The Swamp Sago Industry 9789814377898

Citation preview

The Swamp-Sago Industry in West Malaysia

I5EA5

INSTITUTE OF SOUTHEAST ASIAN STUDIES

The InstItute of Southeast As I an StudIes was estab II shed organ I sat I on In May )968. It Is a reg I on a I research centre other specialists concerned with modern Southeast Asia, multi-faceted problems of stabl I lty and security, economic political and social change.

as an autonomous for scho I ars and particularly the development, and

The Institute Is governed by a twenty-two-member Board of Trustees comprising nominees from the Singapore Government, the National University of SIngapore, the various Chambers of Convnerce, and profess I ona I and clvl c organisations. A ten-man Executive Committee oversees day-to-day operations; 1t Is chaired by the Director, the Institute's chief academic and administrative offIcer.

The Swamp-Sago Industry in West Malaysia a study of the Sunge1 Batu Pahat Floodpla1n

TAN Koonlin

INSTITUTE OF SOUTHEAST ASIAN STUDIES

Pub I Ished by Institute of Southeast Asian Studies Heng Mul Keng Terrace Paslr Panjang Singapore 0511 All rights reserved. No part of this publication may be reproduced, stored In a retrieval system, or transmitted In any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without the prior permission of the Institute of Southeast Asian Studies. © 1983 Institute of Southeast Asian Studies

ISBN 9971-902-66-4

CONTENTS

FIGURES

lX

PLATES

Xl

TABLES

Xili

ACKNOWLEDGEMENTS

XV

Introduct1on

THE PEDIGREE, ECOLOGY AND BIOLOGY OF THE SAGOPALM H1stoncal Role Ecology Palm B10logy Econom1c value of palm florescence

1 1 5 9

12

Chapter I

THE SWAMP-SAGO INDUSTRY IN WEST MALAYSIA Sago 1n the Stra1ts of Malacca Commerc1al Agr1culture 1n Johor Geography of the Sunge1 Batu Pahat Bas1n The Sagopalm 1n Malayan Agr1culture

v

15 15 21 29 37

Chapter II SAGO FARMING IN THE BATU PAHAT FLOODPLAIN Types of Farmers Cult1vat1on since the Japanese Occupat1on Trends 1n Cult1vat1on Attap mak1ng Chapter III SYSTEM OF SAGO

55 55 59 66 71

79

PROOUC~ION

Cultwat10n Hab1tat Plant1ng stocks Agronom1c pract1ces Sago product1veness Manufactun ng Raw mater1al supply Ownershlp of factor1es Process1ng Methods Meal Wet flour or crude starch Dry starch Pearl Process1ng Eff1c1ency

79 79 82 84 87 89 90 100 103 103

107 109

111 111

Chapter IV DEVELOPMENT OF A PROGRESSIVE SAGO INDUSTRY Cap1tal Investment Agronomy and Breed1ng Research Model of a Sago Production System The Soc1al V1nd1cat1on

119 119

123 129 138

APPENDICES I.

F1eldwork

Batu Pahat Dlstrict

141

II.

Gros5 Exports and Pr1ces of Sago and Tap1oca Products from Malaya-S1ngapore, 1925-1939

145

II I.

Sago Area in Malaya, 1951-1962

146

IV(a-c).

Types of Sago Cult1vat1on 1n Malays1a Barat, 1969-1979

147

1n

Vl

v.

MaJOr Sago Areas 1n Malays1a Barat, 1966, 1974 and 1979

150

VI.

Planted Sago Acreage under RISDA 1n Batu Pahat 01str1ct, 1956-1978

152

VII.

Sago Area 1n Batu Pahat 01str1ct, 1978

153

VII I.

Gross Sago and Tap1oca Exports 1n Malays1a Barat, 1969-1979

156

BIBLIOGRAPHY

158

ABOUT THE AUTHOR

175

Vll

FIGURES

1.

MaJor Freshwater Swamps of Malaya and Sumatra 1n 1964, and Sago Ports 1n the Early 20th Century

18

2.

Par1t System of Southwest Malaya

23

3.

Agr1cultural Geography of Bandar Penggaram, 1938

30

4.

Batu Pahat Sago Plant1ngs, 1974

33

5.

Expans1on of Sago Cultivation 1n the Batu Pahat Floodpla1n, 1966 and 1974

36

Sagopalmer1es 1n the Lower Sunge1 Kelantan Floodpla1n

37

7.

Locat1on of Sago Farmers Interv1ewed

57

8.

Compos1te Model of a Sago Enterpr1se on Par1t B1lal

99

6.

9.

Postulated Transport-Drainage System on Dryland and Regularly Inundated Sago Plantat1ons

lX

136

PlATES

A.

The Sagopalm:

Florescent, Fructescent, Senescent

B.

A Characterist1c Floodpla1n Hab1tat

6

C.

Sago Swampforest Complex

8

D.

Plantat1on-Type Sagopalmery

34

E.

The Plantat1on Sago Belt 1n Sunge1 Batu Pahat Floodpla1n

34

F.

Island Sago Groves in R1celand

35

G.

R1ver Bank Sago

72

H.

The Young Sagopalm

73

I.

Attap Sago

74

J.

Sago Harvest1ng

97

K.

Sagolog Pretreatment

98

L.

Pr1mary Sago Processing

104

M.

Sago Washwg

105

N.

Sago Process1ng

106

Xl

2

TABLES

1.

Sago Area 1n Malaya, 1931-1949

26

2.

Sagopalm Area 1n Malays1a Barat, 1960

27

3.

Types of Sago Cult1vat1on 1n Malays1a Barat, 1958-1979

40

4.

Sago 1n Johor, 1970-1979

41

5.

Notable D1str1cts of Sago Cult1vat1on 1n Malays1a Barat, 1966-1974

42

6.

Swamp and Sago Areas 1n Batu Pahat D1str1ct,

1966 and 1974

44

7.

Sago Replant1ng under RISDA in Johor, 1952-1978

47

8.

Rubberland and Replant1ng 1n West Johor Project Phase II, 1953-1975

48

Crops 1n Batu Pahat D1str1ct, 1966 and 1974

51

10.

Types of Sago Farmers Interv1ewed

56

11.

Hold1ngs of Sago Farmers Interv1ewed

58

12.

Sago Areas 1n Batu Pahat D1str1ct, 1966 and 1978

70

9.

13.

Sago and Tap1oca Starch Product1on 1n Malays1a Barat, 1972-1979

90

14.

Product1on of Sago-Tap1oca Factor1es 1n Malays1a Barat, 1967-1975

15.

Types of Sago Factory V1s1ted

101

16.

Cost of Mater1als 1n Sago Factor1es 1n Malays1a Barat, 1968-1973

112

17.

Extract1on Eff1c1enc1es 1n Sago Facton es

114

18.

Product1on of Sago-Tap1oca Factor1es 1n Malaysia Barat, 1973

115

19.

Sago Group Area 1n Malays1a Barat, 1977

126

20.

N1pa Area 1n Malays1a Barat, 1949-1979

129

21.

Plant1ng-Harvest1ng Cycles of the Sagopalm on a Plantat10n

130

XlV

91

ACKNOWLEDGEMENTS

The proJect was nour1shed by the f1rst Research Fellowsh1p 1n Southeast As1an Stud1es awarded by the St1ftung Volkswagenwerk, FRG, at the Inst1tute of Southeast As1an Studies, S1ngapore, dur1ng November 1978-0ctober 1979. I deeply apprec1ate the opportunity gwen by the Foundat10n and the Inst1tute to undertake study "10 a less v1s1ble but no less scient1f1cally excellent f1eld.... The appropnate technology for development needs a great deal of appropr1ate research" (Z1man 1978). Mat en a l part 1cul ar:

was

sought

from

the

fo ll owl ng

depos ltori es

in

the Inst1tute's Llbrary, and Ark1b Negara, Petal1ng Jaya and Johor Bahru, the Department of Agr1culture, Batu Pahat, and the L1brary and So1ls Div1sion of the M1n1stry of Agr1culture, Kuala Lumpur, under the ausp1ces of the Socio-Econom1c Research and General Planmng Un1t of the Pr1me M1mster's Department, Kuala Lumpur, 1n the case of the Malayan orgam sat 10ns. Figures 1, 2 and 4 were or1g1nally prepared by the Department of Geography, and reproduced by the L1 brary, Um vers1 ty of Sl ngapore, I thank a11 these establ1shments for the1r ass1stance. The many facets of the sagopalm industry were eluc1dated Wlth the splend1d ass1stance of Kerk Ah Thor and Abu b1n Korna1n, and the sago farmers and entrepreneurs of the Batu Pahat D1stnct, 1n part1cular HaJl AzlZ bin Taib, Na1b Jaafar bin Abdul Rahman, Enc1k Kamaluddin b1n Abdul Aziz, and Enc1k Mohd. Dom b1n XV

Lat 1m, and the van ous Mr Kerks and thew as soc 1ates, wh1 ch I recall w1th pleasure and apprec1at1on. I am grateful to S. Karth1 and her colleagues at the Inst 1tute for the1 r cheerful cooperat 10n 1n pre pan ng th1 s report and prov1d1ng appropr1ate ass1stance for the work dur1ng my SOJOUrn.

TAN Koonl1n, PhD ISEAS, S1ngapore 1 November 1979 Kuala Lumpur 30 Apr1l 1981

XVl

Introduction THE PEDIGREE, ECOLOGY AND BIOLOGY OF THE SAGOPALM

Historical Role Palms const1tute one of the oldest fam1ly of plants on earth, w1 th ancestors that appear to have been the precursors of the monocotyledons. Part 1a 1 to warmer c l1mes, they are most numerous The few spec1es that l1nger 1n 1n the 1ntertrop1cal zones. temperate latltudes mark the borders of an extenslVe realm that The1r doma1n l1es 1n A£1a, had flour1shed 1n warmer epochs. part1cularly Malaya, wh1le parts of Amazon1a poss1bly shelter More than any other, the palm heartland has another nucleus. surv1ved v1rtually 1ntact s1nce the Cretaceous era, for 1t occup1es that part of the earth least subJected to global cl1mat1c changes- the equator1al belt. Unt 11 the ascendancy of the Gram1 nae fam1ly, to wh 1ch the cereals belong, palms probably were the most bounteous, rel1able sources of food to man and beast. Descnbed as the "Pnnces" of the Vegetab 1e K1 ngdom (Seeman 1856), the1 r useful ness 1n 1 ands where they were endem1 c was noted before cere a 1 s became These dom1nant. d1etet1cally and 1nvas1ve geographlcally perenmals, when cropped, y1eld a bounty 1n shelter, food and dnnk that has g1ven rise to subs1stence strateg1es wh1ch d1ffer markedly from those based on the better-known annua 1s, 1. e. the Some yielded 1mportant commodities 1n cereals and tubers. colon1al commerce, espec1ally the oleag1nous palms. Several cultures have developed a largely selfsuff1c1ent domest1c economy centred round palm arbor1culture, e.g. swamp Amennd1ans of the Amazon-Onnoco on the monche, Maunt1a flexuosa, the Arabs of sub-Sahara on the datepalm, Phoen1x l

dactylifera, the South Indians on the palmyra, Borassus flabeiiJfer, the noneating Roti Islanders of Indonesia on the lontar, B. sundaicus (Fox 1977), the Indo-Pacific islanders on the coconut, Cocos nucifera, the West Africans on the oilpalm, Elaeis guineens1s, and the Papuans and Moluccans on the sagopalm, Metroxylon sagu Rottboll, and its wild kin, M. rumphii Martius. The sagopalm (Plate A) is a Malesian domesticate originating from Maluku-New Guinea; wild species proliferate on islands further east into the Melanesian foreland (Corner 1966). Several other starch palms favouring drier or hillier habitats, notably species of Ar;nga, Borassus, Caryota, Corypha and Eugeissona, were used in ike manner but, culturally and commerciany, they pale in productiveness, extent, significance and sophistication beside the swamp palm. Its selfsustaining ecosystem and ability to manufacture an enormous silo of starch in its stem nurtured planters and gatherers more efficiently, in terms of crop productivity and population capacity, than the environments inhabited by contemporaneous jung 1e hunter- foragers, swi dden and root crop cu 1t i vators (notably Di oscorea and Co 1ocas i a spp), and other arboriculturalists (pandan, Pandanus spp; breadfruit, Artocarpus spp; and banana, Musa spp). Over the centuries, the inexorable shift from vegetative cropping towards seed planting eastwards in Southeast Asia has left conspicuous vestiges only among the easternmost practitioners (Spencer 1966). But while the ubi {Ma~ay =yam) complexes could be delineated, the sago rema1ns hazy. Pre-Co 1umbi an Me 1aka and the ear 1i er northern East Coast cu 1tures of the Malay Peninsula were flourishing emporia sustained by the produce of swamp or river sago. Colonised by the northwesterly wet-rice culture - whose technology is derived largely from mainland civilisations yet manifests some peculiarly vegetative techniques of cultivation scions of the truly native sago culture sheltered in the marginal, remote lowlands of the farflung archipelago. By early this century it had become ext i net in many peri phera 1 is 1ands, such as Fiji and the Nt!W Hebrides. With the expansion of the more prestigious cereal culture in the early 19th century, sago acquired disrepute as food for the poor even in parts of its stronghold. "Sago is most abundant in the islands most distinguished for the production of clove and nutmeg and the geographical distribution seems co-extensive with that of these palms" (Crawfurd 1820). Few modern studies on forest resources consider the Metrox~lon palm as such, or that logging the humid forest for a staple ood or locally processed industrial convnodity ~ould be even more substantial than lumbering and the collection of famous palm products such as gums and rattans, that had only incidental or fraction a 1 va 1ue for its inhabitants. Because of the 3

