, 4, 8, 16 and 28 mexcept for 1997 off-season. The subsurfacedrainage trials yielded an overall 4.57 t/hawhile the station’s average yield withoutsuch treatment was only 3.58 t/ha over thesame time frame. As compared to the non-treated plots, yield difference between themain season and off-season were so muchnarrower from the subsurface drainagesystems in general. Apparently, with surfacedrainage, crop yield in the off-season maybe comparable with the main season’s butless so when without.

Soil bearing capacity increasesexponentially with increasing soil depth.Soil bearing capacity records also indicatedthat soil strength at more than 15-cm depthcan generally support a combine harvester(350 kPa) after complete field drainage.Values of soil bearing capacity at the deptharound 15–20 cm were consistently high

186

Subsurface drainage effect on soil and rice crop

indicating perhaps that the plough panprobably exists at this layer. The averagesoil bearing capacity was normally higher atthe closer drain spacings when the rainfallwas moderate but extreme rainfalls oftennullify this treatment effect. Water tablelevel fluctuated accordingly with changes inweather conditions. During the wet season,water levels of the closer drainage spacingwere usually lower compared to the widerspacing treatments. As the weather becamedrier, water levels of all treatments dropsimultaneously but with wider spacingsbeing faster and finally to a level lower thanthose at closer spacings.

Figure 7. Effects of subsurface drainage spacings on water table levels after complete fielddrainage during 96 main season at MARDI, Seberang Perai

0

10

20

30

40

50

60

70

80

90

100

Cumulative rainfall

807570656055504540353025201510510

5

10

15

20

25

30

35

40

45

50

28 m drain spacing2 m drain spacing

Days after complete field drainage (days)

Cum

ulat

ive

rain

fall

(3 d

ays,

mm

)

Wat

er le

vel b

elow

gro

und

leve

l (cm

)

ReferencesCheong, A.W. (1995). Rice direct seeding in

Malaysia. Paper presented at 2nd Asian CropScience Conference, 21–23 August, 1995, theFukui Prefectural University, Fukui, Japan

Chieng, S. and Visvanatha, N.A. (1997). Subsurfacedrainage design for sustainable developmentof irrigated Monsoonal areas. Paper presentedat 7th Int. Drainage Workshop. Drainage forthe 21st Century. 17–21 Nov. 1997, PulauPinang, Malaysia. Organizer: MANCID

El-Hawary, A.M., Kandil, H.M., Bazaraa, A.S. andAtrees, M.M. (1997). Maximum net presentworth design of subsurface field drainagesystems. Paper presented at 7th Int. DrainageWorkshop. Drainage for the 21st Century.17–21 Nov. 1997, Pulau Pinang, Malaysia.Organizer: MANCID

Kenawey, M.A., Abdel Ghani and Omara, M.A.(1997). Hydrological and economic impactsof subsurface drainage systems in MIT

187

C.S. Chan and A.W. Cheong

Kenana Pilot Area. Paper presented at 7th Int.Drainage Workshop. Drainage for the 21st

Century. 17–21 Nov. 1997, Pulau Pinang,Malaysia. Organizer: MANCID

Matsushima, S. (1970). Crop Science in Rice:Theory of yield determination and itsapplication. Tokyo: Fuji Publishing Co. Ltd.

Murashima, K. and Ogino, Y. (1994). Evaluation ofsubsurface drainage system for excess rainfall– Design on subsurface drainage in paddyfields (ll) – University of Osaka Prefecture,Series B, Vol. 46, 1994

Ogino, Y. and Murashima, K. (1992). Planning anddesign of subsurface drainage for paddies inJapan. Paper presented at 5th Int. Drainageworkshop. Feb. 8–15, 1992, Lahore, Pakistan.Organizer: ICID

–––– (1993). Subsurface drainage system of largesize paddies for crop diversification in Japan.Paper presented at Int. Commission onIrrigation and Drainage. Fifteenth congress.The Hague 1993. Organizer: ICID

Okamoto, M. (1997) Drainage in Japan. Countryreport in the 7th Int. Drainage Workshop.Drainage for the 21st Century. 17–21 Nov.1997, Pulau Pinang, Malaysia. Organizer:MANCID

Accepted for publication on 19 November 2001

Rakesh, H. and Mundra, S.N. (1997) Managingwaterlogging and salinity through subsurfacedrainage: Achievements and experiences ofRajad Project. Paper presented at 7th Int.Drainage Workshop. Drainage for the 21st

Century. 17–21 Nov. 1997, Pulau Pinang,Malaysia. MANCID

Vlotman, W. F., Willardson, L.S. and Dierickx, W.(1997) Drain envelope need, selection, designconstruction and maintenance. Paperpresented at 7th Int. Drainage Workshop.Drainage for the 21st Century. 17–21 Nov.1997, Pulau Pinang, Malaysia. Organizer:MANCID

Warin, F., Kao, C., Hendrickx, F. and Zimmer, D.(1997). Surface versus subsurface drainageimprovements in humid tropical regions.Paper presented at 7th Int. DrainageWorkshop. Drainage for the 21st Century.17–21 Nov. 1997, Pulau Pinang, Malaysia.Organizer: MANCID