1mportance of sago 1n the export economy, espec1 ally 1n Sarawak dur1 ng 1ts f1 rst 50 years of European rule, the north Borneo sultanates were obl1ged to rel1nqu1sh much of the1r terntones to Br1t1sh amb1t1ons 1n the reg1on. In 1ts cultural ecology, the sagopalm resembles the w1ld The sem1w1ld o1lpalm 1n Afr1ca more than 70 years earl1er. Afr1 can pa 1mer1 es were a 1 so gathered to supply a cheap vegetable 011 to Europe, g1 v1 ng r1 se to the most outstand1 ng gathered-crop export economy ever to flounsh 1n West Afnca unt1l plantat1ons were estab 11 shed. Large areas of otherw1 se 1nhosp1 table coast 1 n the Malay Arch1pelago became hab1table because of the sagopalm. Some areas were product1ve enough to trade 1n a gra1n made from 1ts flour, landang, the only notable food of vegetable ong1n 1n the precolonial commerce of the Onent. The Sulu traders supphed "sago of the best k1nd" from a terntory that stretched from northeast Borneo to M1ndanao (Moor 1837). The sago comp 1ex mam fests an 1mpress 1 ve spectrum of techn1cal sk1lls requ1red to fell large trees and m1ll the1r p1thy stems for a staple food that belles 1ts 1mage of cultural pr1m1t1veness. Sago is the meal w1th1n the stem wh1ch e1ther 1s pounded and s 1fted or 1s 1eached vH knead1 ng or tramp 11 ng to y1eld the starch wh1ch 1s made 1nto flour, pearl, b1scu1t and bread; these may be nutr1t1vely enhanced v1a a preparatory fer111entat10n or fort1f1ed w1th meat, r1ce bran, nuts, coconut, etc. Toasted bread, rot 1 papua or sagu maruku, and other dry products keep for years. These once vi ctua 11 ed sh1 ps of the reg1on, were even fed to the Papuan slave cargo, antedat1ng the s1m1lar role of the tap1oca 1n the Negro trade of the New World. So lntngued was Wallace (1898) by the sago food technology 1n the easterly Indones1an 1slands, that he d1gressed from h1s ep1c zoolog1cal pursu1ts there to descnbe 1n cur10us deta1l the art of sago breadmaking in east Seram 1n 1860. The starchy d1et 1s enr1ched w1th other swamp resources, e.g. f1sh, crustacea and w1ld p1g, and maggoty palms prov1de a l1v1ng larder of fat when slashed to encourage a weev1l to lay 1n rott1ng p1th, the "m1crocow" of th1s abong1nal d1et (Stanton 1972). Sugar and w1ne come from the n1pa, thr1v1ng 1n more brack1sh fr1nges. Sago 1s st11l the staple of many pre-Malay populat1ons 1n the huge eastern swamps of the reg1on. M1ndanao, Maluku, Su l awes 1, Borneo and Sumatra conta1 n dwi ndl1 ng numbers of sago eaters; Borneo was long famed for 1ts sago, hence 1ts nat1ve appellat1on, "Pulau Kal1mantan" or "1sland of raw sago". Bas1ca11y an eff1c1ent food acqu1s1tion system, the sago d1et has been found to be super10r to the cereal on occas1ons. In New Gu1nea "populat1on dens1ty ••• per square m1le of 100 to 250 1s largely due to a great rel1ance on sago and/or f1sh1ng for subslstence. In the hllls and mounta1ns there 1s a more general 4

tendency for h1gh populat1on dens1t1es to be assoc1ated w1th 1ntens1ve use of land for cult1vat10n; nevertheless, even here • • • (they) are partly susta1 ned by ••• sago" (HaantJens et al. 1972); the tnbal lands 1n the Mapnk area exceed1ng 400 persons per sq m1le are the most densely settled 1n lowland New Guinea (HaantJens et al. 1968).

Ecology The sagopa l m occup1 es a hab1 tat c 1ose to but di st 1net from that of the coconut: "In most parts of the Ind1an Arch1pelago two k1nds of all uv1 a 1 so1l are found 1n greater or 1ess abundance, one cons 1st 1 ng ch 1efly of sand often thrown up 10 long banks, and the other ch1efly of decomposed vegetable matter.... For these two descnpt1ons of so1l, nature has prov1ded two k1nds of palm adapted 1n a wonderful manner to the necess1t1es of man. On the barren sand she has planted the coconut, and 1n the morass the sago tree.... Low marshy s1tuat1ons shut out, but at no great d1stance from the sea, and well watered by fresh-water seem most productive" (Logan 1849). In the ASEAN reg1on extens1ve tracts of freshwater swamp dom1nate the eastern coast of Sumatra, the north and south coasts of Borneo, and the larger alluv1al bas1ns of New Gu1nea (Plate B), w1th smaller swamps spott1ng the coastal fr1nge of many 1slands of the arch1pelago; 1n favoured s1tes the1r exceptional demzen 1s the sagopalm. Europeans who coasted by th1s env1ronment frequently found 1t unprepossess1ng, even host1le: "The sago and mangrove swamp areas are 1n1 m1 cal to human 11 fe ••• furmshed w1th cond1t10ns wh1ch to European l1fe would be most 1nhosp1table, 1f not almost altogether 1nsufferable ... unf1t for hab1tat1on [with] a tendency to depress the p1oneer1ng enterpn ser and to retard sett 1ement" (Thomson 1892). For all 1ts pest1lent a1r, th1s umque equator1al habitat, "dry land • • • 1 n the wettest sense • • • The Land Beh1 nd the Sea", has had 1nd1spensable value to 1ts 1nhab1tants s1nce preh1stoncal t1mes. "The flora of the wetlands, the swamp, 1s far more 1mportant 1n almost every way than that of the hills1de undergrowth of the ra1n forest.... As a source of produce requ1nng no kind of care, the swamp 1s a mud-mine of last1ng mer1t" (Harr1sson 1970); also "w1th the r1vers and the easy product1 veness of the sago palm an alternatl\le of large, concentrated commun1t1es was added" (Maher 1961). Thus, the 5

PLATE B

A

Char~erls~lc

Floodplain

Habl~a~

Alluvial floodplain of the lower Ramu River, Northeast Territory of New Guinea, about 45 m I I es from the estuary, Is character I sed by scro I Is, I evees, oxbows and A backp I a Ins; It forms a contIguous swampforest with the Sep I k further west. In various stages of sagopalm and herbaceous swamplands, forest, mosaic of sl ltlng up In the cutoff meanders, reflects the Influence of the seasonal flood Sagopalms are abundant as the understorey In swamp woodland, grading to cycle. the cl !max vegetation In the waterlogged water without potential agricultural

sites. control.

This

region

has

low

courtesy

of

In general

(I I I ustration

Dr K. Paljmans-CSIRO, Canberra)

coastal lowlands, up to 2,500 ft in the islands west of Sumatra eastwards to Fiji at least, frequently harboured sedentary, ostensibly nonagricultural people who developed a food technology for the sagopa 1m that para 11 e 1ed in essence and sophistication that of the Amerindians who domesticated for their staple food a M~nihot cassava), (manioc, tapioca the tuber, poisonous esculenta. The sagopalm grows by riparian banks, "all the way down to 6

the mouth of the river" (Oosterwal 1961}, on lake margins, in "creeks and gullies" (Thomson 1892}, "inland forests" (Dilmy 1971), and "damp cuts in the hillsides" (Burridge 1969), "where it seems to thrive equally well as when exposed to the influx of salt or brackish water kept constantly full of moisture by the rains, and by the abundance of rills which trickle down among them" (Wallace 1898}. This early perception on its salt tolerance in the Seram swamps has been affirmed experimentally (Flach et al. 1977}. The sagopalm's ecology recalls that of the wild oilpalm in its own hearth where it naturalised also near saline sites (Zeven 1965; Adejuwon 1970}. A good sago plantation or forest is "a bog knee-deep" (Crawfurd 1820}, in "low marshy situations shut out, but at no great distance from the sea, and well watered by freshwater seem most productive" (Logan 1849). The sagopalm thrives within the environment characterised by silt depositions and a flowing and ebbing water-tab 1e which aerates and stimulates the root systems of swamp vegetation. Even where fastidiously cultivated, manuring is nowhere practised, and would be superfluous within this Nile-like environment, which succinctly illustrates the dictum: "Feed the soil life and it will feed the crops" (Vail 1979}. Systematic land use studies in New Guinea (CSIRO) reveal a cha racteri st i c sequence of sagopa 1m adaptation: Extensive mixing of fresh and brackish waters in a floodplain creates a fairly uniform type of swamp in which sago initially occurs as small scattered circular groves in clearings beneath a forest canopy 70-140 ft high, grading in the wetter sites to a dominant understorey at a height of 40-50 ft, with fronds up to 30 ft long, and the occasional emergence of a flowering spike up to 65 ft (Plate C). Where the drainage is impeded, the tree canopy becomes more open or 1ower. Progressively, sago forms the c 1i max vegetation camp lex with virtually no undergrowth, the only ground cover consisting of its abundant pneumatophores, which project above the regular water table. However, wild palmeries are too poor for convnercial exploitation. A natural palmery becomes useful only "where it is accessible and vigorous enough to come into flower .. • probably the sagopalms growing under forest conditions are usually not Sago production ••• vigorous enough to warrant exploitation. could be considerably increased by the clearing of forest trees, planting, and proper spacing of good-quality sago palms" Edaphic and hydrological (Haantjens et al. 1968, 1972). conditions may also influence the density of palms that flower. In swamps that are 1i ab 1e to season a 1 drought, severe brackish inundation or are deficient in nutrition or aeration, such as waterlogged soils, the palms form a dense cover but few will grow 7

PLATE C

Sligo s-ptores"t Oaaplimperfectly to poorly drained Alluv i al plain In Gulf District (south Papua) subject to wet-season flooding: l and with mi d-height (70-100 ttl, moderately closed mixed forest Interspersed with sago in swampy patches. (Illustration Some sagopalms are In var i ous reproductive stages, particularly evident In right foreground. courtesy of Dr K. Paijmans-CSIRO, Canberra)

to more than 20 ft or flower; instead they vegetate into vigorous stoloniferous clumps. The freshwater swamp is one of the most underrated daerah minus ("deprived region": Fox 1977) in the archipelago;~ popular conception ••• is of vast sago swamps which offer no opportunities for economic development" (Ryan 1972). Yet where there are pure stands of non-timber resources such as the sagopalm, the swampforest has a value equal to the lowland rainforest (Whitmore 1975). "The huge Metroxylon forests of New Guinea represent enormous potential resources in starch for export and industrial uses and deserve new research.... The living conditions of the swamp forest peoples could be improved by an industrialisation and marketing of sago ••• as a source of revenue in these poverty-stricken areas" (Barrau 1959). Logan (1849) made this remarkable assessment of the sagoswamps of insular Southeast Asia: "the Archipelago can furnish any required amount of meal". Hijandi et al. (Stanton and F1ach 1980) estimated that there are about a million ha of exploitable sagopalms in Maluku and Irian Jaya; there may even be 1.5 million ha in Irian Jaya alone (Sinar Harapan 1979). Of unknown qua 1ity is another substantia 1 reserve in Sumatra which still has commercial production in Riau. The Indonesian Minister of Research & Technology made "an appeal to all sago producing provinces in the country not to damage sagopa 1m trees or p 1ant at ions but instead to increase the number" (The Indonesia Times 1979), in recognition of the value of the sagoswamp as a supplementary granary for the country. In 1979, Indonesia had to subsidise the import of 1.5 million tonnes of rice, even more in other years. Papua New Guinea has over a million ha, of which 300,000 ha are accessible and harvestable (Newcombe et al. 1980); Essai (1961} singled out the swamp 2,400 sq mile in extent between the Huntschein Range and the Sepik estuary as a potential sago centre.

Pal• Biology The Metrox~lon spp encompass the more robust members of the Lepidocaryo1deae (scaly-fruited) subfamily, characterised (except for one species in Micronesia) by hapaxanthy,1 a predilection for particularly humid homes, and a prominent relict trait of palm ancestry, spinosity, manifest in a spiny lower stem, spiny spathes and combs of spines along the frond base in the wild sago. The lepidocaryoid palm is also marked by its loricate fruit, covered with imbricating scales; Barrau (1959) distinguished between the fruits of 18 vertical series of scales

9

1n the New Gu1nea centre and of 24-28 further east. Taxonom1c stud1es of the genus are st1ll 1nadequate and there 1s no comprehens1ve 1dent1f1cat1on of all members yet; about seven spec1es are recogmsed. In the 19th century confus1 on dogged the taxonomy of even the cult 1 vated spec 1es, a problem that commonly bedev1lled the study of the spec1es-nch trop1cs. The sagopalm bore other gener1c and spec1f1c names, such as Sa~us l aev1 s. The consensus 1s that M. sagu 1 s probably not a dist1nct spec1es, but an ant1thet1c phenotype or hybr1d of its w1ld kin, because both spec1es cross read1ly, g1v1ng nse to po lymorphl c seedll ngs rang1 ng between the two extreme forms. A proport 1on of sp1 no us or berdur1 progeny 1 s 11 able to result even from nonsp1nous crosses. Open to quest 10n 1 s the degree of "w1l dness" of the westerly palms, although improvement of the two useful spec1es under abor1 ginal agronomy for several m1 llenma has been ent1 rely emp1r1cal. The select1on of smooth forms to prov1de bu1ld1ng mater1als would have promoted the evolut1on of M. sa~u even 1n areas where it was not a foodcrop. Precolon1al 1ntra1sland settlement and trade fac 111 tated the westward d1 spersa 1 of the spec1es, e.g. by Malay traders return1ng to the1r homeland 1n Sumatra, where the smooth palm was to flour1sh 1n an env1ronment 1n wh1ch the pig was suppressed under Arab-Ind1an influence, even while the w1ld palm, w1th 1ts fearsome armature to promote 1ts surv1val 1n cultures in wh1ch th1s pest was pr1zed as food, cont1nued to flour1sh in its hearth. The modif1ed Wallace L1ne (west 1nstead of east of Sulawesi) more or less b1 furcates the d1stribut1on of the two spec1es. The pers1stent, remnant concentrat1on 1n th1s part of the Pac1flc Ocean of th1s 1nsular genus of palms with their uncharacter1 st1 c flower1ng propens1ty has begu1l1ng Atlantean overtones {Whitmore 1973). In Sumatra, Malaya and Sulawes1 the palm is better known as rumb1a, and further east as labia, or their cognates. Largely on tfi"eDas is of the spinous var1 at 10ns, numerous vernacular names arose throughout 1ts geograph1cal span, d1st1nct from those used for other starch palms, wh1ch is regarded as ev1dence of 1ts ant1qu1ty as a foodcrop 1n the reg1on (Crawfurd 1820). The sagopalm 1s cons1dered to be 1ntr1ns1cally Wlld in 1ts genet1c const1tut1on because no sc1ent1f1c breeding has ever been undertaken. Dwarfer var1eties were developed as a more manageable crop by tr1bes where, because of taboos aga1nst male part1c1pat1on, women had to fell the palm; but generally var1etal evolut1on has been towards mass1ve forms that are more productlve and 1nc1dentally require a longer period of growth, a factor of no consequence under abor1g1nal economics of product1on but cr1t1cal for plantat1on explo1tation.

10

The sagopalm shares the root1ng character1stics of most palms. The dense advent1t1ous root system enables the palm to tap the fert1l1ty 1n the alluv1um; l1ke the 01lpalm 1t has breath1ng roots, pneumatophores, wh1ch also funct1on as s1lt traps. Below the frond bases root t1ps are v1s1ble, and a f1ne lacy mat of growths envelop the stem, cours1ng upwards; perhaps th1s 1s a supplementary feed1ng system wh1ch 1mbibes the hum1d1ty of the a1r and supports 1ts surv1val as a hygrophyte part1cularly where groundwater 1s brack1sh or 1nsuff1c1ent. The sagopalm 1s polycarp1c, v1z. 1t produces, even neotenously, stolons that spr1ng from ax1llary buds from the base to the lower stem. Some authors state that 1t suckers only after flower1ng or fru1t1ng. Regenerat1on "by rad1cal shoots, exactly 1n the same manner as the common cult1vated planta1ns, 1s pecul1ar and 1s not observed 1n the true palms" {Logan 1849}. Thus vegetat1ve propagat1on 1s commonly pract1sed under cult1vat10n although the palm seeds 1tself. Sago planters eschew seed1ng also because seedl1ngs are vanable, and palms left to frult tend to be runts that had been reJected for the1 r food worth, hence undes1rable parent matenal. The hab1t fac1litates perpetual clon1ng and leads to the concentrat1on of part1cular vanet1es 1n a spec1f1c area, g1ving rise to fa1rly homotyp1c groves. Where fer a 1 , 1 t vegetates into dense c 1umps 1 ike bamboos; as a weed or squatter, 1t 1s most at home on abandoned r1celand, s1nce the cereal occupies a s1milar hab1tat or has usurped 1ts natural n1che. Be1ng monocotyledonous, 1.e. without camb1al growth, the stem's g1rth and potent1al for growth are predetermined dur1ng 1nfancy 1n 1ts ap1cal menstem, wh1ch caps the "heart" or "cabbage" of the pa 1m; the stem e 1on gates only when 1ts fu 11 d1ameter 1s reached, so that environmental tnfluences dur1ng 1ts early years are cr1t1cal to 1ts future product1veness. "The wtdth of the trunk expresses the welfare of the palm apex" (Corner 1966}; any subsequent th1ckemng 1s due to the expansion of the parenchyma cells and vascular bundles 1n the p1th 1n response to 1mproved cond1t1ons of growth, espec1ally sunl1ght. The sagopalm 1s hel1oph1lous, for 1nsolat1on seems to have a marked effect on 1ts growth, and leaf format1on tnfluences starch accumulat1on dur1ng 1mmatunty (Flach 1980}. The f1rst palms of a new plant 1ng have um form g1 rth of stem, whereas "succeed1 ng pa 1ms mostly show a c 1ear 1 ncrease 1n g1 rth at the top of the trunk" (Flach 1977}; they are also taller (Morns 1953}. About two p1nnate-leaved fronds are produced a month, borne 1n a one-th1rd phyllotax1s, a sp1ral fol1age sequence where the fourth frond 1s emplaced above the f1rst, etc. that 1s character1st1c of the spec1es.

11

Economic value of palm florescence Palms have a relat1vely long vegetatlVe phase of 1nfant1le growth to form the stem and store the food reserves for reproduct10n. In the hapaxanth1c palms, where 1mmatur1ty may endure for 20 years 1 n the sagopa lm and over 40 1n the Corypha, thl s 1mmense store of nour1shment 1s mob1l1sed and exhausted 1n a s1ngle fatal outburst, aptly descnbed as "the cap1tal1st1c nature of plants" (Corner 1966). It 1 s not known what factors cause the pa 1m to bolt, but photopenod1sm, or l1ght durat1on, rather than l1ght 1ntens 1ty may be the dec1 s 1ve 1nfl uence. The hapaxanth1 c tra 1t, comparat1vely rare 1n palms, may spnng from a term1nal ax1llary bud rather than the mer1stem 1tself (Corner 1966; th1s appears to be 1llustrated by Tupamahu 1909). Flower1ng is protandrous, 1.e. the pollen 1s shed before the st1gma 1s recept1ve. The Papuans occas1onally d1sbud the palm shortly before flowering. Abort1ng its development at th1s cr1t1cal moment prevents bolt1ng and probably prolongs starch accumul at 1on or conserves the starch reserves, because a hapaxanth1c palm left to fru1t has no value as a food source. Where used for food then, the sagopalm 1s unable to seed 1tself. Some cult1vars develop 1nto cult1gens probably because vegecu 1ture depresses the1 r ab1l1 ty to flower or produce fert 11 e seed (Crawfurd 1820; Burk1ll 1935). Low (1848) and Burkill noted th1s 1nfert1l1ty 1n some westerly cult1vars 1n part1cular. Although 1n 1ts phys1ology and habltat sago hardly resembles the tap1oca, 1t shares two outstand1ng features w1th th1s other hardy foodcrop. The f1rst 1s "the capac1ty to y1eld a crop of sorts even under the poorest cond1t1ons ••• and be left 1n the ground for cons1derable per1ods w1thout ser1ous loss and harvested whenever needed" (Coursey and Haynes 1970). Second, the useful part lS produced w1thout the operat1on of the sexual process and often before the plant 1s sexually mature, 1n contrast to the cereals, w1th 1ts "structural eng1neenng" contr1buting to 1ts h1gh b1olog1cal eff1c1ency. The sagopalm's actual product1on of dry metter 1s h1gher than the potent1al product1v1ty of r1ce (Flach 1977), v1z. the prim1 t lVe pa 1m 1 s a 1 ready super1 or 1n b1 omass product 1on to the overbred cereal. The biomass 1s translocated dunng 1mmatur1ty to 1ts 1mmense photosynthetlc mach1nery of fol1age. When the vegetat1 ve phase peters out, there 1 s a reverse flow towards the stem, the "mere shell" (Crawfurd 1856) wh1ch 1t reverts to after fru1t1ng. The stem has cons1derable structural res1l1ence and an enormous starch-stor1ng capac1ty, wh1ch may be l1kened to the puny organs of dormancy of the aro1 ds and tubers. Protected by a half-1nch th1ck, water and pest res1stant cortex, the sagopalm 1s vutually 1ndestructible after surv1v1ng 1nfancy, even by f1re,

12

except for weev1l damage to soft crown t1ssue. From the ev1dence on tapp1ng pleonanth1c palms (w1th success 1 ve ax1ll ary fl ower1 ng), especially ~renga, Borassus and Caryota spp, starch content 1n the hapaxant 1c palm lS presumed to be at 1ts peak wh1le the 1nflorescence 1s form1ng but before 1ts resources are d1ss1pated 1n reproduct1on. The emp1rical 1ntu1t1on of 1nd1genous cultures that explo1t carbohydrate-r1ch palms for food and dr1nk 1n 1nsular Southeast As1a 1s ev1dent 1n the common pract1ce of harvest1ng them at the cr1t1cal phase when vegetat1ve growth g1ves way to the reproduct1ve. However, the prec1se stage when starch accumulat1on 1s at 1ts hlghest or the earl1est stage when 1t 1s econom1cal to harvest the palm 1s not known. The per10d of 1mmatu r1 ty 1s as short as s 1 x years for seedl1ng palms (Cantley 1886) and seven years for the w1ld, smaller palm (Forrest 1780), but most modern wr1ters are 1ncl1ned to attr1bute a longer 1mmatur1ty to 1t. The general pract1ce 1s to harvest between 1nc1p1ent emergence of the 1nflorescence spad1x and 1ts 1mm1nent fru1t1ng, wh1ch co1nc1de w1th two apparent peaks 1n starch content (Flach 1977; Wee: Tan 1977). The general bel1ef among planters 1s that the domest1cated sagopalm averages 9-10 years of growth on more fert1le so1ls before 1t can be felled for 1ts starch. Matur1ty 1s JUdged 1n New Gu1nea at 1nc1p1ent to full flowenng, in Sarawak from early florescence to early fructescence, and 1n Batu Pahat even earl1er. Impend1ng matur1ty 1s heralded by "a mealy appearance" on the frond bases, wh1ch 1s "the earl1est of s1x stages 1n the progress of the matun ty of the medullary matter" recognised by the Moluccans (Crawfurd 1820); other features are displacement of the fronds from the1r character1st1c phyllotact1c arrangement, reduced frond growth, fall owed by taper1 ng of the crown. The qual1ty of the starch may vary at the d1fferent stages of matunty. The sagopalm probably saturates 1ts pith w1th starch from the base of the stem upwards (Flach 1977), though 1ts 1owest and uppermost segments are never the starch1est; the starch at the base 1s presumed to be d1verted for the nour1shment of 1ts progeny. Understand1ng the process of carbohydrate format1on would a1d 1n determ1n1ng the earl1est useful harvest1ng t1me before fullterm b1olog1cal matur1ty, as the sagopalm's long 1mmatur1ty 1s regarded as the solltary, overwhelm1ng 1mped1ment to 1ts extens1ve culture. The ster1le cultivars y1eld 400 kg of flour at 35-40% mo1sture (Barrau 1959), wh1ch 1s h1gher than most estimates, rang1ng from 152 kg (Forrest 1780) to 408 kg (Wallace 1898). The best of the New Gu1nea cult1vars are sa1d to be more mass1ve than those found 1n Malaya (T. Power 1981, Angoram, pers. comm.), as 13

are also the Sarawak1an (Flach 1977); 1n Ir1an Jaya 30 tumang of wet sago (30 kg each) can be obta1ned per palm (S1nar Harapan 1979). A well-tended sagopalmery could annually y1eld 22 tonnes/ha of flour, equ1valent to 35 m1ll1on Calone, w1th less agronom1c 1nput per umt volume and area than any other staple foodcropping system. The sagopalm was found to be at least 10 t1mes more product1ve a food suppl1er (after cons1der1ng an 1mmatunty phase of 15 years) than wheat or potato before the temperate countr1 es embarked on the genet 1c and agronom1 c 1mprovement of the1r staple foodcrops (Logan 1849).

NOTE Hapaxanthy

(once-flowering) and pleonanthy (multiple florescence) are used here as by palm botanist, J. Dransfield (Tan 1977), and are not synonymous respectively with monocarpy (sol ltary, non-clumping) and polycarpy (clustered),

14

I

THE SWAMP-SAGO INDUSTRY IN WEST MALAYSIA

Sago in the Straits of Malacca In the southwest Malayan adm1n1strat1on observed:

pen1nsula,

the

early

Br1t1sh

"The occupat 10n of the 1owly1 ng 1and on the coast and nver banks 1s preh1stonc. The wnter of the Malay Annals states that 1n the days of Sultan Mahmud a man m1 ght pass from Ma 1acca to Batu Pahat w1 thout a torch, wherever he should stop, there would be houses and men" (Johore, Ann. Rep. 1911). Yet 1n 1856 Crawfurd wrote thus of the southernmost state: "The large an1mals of the forest are probably far more Earll er Moor ( 1837) noted numerous than 1ts human 1nhab1 tants". "almost umnterrupted mass of JUngle and swamp ••• except for a few straggllng v1llages and cleanngs" 1n the Muar area 1n the northwest border of the State, w1 th few trad 1ng boats and fi sh1 ng canoes, although accord1 ng to 1oca 1 trad 1t 10n there had earl1 er been large and populous towns, "cult1vated r1ce grounds have degenerated 1nto barren marshes" prol1f1c 1n wlldlife. Th1s state of the hab1tat10n, of what was to become "w1thout doubt the most r1ch settlement 1n nat1ve agr1culture 1n the Malay pemnsula" by 1914 (Johore, Ann. Rep. 1915), had been caused by local ch1efs who exacted gratu1tous labour from the1r subJects and by pirates whose depredat1ons pract1cally depopulated the coasts (partly aggravated by self1sh Dutch mercantile pol1cy). Both bl1ghted the once-thr1v1ng forest-collect1ng economy. It 1s Logan's percept1ve observat1on of 1849 about th1s reg1on during 1ts doldrums that 1s part1cularly sign1f1cant to my study: "Along the

Sumatra coast

from S1ak to the

15

Lampongs,

and 1n the large pla1ns of R1o Formosa and the Muar, there are hundreds of m1les of sago land unoccup1ed and unproduct1ve, every acre of wh1ch 1s capable of y1 el d1 ng at the rate of about twenty thousand pounds of meal yearly." He added, "Malays all agree that the cult1vat1on of sago 1s the most prof1table of agncultural pursults, not y1eld1ng even to the cult1vat1on of nee by Sawas, for once 1n beanng the trees are ad 1nf1n1tum equally prof1table and requ1re l1ttle or no labour". These palmer1es 1n the Batu Pahat-Muar area were commerc1ally useful, because sago 1n "bundles" was an early export to S1 ngapore, a 1though the greater vo 1ume of export was 1n forest products and, later, 1n cult1vated produce. Newbold (1837) added: "the trade 1s earned on almost ent1rely by nat1ve craft from one to e1ght coyans burthen, and small open boats called sampan Pucats" (coyan = 2.4 tonnes). The Malays of southwest Malaya of Logan's t1me d1d not seem to use the palm for the1r staple food, prefernng 1mported nee when they could purchase 1t, and probably they d1 d not estab 11 sh the palmenes he saw there. But the earl1er p1oneers probably d1d use the palmer1es as d1d the Malaccans the1rs, for as Wheatley (1964) observed: "Despite the wealth and 1mportance of Melaka, the 1mmed1ate h1nterland of the port appears to have been very l1ttle developed. At the t1me of Tome P1res' v1s1t, 1t was clothed w1th an almost un1nterrupted mantle of forest, d1vers1f1ed only occas10nally by an 1solated kampong. Muar and Batu Pahat were small farm1ng commun1t1es; S1ngapore was pract1cally deserted and of l1ttle 1mportance. Later arr1vals scaveng1ng along the coast had taken up the1r preferred s1te, close aga1nst the water's edge. Few of the 1mm1grants showed any 1ncl1nat1on to undertake cult1vat1on of the so1l. Indeed, farm1ng was but poorly developed throughout the south of the pen1nsula at th1s t1me. Furthermore, the alluv1al so1ls adJacent to the town were probably st11l too sal1ne for pad1-farm1ng. In any case, sago, not nee, was the staple food of these early colomsts, supplemented by qu1ck-grow1ng crops." In the East coast cultures "r1ver sago" was already cult1vated as the staple 1n the early Chnst1an era (Tregonn1ng 1964), and the Kelantan nvenne 1nhab1tants cont1m:e to nurture slulls 1n mak1ng a var1ety of foods based on sago. It 1s not known when or by whom the ea rl1 est sa gop a 1ms were 16

d1 ssem1 nated to the Batu Pahat bas 1n, but the r1 se of S1 ngapore probably st1mulated the extens1on of sago cult1Vat10n from the ne1ghbour1ng old produc1ng centres, part1cularly Melaka and sumatra, wh1ch were among her pnnc1pal suppl1ers of raw sago. R1dley (1905) thought they were "probably 1ntroduced very early 1n Malacca"; he found them "scattered all over the Pen1nsula, and S1nce 1t 1s generally form an 1mportant art1cle of export". accepted that the sagopalm 1s not 1nd1genous to the westerly parts of the Malay Archlpelago, nor 1s 1t equ1pped to be an 1sland-hopper, as 1ts ab1l1ty to seed 1tself 1n these fragmented areas to form the 1mpress1ve groves 1s 1n doubt, these palmenes cou 1d not have been w1l d stands. Thl s assumpt 10n 1s supported by the fact that the sp1 no us pa 1m 1s rare 1n th1 s reg1 on, whereas even on the eastern coast of Sumatra sp1 ny forms grew a 1ongs 1de the smooth sago; both are absent from the natural flora of the pandan-r1ch swampforest 1n south Johor (Corner 1978). Another poss1ble area of 1nfluence 1s Sulawes1. Rumb1a 1s a Maccarase term for the palm that 1s used also 1n southwest Malaya up to Negr1 Semb1l an, and 1n Sumatra; however, these areas also have M1nangkabau 1nfluence. The Bug1s used to organ1se substant1al perahu trad1ng fleets to Melaka, and later to S1ngapore (each perahu capable of carry1ng 2,000 tamp1ng of a p1 kul or 60 kg each). They congregated 1 n WaJu 1 n Su 1awes 1 to collect a vanety of produce wh1ch 1ncluded sago from Maluku (the Sp1ce Islands) and ne1ghbour1ng coasts, gathered by the abor1 g1 na 1 popul at 1ons ( Crawfurd 1820, 1856; Cameron 1865; Treacher 1889); R1au was the1r ch1ef haven 1n the Stra1ts of Maiacca. Thus Bug1s settlers on the southwest Malayan coast could have started or encouraged some of the commerc1al palmer1es to ensure a dependable food source for themselves 1n lean t1mes. Once core plant1ngs were establ1shed, the palm read1ly spread across the natural landscape; Banng-Gould and Bampfylde (1909) also had observed 1n Sarawak "Jungles of the cult1vated palm where f1fty years ago there were patchy plantat1ons". Ma 1ay traders monopo l1 sed the 1 ucrat lVe sago trade 1 n the ent1re arch1pelago, but 1n the western parts th1s had become slugg1sh unt1l Slngapore st1mulated it, part1cularly from the early 19th century. Earll er the Malay traders were afra1 d to cross the Stra 1ts frequently because of p1 rates, or to venture deep 1nto the marshes 1n east Sumatra to collect sago from the abor1 g1 nes. Furthermore, the chief of S1 ak was reputed to be rapac1 ous to the pol nt of deterr1 ng his subJects from earm ng a l1velihood under h1m or work1ng too hard for h1m by produc1ng the sago he sought. W1 th the stab 11 sat 1on of the col om a 1 spheres of explo1tat1on by the md-19th century, p1racy was brought under control, and the sago trade from the southwest Malayan coast resumed. Gnfflth (1850) recorded 1ts extens1ve cult1Vat1on around Me 1aka shortly before tap1 oca dom1 nated the economy of the 17

port and m1n1ng centre from the m1ddle of the 19th century. The centre of commerc1al sago trad1ng was S1ngapore (F1gure 1), because the unfavourable taxat1on pol1c1es of the Dutch 1n the1r own ports drove entrepreneurs to a less restnct1ve base. Wlth the decl1ne of Melaka around 1820, the Br1t1sh had planned to develop R1au 1sland as the1r pnnc1pal port to secure a "fa1r" part1c1pat1on 1n the lucrat1ve trade of the arch1pelago, from wh 1ch they had been a 1most comp 1ete ly exc 1uded; but the Dutch forestalled the1r amb1t10n, and the Bnt1sh turned to S1ngapore 1sland. Sl ngapore' s trade orb1 t stretched from Tennassen m, through the Federated Malay States and Straits Settlements (along much of the west coast of the pen1 nsul a of Ma 1aya), and across the Dutch

Q

25

50

102 f

IOOE

104

~~-~m1le

~

e

major freshwaTer

=-~ swamps

~ S 0 U T H

MA L A Y A

C H I N A S E A 2 •

FIGURE 1

MaJor Freshlfllfar S-ps of Mlllaya and S-1T11 In 1964. and Sago Ports 1n 'the Ear I y 20'th Can'tury

18

terr1 torl es up to Ma 1uku. The produce of the sagopa l m from the ent1re arch1pelago supported the only notable manufactunng 1 ndustry of Swgapore 1n the 19th century, v1z. be1ng reprocessed 1 nstead of merely graded !1 ke other produce. Java too had sago factones, but Arenga P1nnata was apparently the palm used there by that t1me, hav1ng superseded swamp sago. The wet flour was shl pped to the port 1n leafy bags after bel ng scraped out and washed, for further clean1ng and dry1ng wto starch or pearl for export to Europe, Ch1na and India where 1t was used 1n the paper and text1le 1ndustr1es (Seeman 1856; Kn1ght 1969). Even "one or two square- r1 gged vessels arn ved annually from Arab1 a • • • tak 1 ng back ••• a small quantity of sago" (Newbold 1839). The Ch1nese manufacturers 1n S1ngapore were supposed to have been the f1rst to adopt commerc1ally 1n 1819 the method of pearl1ng "wh1ch has done so much to render 1t an art1cle of consumpt1on.... In S1ngapore all sago 1s of the1r manufacture • • • and 1 n Mal acca • • • all the sago 1s of the1 r product 10n" (Cameron 1865). Th1s pearl1ng techn1que was apparently un1que to the westerly parts 1n contrast to the bread mak1ng of the eastern 1slands, and was sa1d to have been denved from Buk1t Sau on the S1ak coast.2 Unt1l S1ngapore used 1t commerc1ally, 1t was sa1d to have been kept a secret (Moor 1837; Logan 1849); 1t was 1mproved at the same t1me by the Melaka Ch1nese to g1ve sago "a f1ne pearly lustre" (Crawfurd 1820), wh1ch had thnce the value of ord1nary sago. The Melaka tap1oca processors later usurped 1t to produce an 1dent1cal product, for wh1ch thereafter "sago" became synonymous. By the early 19th century S1 ngapore manufacturers recogn1 sed three grades of sago: the f1 nest came from S1 ak, next was the Bornean product, and the least des1red was the Moluccan brought over by the Bug1s (Crawfurd 1820). Wh1le Slak itself apparently produced none, the entue Bengkal1s-L1ngga reg1on was noted for 1ts commerc1al sago; the S1ak hwterland was the deepest because the r1ver was nav1gable for 80 m1les 1nland for vessels of over 200 tons and 150 m1les for smaller boats. S1ak traders who operated from that entrepot dun ng its preem1 nence 1mported the f1nest sago from across the Stra1ts v1a Slak and R1au, which were collect1ng centres of substant1al quant1t1es of sago as well as the abundant produce of the vast Siak-Indrag1r1 h1nterlands. R1au's sago was also U$ed to adulterate gamb1er to make 1t heav1er (Newbold 1839). Rantau conta1ned the most extens1ve "plantat1ons" (Logan 1849); Anderson (1826) 1n h1s reconna1ssance of east Sumatra regarded "Rant ow 1 s 1and • • • a great p1ace for sago". The Ma 1ays who owned palmenes there organ1sed abor1g1nal serfs, and employed methods and 1mplements st1ll used by the Papuans, to produce sago for export on a sharecropp1ng bas1s, or they 19

purchased the surplus off abor1g1nes who worked the palmer1es for the1r staple food. Unl1ke the coastal culture, the 1nland tr1bes were "w1ld" for they st11l foraged for the1r 11Vlng 1n the almost 1mpenetrable sago swamps. Johor had also been part of the vo l at 11 e Me 1aka-Sumatran emp1 res and by the early decades of the 17th century 1ts west coast was sa1d to produce a l1ttle sago and r1 ce. Groeneve 1dt (1877) ment 1 oned Ch1 nese references to sago 1 n Johor 1n the M1ng era (1368-1644), as an art1cle of trade. The Ma 1ays 1an-S1 ngapore sago 1ndust ry may be regarded as an extens10n of Ch1nese enterpr1se w1th1n the Br1t1sh sphere 1nto a fleld of act1v1ty that had earl1er been dom1nated by the marit1me Indones1an traders and collectors of produce 1n the former Dutch 1slands. The 1ndustry was encouraged to ensure Br1t1sh control over the bulk of the supply wh1ch the Singapore ref1ner1es requ1 red. The 1933-34 S1 ngapore Comm1 ss 1on descr1 bed the sago trade of the port as a su rv1Va 1 of the o 1 d barter system wh1 ch developed, as the enterpr1se became more commerc1al1sed, 1nto one of "barter w1th the pnces flxed". The S1ngapore f1nanc1ers, l1ke the1r counterparts 1n eastern Indones1a, had a l1en on the produce collected from growers, for 1n the1r system of control of pnmary product10n 1n the archlpelago they gave advances of cash and prov1s1ons to outport m1ddlemen who 1n turn were bound to sell to them only. The S1ngapore Ch1nese dom1nated 1n sago because the Europeans d1d not l1ke low-prof1t produce, preferr1ng to earn comm1ss1on on fre1ght 1nstead. Although the Br1t1sh adm1n1strat1ons frequently condemned the Ch1nese hold on native product1on 1n terms suggest1ng a well-organ1sed consp1racy, they barely remembered that European dealers had refused to take the phys1cal or f1nanc1al r1sks 1nvolved and had left 1t to the As1an m1ddlemen to forage around the arch1pelago to f1ll the1r sh1ps w1th worthwhile cargo. Instead, the Europeans exerted a v1rtual stranglehold on the trade by monopol1s1ng ocean-sh1pp1ng. Hardly an As1an dealer could sell goods outs1de the reg1on w1thout going through a Th1s was the case with sago, where only one European agency. Ch1 nese exporter managed to emerge. The Ch1 nese dea 1ers took the opportum ty to v1 nd1 cate the1 r unenv1 ab 1e pas 1t 1on to the Comm1ss1on. The Ch1nese owned most of the smaller steamsh1ps wh1 ch p11 ed the coast of Johor, the east coast of Sumatra and the R1au-L1ngga arch1pelago, but the two largest sh1pp1ng compames were European - the St ralts Steamsh1 p and the KPM ( Kom nk 11 J ke Paketvaart Maatschapp1 J) wh1 ch creamed off the trade at exorbltant prof1 t. The1 r vesse 1s a 1 so appropr1 a ted much of the Bug1s trade, relegat1ng the pera~~ fleets to the status of feeder vessels (D1ck 1975), the only 1n 1genous compet1t1on to endure 1n th1s reg1on's trade.

20

commercial Agriculture 1n Johor compared to the rest of Malaya, Johor came under the 1nfluence of an occ1dental system of commerc1al agnculture rather late, for the Bnt1sh became 1nvolved 1n the State only 1n the tw1l1ght of the 19th century. When they arr1 ved, they noted a 1 ready flounsh1ng areas of commerc1al agnculture on 1ts west coast, espec1ally fru1t and arecanut 1n the Sunge1 Muar bas1n geared for export ma1nly to S1ngapore and newly extend1ng down to Kukup, w1 th "sea rce ly an uncu l t 1Vated acre amongst them" (Ark 1 b Negara: GA 143/1918). Unt1l the r1se of Johor Bahru, the D1stncts of Muar (noted for the f1nest orchards, a Malay monopoly) and, from 1920, Batu Pahat produced the most substant 1a 1 revenues of a 11 the D1stncts 1n the Unfederated Malay State 1n the early years of the Bnt1sh adm1n1strat10n. The Batu Pahat-Kukup coast was among the most densely settled areas 1n the pen1nsula, covering a seventh of the state area. So 1mportant d1d the D1stnct become that the Brlt 1 sh Ass 1stant Adv1 ser was stat 1oned there to adm1 n1 ster 1t separately; unt1l the Causeway was bu1lt 1n 1924, the only mode of commun1cat1on w1th Johor Bahru was by boat. The Bnt1sh admlnistrators were del1ghted by the affluence surround1ng them 1n Batu Pahat, marvelling at the ex1stence of "an 1ll1m1table system of nat1ve agrlCulture far more extens1ve as regards nat1Ve plantat1ons, than anyth1ng one has seen anywhere 1n the Federated Malay States, even 1nclus1Ve of the Bagan Datoh d1str1ct 1n Perak" (Johore, Ann. Rep. 1915). The Ass1stant Adv1ser exulted, "Batu Pahat ••• llttle known to those outs1de Johore save the S1ngapore Bazaar where 1t has by no means a bad reputat10n ••• seems to be full of great poss1b1l1t1es" (GA 143/1918). The Johor rulers and the colomal adm1n1strat1on encouraged 1mm1grat10n from ne1ghbounng 1slands 1n the late 19th and early 20th centur1es to develop the State and to d1lute the Ch1nese dom1nance 1n commerc1al agr1culture earl1er encouraged to develop the pepper and gambler 1ndustr1es. The undes1rable penghulu system of land adm1 n1 strat 1 on was rep 1aced by the sunge1 surat to encourage a more sedentary and act1ve 1mm1grant populat1on. TyplCal of anc1ent Malay colomsat10n, seafarers had settled on the estuanne tracts, but few travelled much farther 1nland than they could 1n boats (1n contrast the 1nland area was developed by Ch1nese p1oneers and the1r European successors). Waves of 1mm1grants from Melaka, Muar, Sumatra, Java and Borneo crossed to the State, d1 sp 1ac1 ng the abor1 g1 nes who were absorbed 1nto the evolv1ng Malay culture or retreated 1nland; there was a h1atus dur1ng 1912-19, but expropnat1on of Japanese estates after World War One started another "locust swarm" of Javanese wh1ch endured unt1l the Great Depress10n (GA 602/1938). 21

But to the chagnn of the colon1al adm1n1strat10n, "they show the same tendency to become the servants of the Ch1 nese as they are over almost the ent1re coastl1ne of Negn Semb1lan" {GA 172/1915). Furthermore, 1t prov1ded no susta1ned market1ng or cred1t serv1ce to promote th1s v1tal part of commercldl farm1ng, and the Ma 1ay peasantry had to turn to Ch1 nese moneylenders and m1ddlemen to develop the1r land and market 1ts produce. These p1oneer farmers effected two s1gn1f1cant changes to the almost un1nhab1table wetlands. They establlshed ent1rely on the1r own an 1mpress1ve pal1t {dra1n) system wh1ch now ram1f1es the southwest coast of Ma aya from Muar {Bandar Maharan1) to Pont1an Kec1l for about 15 m1les wland, deeper 1nto the Sunge1 Batu Pahat bas1 n wh1 ch even then was st 111 the most flood-prone reg1on 1n southwest Johor {F1gure 2). The dra1nage in turn enabled the reg10n to be converted 1nto cult1vated f1elds of a vanety of crops, to form a coastal coconut fr1nge and a floodpla1n belt of mlXed cropp1ng soon dom1nated by rubber, and gambler-pepper 1nland. Produce such as p1neapples, sagologs and coconuts were strung together and floated downstream to market. Trad1ng boats penetrated past Tongkang Pechah, hence 1ts name "broken boat", ferry1ng farm and forest produce and ores to the or1g1nal coastal settlement of Bandar Penggaram, named after a stream, core of the present Batu Pahat town. By 1933, Tempany {1934} drew attent1on to the fact that Johor possessed a 11 crops of m1 nor 1mportance as we 11 as the largest acreages of the maJOr crops 1n the pen1nsula except r1ce. He noted that noth1 ng had been done for sago and lts "poss1bl e evolut10n 1nto a regular crop ... 1s a matter that m1ght well repay eluc1dat1on". {Older farmers 1n the outlying rural areas can recall the1r earl1er polycultural landscape; one descnbed coconut-r1ce-sago as "asal", 1.e. or1g1nal, trad1t1onal.) Desp1 te th1 s feat, that the Johor Ma 1ays were not except 10na l wet-nee cultlVators l1ke the1r k1n 1n Kedah, Kelantan and Negeri Semb1lan, was a source of tr1bulat10n to the Br1t1sh unt1l the Occupation. The var1ed Malay populat10n was sk1lled 1n var1ous farm1ng enterpnses (Tempany 1936}, and the establ1shed Malays preferred commerc1al arbor1culture to foodcrop farm1ng, l1ke the1r counterparts 1 n southwest Sarawak: "The r1 ce-dependent Ma 1ay has never regarded r1ce as a permanent occupat1on, but fits 1t prec1sely 1nto the pattern of p1oneer1ng JUngle back 1nto trees aga1n ... r1ce-plant1ng was nearly always cons1dered one stage 1n br1ng1ng ground under some more mass1ve and less energy-consum1ng method of cult1Vat1on" (Harnsson 1970). Even those of Javanese ori g1 n had the unusua 1 reputat 10n at the advent of co 10n1 a l rule of not hav1ng had a wet-nee trad1t1on (unl1ke 1n the Perak-Selangor coast where a strong Javanese element was l1nked 22

\

"•,

.•.

Stra~ts

\

FIGURE 2 Perl1" Syu- of Soutllwes1" Malaya

- par1t(dra1n - r1ver mma]or coastal Malay Reserves

of Malacca

,.,.

. ~

\

5

8

·~

IH..,

to successful nee plant1ng), preferr1ng where poss1ble or necessary "undes1rable" dry-nee (pad1 tugal) 1ntercropp1ng on the1r coconut and fru1t gardens, and purchas1ng the rest of the1r d1etary needs w1th the frults of other farm efforts. They took scant not1ce of the bonus pa1d for new sawah (GA 177/1939). Nor d1d these anak kebun (= peasant farmer: Johore, Ann. Rep. 1914) develop one even when offered 3-year rent- free sa wah and belaboured by success1ve adm1n1strat1ons (GA 172/1915; GA 177/1939; GA 2165/1940), who perce1ved an 1nexpl1cable preJUd1ce aga1nst the hard1er dry-r1ce, desp1te all the1r wet-r1ce prOJects fanng d1smally due to pest attack, flood1ng, the 1nexper1ence of the early r1 ce farmers, and not 1east to thelr own 1neffectua 1 manangement of an env1ronment where the rel1ef and cl1mat1c reg1 me were d1 fferent from those they had become used to 1 n the Federated Malay States. The State Dra1nage Board was not even establ1shed unt1l 1938 (GA 328/1939). The threat of taxat10n to stop the convers1on of sawah to palmery was to no ava1l too, for soon enough, large numbers of farmers who had persevered w1 th the1 r sawahs had to abandon them because they could no l anger cope w1th the lack of water control, part1cularly ev1dent dur1ng 1935-37. By then 1t was also real1sed that 1t was fool1sh to turn every farmer 1nto a food producer "whether 1t 1s economic or not" (GA 172/1915; GA 806/1939; GA 465/1940). In the1 r endeavour to nurture a homogeneous Malay peasantry among the 1mm1grants, the adm1n1strat1ons had discouraged Ch1nese part1c1pat1on (GA 02/1938), although it 1s qu1te doubtful that had 1t been otherw1se there would have been any 1mprovement 1n the state of wet-r1ce cult1vat1on so long as the farmers d1d not have an appropr1ate 1rr1gat1on system. It 1s most probable that the sa gop a l m under these c1rcumstances was relied on as a reserve staple food, for even dry-nee plant1ng was a speculat1ve act1v1ty, wax1ng and wan1ng 1nversely w1th cashcropp1ng. Dur1ng the Great Depress1on, forag1ng was a more noteworthy subst1tute l1vel1hood than r1ce-farm1ng. In World War One, the Johor adm1n1strat1on depended on both sago and tap1 oca to bu 11 d up food reserves, yet for a wh1le the Batu Pahat D1stnct managed to export monthly 180 tons sago bes 1 des 655 tons tap1 oca (AA BP 12/20). Johor 1 s sago contn but 10n to S1 ngapore was never s 1 gn1 f1 cant, as 1 ndeed was the case w1th the ent1re Malayan product1on because Malayan tap10ca and Sarawak sago were far more subsant 1a 1 (Append1 x II); a 1 ready 1 n 1906 S1 ngapore 1 s sago exports exceeded 50,000 tons (Malaya, Stra1ts Sett., 1907). under

Thus wh1le the peatlands of Pont1an were extens1vely dra1ned a susta1 ned off1 c1 a 1 programme to extend p1 neapp l e 24

cult 1 vatlon, and Benut for coconut, Batu Pahat became home to the sagopalm even wh1le the pre-ex1st1ng sago centres were on the decl1ne. Commerc1al product1on of sago certa1nly throve there by the end of the last century (Dennys 1894), or had been revlVed by then, 1n the Stra1ts Settlements, too, 1nterest 1n sago was apparent (Malaya, Stralts Sett. 1907) and there were plantat1ons ostens1bly 1n Selangor and Johor (Dennys 1894; Tregonn1ng 1964). Its trade was recorded fa1rly cont1nuously for a penod from 1911 when the Johor adm1n1strat1on started document1ng agr1cultural progress 1 n the State. Fa1 rweather and Yap (1937) cl a1med that the Batu Pahat sago 1ndustry was establ1shed 1n the clos1ng years of the 19th century by the Malays, and that 1t was usurped by the Ch 1nese on the dem1 se of the1 r gamb1 er-pepper trade 1 n Johor. Both R1dley (1905) and N1cholson (1921) fa1led to elaborate on sago here, though they were fam1l1ar w1th the Melaka 1ndustry. Sagu basah (equ1valent to the Melanau lementa), a wet flour after the bulk of the f1 brous matter has been removed through a prel1m1nary wash, was produced for export to S1ngapore, by the parut trampl1ng method, 1.e. ent1rely manual; the res1due was aTSoa profltable export commodlty. By 1923 the sago 1ndustry 1n the D1str1ct was "fa1rly large" (Johore, Ann. Rep. 1923), w1th numerous factor1es strewn along both tr1butar1es of 1ts main r1ver attest1ng to an apprec1able area of palmer1es 1n the h1nterland, w1th a subs1d1ary core of 800 acres of plantat10n palms 1n Kukup; export 1n 1925 amounted to nearly 1,430 tons (though 1t was qu1te overshadowed by tap1oca, wh1ch was a catchcrop with rubber or under sh1ft1ng cult1vat1on 1nland), produced by 24 factor1es (Johore Dep. Agr1c., Ann. Rep. 1925). However by 1929, unable to compete aga1nst the ava1lab1l1ty of a h1gh volume of better quallty tap10ca, espec1ally from the more mechan1sed factones of Java, sago product10n 1n the D1stnct was "on the decl1ne". Many small factones formerly ope rat 1 ng a 1ong the r1 ver banks ceased to work w1 th "the plantat1on at Kukup gradually bel ng removed"; st1ll over 690 tons were exported (compared to over 15,300 tons of tap1oca from the State, wh1ch was then the refuge of th1s 1ndustry too). By 1935 the sago 1ndustry, l1ke tap1oca, had dw1ndled, and the palm seemed to be used only to "augment the short supply of tubers" caused by restnct10ns on sh1ft1ng cult1vat1on of the tuber crop (Johore Pr1nc1pal Agr1c. Off., Ann. Rep. 1929, 1930, 1935), desp1te 1mprov1ng prlces.3 Johor had 1,814 ac w1th seven factor1es st1ll operat1ng by Par1t B1lal (Fa1rweather and Yap 1937); the Department of Agnculture recorded 4,596 ac 1n the pen1nsula, of wh1ch Johor had 1,500 ac, and sago was one of the few products to show a pos1t1ve trad1ng pos1t1on (Table 1). Dur1ng the Occupatlon 1n Sarawak 1t was noted that "the Japanese were hungry for starch of all klnds" (Ong 1948). That 25

TABLE 1 Sago Area In Malaya. 19:51-1949

1939

1940

1947

1948

1949

1,542

1,877

1,837

1,600

1,627

1,634

769

809

819

836

957

990

1,080

200

200

1,685

1,891

1,748

1,303

1,303

1,186

523

523

599

592

660

740

542

544

561

20

20

22

28

73

36

37

364

374

380

10

19

20

21

40

47

90

105

112

96

96

100

248

407

481

522

637

665

788

1,021

1,053

1,118

1,159

na

-

-

172

402

206

193

197

367

148

148

148

9

132

260

239

229

225

225

241

241

237

253

254

na

na

na

400

400

400

115

129

144

240

240

240

1931

1933

1934

1935

1936

1937

Johor

200

2,096

1,655

1,729

1,588

1,492

Kedah

-

615

615

615

387

Kelantan

na

200

200

200

Mel aka

61

144

163

Neger 1 SembI Ian

20

20

Pahang

na

Perak Perils

State

Selangor Terengganu

NOTE. SOURCE:

1932 and 1950 na; no sago In P. Plnang & S, Peral, Pee (1967!; continued In Appendix Ill,

1938 (acre)

TABLE 2 Sagopal• ltrea In Malaysia

Bara~,

1960

Compact Planting Area No. of Farms (acre) over 4 ac

Scattered Planting No, of No, of Farms Palms

State

No. of Farms

Johor

852

1,490

88

8,680

42,122

410 246 150 44

1,011 239 151 84

0 0 0 0

3,980 436 2,358 1,604

19,578 1,920 11,726 7,056

Kedah

164

79

0

7,902

75,408

Kelantan

106

48

0

9,302

119,010

Mel aka

336

286

2

3,348

26,648

40

10

0

8,978

37,696

Pahang

0

0

0

700

3,446

P, Ptnang & S, Peral

0

0

0

332

1,902

152

402

46

7,442

43,588

Batu Pahat Keluang/Segamat Muar Pont tan

Negert Sembi Ian

Perak Perils Selangor Terengganu

Malaysia Barat

Sa51o on

T}:~e

of Farm

0

0

0

2,052

37,444

102

166

0

1,912

12,840

4

2

0

2,618

17,734

136*

54,070

417,838

1, 756*

2,483*

Plantation

Scattered

Mixed crop

61

54

Ousun/Kampong

20

11

Rubber

13

14

Sa wah

6

17

NOTE:

SOURCE:

*Excluding 74 "urban" plantations reporttng 122 ac In Johor and 9 ac tn Selangor, Selvadurai 1960; cf, Appendix 111.

27

2,495 tons of wet sago could be exported from Batu Pahat D1str1ct 1n 1950 and that Kukup had a 300-acre "estate" even 1n 1947 1nd1cate that cores of the palm had been prudently reta1ned desp1te the very c1rcumstances that had caused 1ts gradual decl1ne as a commerc1al crop 1n the years before the Occupat1on. The flour was 1 oca lly consumed and the refuse was an 1 nva 1 uab 1e p1 gfeed at a t 1me when the pem nsul a was recoven ng from the depr1Yat1ons and destruct10n of the Occupat1on. In 1947-48 the market had 11 vened to such an extent that some fact ones could not obta1n enough palms (Johore, Ann. Rep. 1947, 1950). The Korean Boom s 1 gna 11 ed the postwar resuscltat 1 on of the useful agr1cultural product1ons, wh1ch 1ncluded sago. In 1952 there were 10 factones scattered 1n the D1stnct, w1th others also 1n Segamat and Senggarang; by 1956 the number had doubled w1th 12 chang1ng to mechan1cal methods. Cult1vat1on 1n the D1str1ct "cont1nued ga1n1ng popular1ty" (AO JW 258/ll), espec1ally 1n the L1nau and S1mpang Kanan muk1ms; the area 1ncreased to 684 ac out of the 1,600 ac 1n the State (cf. Table 2). There were another three factor1es 1n Muar and four 1n The res1due was worth up to $10/p1kul and Pont1an D1stncts. logs at 60-70¢/ft; these were sold to an1mal rearers (AO JN 125/52; AO JN 108/53; AO JN 108/54; DO BP 127/50). The State exported 4,285 tons of wet sago and 1,130 tons of flour 1n 1958, but demand exceeded the product1veness of the ex1st1ng palmenes so that not a 11 the 25 fact ones could operate fully; some c 1osed down dur1ng 1958-60. Th1s st1mulated new plant1ng, as a result of wh1 ch Segamat and Tangkak reemerged as 1mportant subs 1 d1 a ry produc1ng centres from the m1d-1950s; ow1ng to better dra1nage and market cond1t1ons, they d1d not develop to the SlZe of the Batu Pahat 1ndust ry. A notable change occurred 1n the postwar process1ng of the palm 1n the D1str1ct due to econom1c changes 1n S1ngapore and 1ts subsequent pol1t1cal al1enat1on from the pen1nsula: the product10n of "flour". Th1s was a1ded by techmcal 1mprovements w1th parts d1smantled from the S1ngapore ref1ner1es, wh1ch were becom1ng anachromst1c 1n the republ1c's 1ndustr1al sett1ng. By 1961 flour was equally 1mportant w1 th sagu basah, when Muar-Batu Pahat y1 e 1 ded 1,880 tons of wet sago and 1,644 tons of flour, w1th Batu Pahat prov1d1ng over 90% of the latter. Pant B1lal rema1ned the nucleus of process1ng. Sago was also "qu1te a prof1table 1ndustry to ammal rearers and attap makers" (Muar D1str1ct Ann. Rep. 1961). Subsequent progress 1n sago product1on 1n Batu Pahat D1str1ct 1s d1scussed below.

28

Geography of the Sungei Batu Pahat Basin The southwest Malayan pla1n was once a reg1on of swamps, some cont1guous ones of relat1Vely large extent rema1n1ng even after the p1oneenng dra1nage works had been establlshed. The Sunge1 Batu Pahat nses from the southernmost extens1on of the central mountaln range, the spr1ng of the maJor r1Ver systems of the south Malayan penepla1n. It formed the southern border of the Portuguese enc 1ave 1 n Ma 1aya after the1 r conquest of Me 1aka 1 n 1511; 1t 1s depleted as the R1o Formosa 1n the1r maps. The r1Ver 1s a 1so famed as the western connect 1 on of the anc1 ent southernmost overland route between the Stralts of Malacca and the South Ch1na Sea. It shares an 1nland tr1butary, the Sunge1 Semberong 1n the V1c1n1ty of Keluang, w1th the Sunge1 Endau on the east coast of Johor; th1s once vast swampland fac1l1tated 1nland nav1gat1on (F1gure 1). The Sunge1 S1mpang Klrl, 1ts r1ght tr1butary (contrary to occ1dental custom, the mant1me Malays take the1r r1ver d1rect10n from 1ts mouth, not 1ts source), also marked the northern boundary of the Orang B1nua, sa1d to be the most southerly of the Indoch1nese abon g1 nes; now ext1nct, they were numerous enough 1 n the mld-19th century to have been stud1ed also by Logan (1847). The 01stnct 1s about 1,000 sq m1les and half of 1ts area 1s peaty, most lowly1ng. Batu Pahat 1s the urban-commerc1al centre of the central west Johor reg1on. Its name comes from one of 1ts earl1er act1v1t1es, the quarry1ng of stone, batu, w1th the ch1sel, pahat, along the Sunge1 Kondo, renamed the Sunge1 Penggaram. A 1938 map (lodged at the Ark1b Negara, Petal1ng Jaya) of the r1ver1ne env1 rons of Batu Pahat town shows that the present urban area earl1er had the typ1cal equatonal littoral vegetat1on (F1gure 3): between the mangrove forest (bakau) fnnge establ1shed on the sal1ne t1dal flats and the 1nTii"'ii'CI alluv1al plain was the swampy bas1n, bu11t up from the fluv1al sed1ments of the r1ver, veneered by per1od1cal tldal depos1t1on on wh1ch pockets of cult1vat1on were develop1ng. The 1ntert1dal zone, poorly dra1ned, was dom1nated by two types of edaph1 c ecotones: the brack 1sh t 1da 1 banks supported groves of d1sturbed forest, belukar, and mpa, wh1ch was absent from the coast but appeared more 1nland than would be usual because of the flatt1sh grad1ent, backed landwards probably up to the h1gh-t1de mark by permanent freshwater swamps where undra1 ned, wh1le the dra1 ned areas probably had a h1gh water-table at or JUSt below the land surface on wh1ch sago palmenes were developed. On th1s map, some groves were marked "sago" and others "rumb1a", but 1t 1s not known why thls d1st1nct10n was made, unless perhaps to d1fferent1ate between the 29

coconut

0

*

i

mile N

f road par1 t sago palmery

N

R

rumb1a

s

sagu

+

sago m1xed w1th other crops, bel ukar

After "Bandar Penggaram (Batu Pahatl", 1938, scale 16 1nch = I m1 le: Je.4/J93, Arkib Negara, Malaysia

.. lB-...•

FIGURE 3 Agricul-tural Geography of Bander Penggar•. 1938

1ntroduced or commerc1al and the natural1sed or domest1c stands, or even between the smooth and sp1ny var1et1es used. Parts of the present sago pa 1meri es were estab 1i shed on former sawah and belukar. Two other palms, the coconut and the betelnut, Areca catechu, were also cultiVated to a large extent. The penchant for palm cult1vat1on 1nd1cates some knowledge of crop opt1ons, wh1ch pers1sted 1nto the early postwar years when 1ngress of bracklsh waters cont1nued to be a hazard to cult1vat1on. 30

The old dra1ns had tended to follow the topography; where strearn flow was slugglSh they pursued a longer course. Haphazard s1t1ng and w1de spac1ng of the par1ts rendered dra1nage even less effectlVe when 1nland agncultural development dunng the 1nterwar penod followed road and ra1l construct1on through central Johor, overload1ng the r1vers w1th sed1ment; neglect dun ng the Occupat 10n caused the system to be 1 neffect 1ve dur1 ng the bdd flood of 1949 (Johore, Ann. Rep. 1946-47, 1949). The propens1ty to flood was aggravated by stra1ghtemng and scounng of the upper reaches of the rlVer w1thout correspond1ng work on the lower; the deepen1ng of the rivers led to 1ngress of seawater .at hlgh t1de, wh1le the shallow depth of the pants, unsu1table locatlOn of bunds and 1nsuff1c1ent number of t1dal gates encouraged s1lt1ng and pond1ng, so that flood1ng was not1ceably extended 1n t1me and space; crabs aggravated eros1on (GA 328/1939; DO BP 159/50; DO BP 170/50). The b1ology and hydrology of th1s basin have yet to be stud1ed adequately. There 1s a c1rcadian cycle of t1dal 1nundat1ons in the lower r1ver banks, every 25 hours, wh1ch decreases 1n 1ntens1ty further upstream and away from the banks, and brack1shness 1s encountered on sago holdings under tidal 1nfluences; there 1s also an annual spr1ng t1de. The t1dal waters r1 se up to 6 ft up to the Sunge1 Semberong tw1 ce da1ly, and to 10 ft dur1ng the tw1ce-monthly (lunar) neap t1des when co1 nc1 d1 ng w1 th the Northeast Monsoon ra 1ns; 1t used to spread water a quarter-m1l e or deeper 1 nland from the Sunge1 Simpang Kanan dur1ng the monsoon. The maJor rivers of the Di str1 ct flow 1 nto the Sun gel Batu Pahat, and ow1ng to the relat1ve flatness of the region any pressure on the dra1nage system 1n one area affects all, part1cularly w1th1n the confluence pla1n. Waterlogg1ng, e.g. in depress1ons of the confluences of the Sunge1 Bekok-Semberong and the tw1n Sunge1 S1mpang, 1s due to ra1nwater accumulat1on 1n the 1mpermeable areas, 1nadequate dra1nage and seawater 1ntrus1on (attested to by groves of n1 pa up to the footh1ll s of Buk lt Payong 1n older maps; a fa1rly large grove pers1sted east of Pant Sulong, near Bt. L1nau: L7010, 1973, ed. PPNM, Sy1t 122, 1: 63360). T1dal backup effects reach Ser1 Medan on the Sunge1 Slmpang K1 n, on to Yang Peng on the Sun gel Bekok and Sawah Assam Bubok on the Sunge1 Semberong, causing severe flood1ng when high-t1de cond1 t 10ns co1 nc1 de w1 th floodwater flows dur1 ng the Northeast Monsoon, the oversp1ll 1nundat1ng a large part of the floodpla1n; there 1s also s1lt1ng of the m1ddle course. The effects of waterlogg1ng are aggravated by the pedology of the reg1on, wh1ch lS character1sed by organ1c so1ls, the Llnau-Sedu Ser1es, the most ext ens 1 ve of the three maJor so1l groups, Wl th 94% of the 31

peat over 5 ft th1ck, down to 25 ft, underla1n by mar1ne clay. The trad1t1onal sago area 1n thls bas1n 1s now centred 1n the tongue of 1and, kua 1a or muara, at the confluence of the nver (F1gure 4). The left bar1i("(if the Sunge1 S1mpang K1r1 1s reputed to be the geograph1c core of sago farm1ng, cult1vat1on of the palm here dat1ng back two centunes accord1ng to local lore. The best palms are cla1med to come from th1s hearth, mula mula, through wh1ch a sunge1 mat1 traverses, the Sunge1 Selulon, now a of Parit Selulon {although 1ts earl1er reg1mented part unpred1 ctab 1e meander1 ng 1s marked on var1 ous maps). It 1s not clear whether the alleged product1veness of the palms here 1s due to the cult1var used, the so1l or the expert1se of the farmers; over the centur1es, accret1ons from the suspended sed1ment ca rr1 ed 1n floodwaters and runoff from h1 gher e 1evat 10ns formed fert1le banks of alluv1um on wh1ch the palm flour1shed. Impeded dra 1nage 1n the kua 1a a 11 owed sago to be cult lVated nght up to the footh1l1s of Buk1t Payong and 1ts leeward slope wh1ch 1s dra1ned by the Sunge1 S1mpang K1 r1. The water-table here 1s st1ll h1gh, often less than 3 ft below the so1l. The terra1n had promoted the development of organ1c so1ls, 6 1nch to 10 ft deep (AA BP 14/57), thus render1ng the area even more unsu1table for convent1onal cropp1ng; 1ts peaty nature can be read1ly d1scerned from many par1ts adJo1n1ng the upper Sunge1 S1 mpang K1 r1 , where the water 1s perpetually of a tea co 1our. The Department of Agnculture {1922; Gr1st 1936) noted that sago grew well where r1ce and coconut repeatedly fa1led. One may therefore cred1t the sago farmers w1th hav1ng made product1ve an otherw1se agnculturally unmanageable part of the peninsula at a t1me when the adm1nstrat1on fa1led to prov1de any useful a1d to these p1oneers. In the last few years, the Dra1nage & Irngat10n Department has 1mproved the already dense dra1nage network, by deepen1ng and w1den1ng the channels rather than 1ncreas1ng the dens1ty of shallow dra1ns (contrary to 1ts prewar efforts wh1ch were m1ndful of the problem of seawater 1ngress), 1n read1ness for the 1mp 1ementat 10n of Phase II of the West Johor Agn cu 1tura 1 Deve 1opment PrOJect, a soph1 st 1cated agroeng1 neen ng scheme that w1ll cover about a m1ll1 on acres of mu lt 1cropp1 ng of southwest Malaya. A perceptlble lower1ng of the water-table has occurred over large areas of the D1stnct such that farmlands that were once subJected to the most unfortunate cycles of fl ood1 ng are now su1table for the cult1Vat1on of those very crops wh1ch earl1er could not flour1sh. On the other hand, w1thout the per1od1c 1nundat1ons or w1th the lowenng of the water-table under the dra1nage scheme, sago plant1ngs are suffenng from drought stress, and there are fears that the palm w1ll d1e out. 32

.

~

~,

Sen Hedan

:l~

f ,..

·

Bagan

1966

Swamp 1974

Sago 1966

1974

847

809

6,215

18,923

Kampong Bahru

347

151

16

4

Llnau

370

362

621

2,327

Lubok

636

657

3,320

2,232

43

2

Peser!li

550

421

37

79

S 1m pang Kanan

322

298

1,026

1,716

42

460

828

3,312

Ser1 Gadlng

7,636

7,303

74

100

Ser 1 Madan

3,084

10,031

190

1,122

Sunge I Keluang

2,105

2,486

sungel Punggor

2,583

1,981

10

Tanjong Semberong

10,085

9,454

60

535

Batu Pahat Dlstr1ct

38,142

55,568

2,905

9,208

Cha 1 ah Baru

M1nyak Beku

Slmpang Kir1

SOURCE

11

WOng 1979.

M1scellaneous Crops data have not proved to be rel1able enough to encourage a conf1 dent conJecture on the fate of the new and 1ncreased plant1ngs s1nce 1974, for 1n that year alone they 44

dl verged severely 1n sever a 1 States from the Land Use data; the 1atter source a 1so fa 11 ed to document the ex1 stence of the 1arge number of small groves 1n the Kelantan floodplaln (the maps cannot dep1ct plant1ngs of less than 2 ac). Nonetheless, 1n the evolut1on of the sagopalm as a commerc1al crop 1n Malaya, the southwestern coast of the pen1nsula appears to be 1ts last man1festat1on; all earl1er centres are on the verge of ext1nct1on, 1f not already defunct, as around Kota Bharu, Muar and all of Melaka (Table 5; Appendl>' V). But even th1s core may not be durable. Younger farmers are not devoted to the crop like the1r predecessors and the s1tuat1on 1s further underm1ned by the government po 11 cy of promot 1ng v1 s1 b ly more attract lVe crops, to wh1ch end the sago hab1tat was progress1vely or 1nadvertently destroyed. There seems to have been an apprec1able reduct10n 1n the area at least s1nce the 1974 reconna1ssance. Off1cial stat 1st 1cs and reports 1nd1 cate nelther the extent nor character of th1s trend although the D1stnct's Department of Agr1culture conf1rmed 1ts d1rect1on 1n certa1n subd1str1cts. Whereas sago plant1ngs by the Sunge1 S1mpang Kanan all adJ01n 1ts banks of alluv1al clays, averag1ng 0.5 m1le landwards, a d1ffused d1str1but10n 1s apparent 1n the peaty Sunge1 S1mpang K1r1 bas1n, where 1mpeded dra1nage on the organ1c so1ls of the western part of the confluence tongue around Sunge1 Selulon allowed a greater d1spers1on of sago farm1ng to the 1nterior (F1gure 4). The sago h1nterland became geograph1cally consol1dated, w1th expans1on occurr1ng w1th1n the core area 1tself after 1966. The largest plant1ngs totalled 3,312 ac 1n S1mpang K1r1, the area quadrupling dur1ng 1966-74; the other large sago areas are 1n L1nau, Ser1 Medan, S1mpang Kanan and The Ch1nese apparently dom1nate TanJong Semberong (Table 6). sago farm1 ng 1n S1mpang Kanan and Ll nau, but there has been a decrease 1n the area 1n these two muk1ms accord1ng to the Department of Agnculture 1n Batu Pahat; 1n relatlVe terms the 1ncrease was greatest 1n TanJong Semberong. Much of the cont1guous planting along the Sunge1 S1mpang Kanan 1s sa1d to belong to the process1ng sector, wh1ch would account for the unusual concentrat1on of the maJOr starch factor1es below Tongkang Pechah, cons1der1ng the cons1derable d1spers1on of sago cult1Vat10n elsewhere and the 1ncreased area 1n S1mpang K1r1. Before the Occupat1on, there were a few factor1es s1ted near the Sunge1 S1mpang K1r1 too, whereas at present there are none. Th1s suggests that Malay planters could have p10neered sago cult1vat1on 1n the D1str1ct 1f not plantat10n sago, most probably for domest1c use and were encouraged by Ch 1nese manufacturers who entered the sago trade. But to meet the1r need for a h1gher volume of raw mater1als, the Ch1nese themselves seem to have extended cult1vat1on to other nvenne hold1ngs for wh1ch too there was no agr1cultural use; booms 1n 45

the sago trade cons1derably too.

soon

encouraged

the

Malay

areas

to

expand

The Rubber Industry Smallholders Development Author1ty (RISDA) promoted sago cult1vat1on among the Malay peasantry dur1ng 1ts most st1mulat1ng per1od of evolut1on as a modern cashcrop, support1ng the convers10n of over 11,000 ac to sago (Table 7), or an average of 950 ac annually dunng the penod 1966-74. Farmers were encouraged to substltute rubber w1th other crops (Table 8), because 1t was recogmsed that 1n th1s bas1n 1n part1cular the env1ronment was far from su1table for th1s perenn1 a 1 and because of the recommendat 1on that "coastal alluv1um should be reserved for the cult1vat1on of crops other than rubber" (AO JN 114/52). There 1s a standard plant1ng grant of $1,500 per acre for crops other than rubber, wh1ch 1tself mer1ts $2,200). In the case of sago 1t 1s glVen 1n 1nstalments of $300 for the f1 rst year and $200 1 n each of the succeed1 ng three years; unllke for other crops, 1t barely t1des the farmer over the unproduct 1 ve pen od of 1mmatur1 ty, though 1n the past needy farmers sold forward the more v1gorous palms to compensate for th1s def1c1ency. The sago 1ndustry cons1sts of two sectors: the almost selfsuff1c1ent Ch1nese plant1ng-process1ng sector and 1ts Malay plant1ng adJunct, wh1ch appears to be largely Javanese; the latter denves 1ts commerc1al surv1Val from the processing sector, wh1ch 1tself 1s completely at the mercy of external market 1 nfl uences. The process 1 ng sector probably controls d1rectly the bulk of the sago area 1n the D1str1ct for sound econom1 c reasons but 1 ts pattern 1s unknown as off1 c1 a l data on sago land ownersh1 p have not been compll ed; the process 1 ng sector draws lts ep1sod1c need for raw mater1als from the Malay sago farms. Some of the old sagolands under Ch1nese ownership could have been purchased from 1mm1 grant farmers who returned to the1 r homelands 1 n earl1 er decades or who became 1mpoven shed during earl1er cycles of depress1on; 1n recent years, more were sa1d to have passed 1nto Ch1nese ownersh1p when Malays J01ned land development schemes such as those of the Federal Land Development Authorlty. Certa1nly, 1mm1grant Malays 1n part1cular had tended to farm out to the Ch1nese the collect1on of produce from land they had l abonously worked w1th borrowed cap1tal (GA 172/1915). After such debts were pa1d off a mod1fled form of the relat1onsh1p evolved wh1ch d1splays a few unusual features that are d1scussed later. The Malay sago farmers have shown ne1ther w1ll1ngness nor ab1l1ty to part1c1pate 1n the process1ng and market1ng of their sago produce, rem1n1scent of the1r role 1n the earl1er copra 1ndustry, and appear content wlth the preva1l1ng system of explo1tat1on. Th1s perm1ts them to pursue other farm1ng

46

TABLE 7 Sago Replan"tlng under RISDA In Johor. 1952-1978

Year

Acreage

1952-56 1957 1958 1959 1960 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978

235 793/4 1473/4 1913/4 763/4 583/4 gz1f2 2003/4 5051/4 8861/2 522 8373/4 1,2713/4 1,1141/4 7431/4 301 980 1,220 1,039112 5773/4 1221/4 753/4 273/4

Johor Distnct

Acreage

Batu Pahat Ponti an Muar Keluang Kot" Tingg 1 Segamat Marsing Johor Bahru

9,928112 6853/4 4223/4 1643/4 481/4 341/2 171/4 6

Johor

SOURCE:

RISDA, Batu Pahat. v. Appendix VI.

47

TABLE 8 Rubberland IHid Replant-Ing In 11185-t Johor Project Phase 11. 1953-1975

No, of Farmers

Crop

Acreage

na

127,468

Coconut

2,254

9,018

Sago

2,229

8,919

Ol I palm

1,438

7,193

Non-citrus trees

1,210

3,633

Pineapple

265

798

Coffee

281

846

73

220

7,750

30,627

Rubber

Citrus fruit

Total.

SOURCE:

Other crops

RISDA, Batu Pahat,

act1v1t1es, 1n keep1ng w1th the1r trad1t1on of and still marked preference for polycultural farming. The b1ology of the palm, wh1ch allows 1t to be harvested anytime over a long penod, and 1ts ab1l1ty to thnve desp1te a great deal of neglect, and the mutually useful commercial relations of the two commun1t1es, promote the almost indifferent Malay attitude to sago farm1ng and Chinese entrepreneunal skills 1n organ1s1ng commercial production. In the early 1970s, there were 33-36 sago factor1es 1n Johor, the maJOrity concentrated 1n th1s D1str1ct;4 now about half are left, st1ll ma1nly clustered near Par1t B1lal. Water transport IS VItal to the Industry; Fairweather and Yap (1937) believed "1t IS this cheap form of transport that enables the Industry to survive". There are two pearl factones, pearl1ng requ1r1ng more sophiSticated machinery; half a dozen factories

48

produce a pr1me-qual1ty starch because they have access to clean water, usually p1 ped from h1ll streams. All starch and pearl factones requ1re l1beral access to water, although not all can obta1n clean water; hence they are t1ed even more closely 1n the1r locat1on to the ctra1nage system. All sago factones are Chlnese-owned. In the early 1970s so opt1m1st1c was the mood 1n the 1ndustry that some operators moctern1 sed thel r plants; but the 1eve 1 of techno 1ogy 1s not yet on par w1th the Chemor tap1oca factor1es, where automated centr1fuges for extract1ng the starch were already 1nstalled. None of the sago factory owners appeared to have knowledge of th1s method of process1ng or was 1nterested 1n us1ng 1t, presumably because of the unstable market s1tuat1on. The 1977 s 1ump 1n the 1ndust ry proved a t raumat 1 c exper1 ence for both processor and peasant farmer due to two ma1 n factors: f1rst, off1c1al pol1cy favours other crops, and 1n the sago area th1s 1s at the expense of the hygrophyt1c palm, because the most popular1sed crop 1s the o1lpalm, wh1ch has a s1m1lar ecolog1cal preference; second, accord1ng to the processors, by late 1978 there was a marked decl1ne 1n the demand for hampas, the p1thy byproduct of sago process1 ng that was once a pr1 zed p1 gfeed. The marg1n of prof1tab1l1ty for sago processors thus has been severely reduced, a s1tuat10n aggravated by the dump1ng of starch of one k1nd or another on the reg1onal market by ne1ghbour1ng states, so that prices are even more depressed. Processors 1n the past persevered w1th starch product1on even under slump cond1t1ons because they could get by from the sale of the hampas; 1n earl1er decades a hampas shortage, be 1t from tap1oca or sago process1ng, could even threaten the survlVal of p1 g-rear1 ng, and 1t had to be compensated w1 th 1mports from Sumatra and Java, where a commensurate p1 g-rear1 ng 1ndustry d1 d not ex1st for rel1g1ous reasons up t1ll the early postwar era, In an even earl1er era p1g-rearing was an 1ntegral part of both starch 1ndustr1es, prov1d1ng a subs1d1ary l1vel1hood and add1t1onal food to the labour force on the Ch1nese plantat1ons. The reasons for the extraord1nary change 1n the status of hampas are the modern1 sat 10n of 11 vestock fa rm1 ng and Government rest r1 ct 10ns on haphazard p1 g-rean ng 1n deference to re 11 g1 ous sens1t1v1ty. Smallt1me p1g-farmers and householders who once absorbed the hampas have been put out of bus 1ness; wh1l e ducks thrlVe on th1s feed, demand from th1s sector is too small. In 1975, 31 1arge feedmlll s produced $195 m1ll1 on worth of products; the annual turnover exceeded $225 mill10n dur1ng 1975-79 {Dep. Stat1st1cs). Because of propaganda about the super1or1ty of lmported cereal teeds, the more substant1al an1mal-farmers are turn1ng to these more nutr1t1ous feedstuffs, wh1ch conta1n no 49

sago. The larger feedm1lls are d1s1ncl1ned to use local raw mater1als; unw1ll1ng to pay equltable pr1ces, they tend to reJect these for the1r lower qual1ty. What was once a remunerat1ve byproduct and the cheapest feedstuff 1 ngred1 ent now encumbers the very env1 ronment 1n wh1 ch subs 1stence farmers cannot afford to follow more modern costl1er methods of l1vestock farm1ng. (The latest stat1st1cs contrad1ct th1s s1tuat1on descr1bed by the processors, the average FOB pr1ce be1ng h1gher than earl1er years [Append1x VIII]; but at the t1me of my survey, the hampas problem was the sorest subJeCt. The contrad1ct1on bears future study.) Further, w1th the technology preva1l1ng 1n the D1str1ct, sago product10n 1s labour-1ntens1ve, and the 1ndustry 1s fac1ng a severe labour shortage because the slump has prevented processors from compet1ng effect1vely for sk1lled labour;5 due to 1ts prox1m1ty to S1ngapore, Johor labour 1s the most expens1ve 1n the pemnsula, and sk1lled labour 1s most scarce. Investment 1n raw mater1al supply used to be a cons1derable burden, v1nd1cated by the h1gh proflts that could be expected, but only a few processors are now able to take advantage of the slump to purchase stems at the low preva1l1ng pr1ces. The future 1s equally bleak for peasant farmers who persevere w1th sago cult1vat1on when they are unable even to sell forward l1ke before. Desp1te the fact that sago has been an 1mportant secondary crop 1 n Johor 1t appears to have been 1 gnored 1 n the current amb1t1ous scheme to develop the western coastal belt of the State, v1 z. under the West Johor Agr1 cu 1tu ra l Deve 1opment ProJeCt, f1nanced by the Internat10nal Bank for Reconstruct10n and Development, though 1t mer1ted attent1on 1n the prel1m1nary stud1es for the Pahang Tenggara scheme (Flach et al. 1971). Phase II of the Johor ProJect emcompasses the ent1re Sunge1 Batu Pahat bas1n, w1th 12 out of the D1str1ct's 13 muk1ms 1ncorporated 1n the PrOJect area. Most plans 1gnore the sagopalm, cons1der1ng p1neapples as the fourth most 1mportant current crop (Table 9), although the acreage of sago (2.6%) 1s certa1nly larger than that of pineapple (0.6%) and smallholder o1lpalm (2.1%). Yet, about 8% of the ProJeCt area of over 0.5 m1ll1on ac, or 38,000 ac, are marshy (cf. Table 6). Some 124,000 ac of the hum1d alluv1al clays and organ1c so1ls now under rubber are largely marked for replacement by the o1lpalm, deeper peats elsewhere by annual crops. To fac1l1tate th1s, o1lpalm cult1vat10n w1ll be supported by a complex comb1nat1on of water-table management and organ1c potash manur1ng (Shepherd 1977). The ProJeCt authont1es were aware that Phase II had more complex hydrolog1Cal problems than Phase I to the south - also, the former covers tw1 ce the area of the latter - and that the water-management system had to be a m1t1gat1ve rather than a 50

TABLE 9 Crops In Ba"tu Paha"t 0 I s1T lc"t •

1966 and 1974

Crop

1966

1974

Rubber

194,066

177,903

Coconut

72,322

78,590

01 I palm

2,478

19,026

Sago

2,905

9,206

Pineapple

3,096

6,042

Coffee

1,646

3,673

Pad I

2,379

2,068

Arecanut

2,143

32

290,967

323,561

(~ere)

Total cultivated land

SOURCE·

Wong 1979,

correct 1 ve one. Because the reg1 on had been so extens 1ve ly and 1ntens1vely culttvated and densely populated for a long t1me, 1t waul d have requ1 red an even more elaborate programme of changes to the ex1st1ng agrartan system and tnvolved far greater capttal costs, e.g. 1n land acqu1s1t1on, and could cause Incalculable soctoeconomtc dtsturbance to the peasantry. Desplte 1ts coastal pos1t1on and the presence of swamps, the bastn 1s one of the drter areas of the pentnsula, wtth ratnfall averaging 70-80 tnch per year as aga1nst the nattonal average of 12D tnch; 63% of the ProJeCt area 1s dratned by the Sunge1 Batu Pahat, and surface water resources are only 30-40 1nch per year. The pre 11 m1 nary 1ntens 1 f1 cat1 on of dra 1nage 1 s beg1 nnt ng to show the not unexpected detrtmental effects of dryland cropptng on the ecology of the reg10n, e.g. 1ncreased ac1d1ty of the peaty so1l and drought effect 1n marg1nal areas. Thts would suggest that

51

unt1l suff1c1ent water catchment areas are found and conserved, the dra1nage 1ntensity already ach1eved after decades of peasant endeavours may have reached the opt1mum level, 1f not gone past 1t, Eng1neers, apparently w1thout proper cogn1sance of env1ronmental pnnc1ples 1n land use plann1ng, tend to d1sparage a system that, w1th a llttle adJustment, may be env1ronmentally sounder than the planned total dess1cat10n of a spongy env1ronment to render 1t ent1rely amenable to dryland cult1vat1on. For compar1son, hydrology stud1es of the Sunge1 Ke 1ant an fl oodp 1a 1n 1nd1 cate that ow1 ng to the rap1 d aggradat 1on process 1n the nver bed, "apart from nver bank protect1on work ... and careful management of development 1n the h1nterland, no 1mmed1ate measures that are econom1cally JUStlflable can be taken aga1nst th1s aggradat1on process" (Tonk1n &Taylor 1977),6 There 1 s a water def1 c1 ency of some grav1 ty 1n the ne1ghbounng densely settled Pant Jawa area, wh1ch was dra1ned some 20 years earl1er, a portentous m1crocosm of what could happen 1n the present Batu Pahat floodpla1n (Parlt Jawa v1llagers had to 1eave Jars by the roads 1de to be flll ed by water tankers), The 1mprovement of dra1nage had been expected to result 1n 1ncreased so1l fert 111 ty, but "the effects have been qu1 te the oppos1te the problems of dryland farm1ng on dra1ned pyr1te so1ls are obv1ous (ow1ng to the1r 1ncreased ac1d1ty). Over the last seven or e1 ght years very great 1mprovements of o1l pa 1m on ac1 d su 1ph ate so1l s have resulted from the suggest 1on that the water-table should be ra1sed and ma1nta1ned as near to the surface as 1s pract1cable" (Joseph and Narudd1n Maarof 1977). The farmer of th1s reg1on have had more than a century of recorded expenence w1th the problems of cult1vat1ng the fluv1al bas1n w1th orthodox crops, pnnc1pally rubber and coconuts, wh1ch were spread out over a versat 11 e spectrum of crops, of wh1 ch the sagopalm used the wettest parts far more eff1c1ently 1n econom1c and ecolog1c terms than any of the other crops they had exper1mented w1th. The unt1lled sagopalmenes fnng1ng the Sunge1 Batu Pahat system may yet prove to be the most durab 1e bankhold1ng vegetat1on ever establ1shed for controll1ng and stab1l1s1ng the seasonal sp1llover effects 1n a region that already has water-def1c1t enclaves of settlement and cult1vat10n border1ng the floodprone valleys. It 1s a measure of the perenn1al off1c1al d1sda1n for the sagopalm that 1n the crop capab1l1ty assessment of a reg10n noted for 1ts poor dra1nage, a v1able rural 1ndustry, wh1ch made 1ts least cult1vable land h1ghly product1ve even before colon1al t1mes, was d1sm1ssed; the State Econom1c Development Corporat10n man1 fested no 1nterest 1n the crop e1 the r, The att 1tude that wet 1ands not used as sawah should be dra 1 ned for dryl and crops preva1ls, even 1f th1s 1nvolves alterat10n at great cost of an

52

env1ronment of some complex1ty, even frag1l1ty. S1nce the last refuge of the sagopa l m seems to be the Batu Pahat 01 st r1 ct, the west Johor PrOJect would sound the deathkne ll of the 1nd1 genous plantat1on palm 1n the pen1nsula unless there 1s a change of po 11 cy, wh1 ch wou 1d need to be supported by a rat 10na l programme of redevelopment and research, as had been prov1ded for the p1neapple earl1er. The Malays1an Agr1cultural Research & Development Inst1tute has been d1rected to render appropr1ate a1d. But the 1mmed1ate concern of the 1ndustry 1s access to a suff1c1ently large hlnterland of supply, 1n order to ensure 1ts cont1nued ex1stence as a compet1t1ve rural 1ndustry.

NOTES 2

Pea-s 1 zed sago was mentIoned at the turn of the 13th century by Chau Ju Kua ( H 1rth and Rockh II I 1911) as beIng used as a standard of exchange 1n the Javanese dependencIes, wh 1ch Inc I uded the Sundas and parts of Borneo then. According to Dalrymple (1770), granulated sago was made by toasting flour on hot pans over a fire, in Mindanao, "sago is dried in small pieces I ike ltttle seeds or comftts" oe> oe> c::> C>

C)O

C> 0

00

co

j

0

(~

oo.n c r ' ·· om.;)n~ gt:! r' s r ~s 1

dcncc

0.

~

0000

ogq

00

at t rtt·

r-- ___ _ __

El Q lJ

1 l ha l chea

lJ ~ fl

L

j

s h e I t e r f vr M t ilp - ma k e r

illJIHf---- --

oooo ooo

t>un d I es o f s ago l ea11es

n o tt~

FIGURE 8 ec.poslte Model of • Sligo Efttwprlse on Pllrlt Bllel

; :

The locat1on of the sago factory, w1th the except10n of the few wh1ch produce meal, 1s pr1mar1ly governed by water supply for two purposes : transport 1ng raw mater1 a 1 s and wash1 ng the products. Thus locat10n to a parlt 1s essent1al, somet1mes 1t may be necessary to be near two par1ts, one for discharg1ng effluent 1nto and for stor1ng logs, the other, the cleaner one, to supply water for the wash1ng. In the case of pearl manufacture and grade-1 starch, access to a h1ll stream is essent1al; th1s 1s p1ped to storage tanks. Two or three factor1 es have the reputat1 on of produc1 ng pr1me-qua 11 ty starch su1table for noodles. Wh1le the t1dal effect 1s useful 1n ass1st1ng in transport, 1t 1s a bane 1n that if 1t br1ngs brack1sh water upstream, the qual1ty of the starch and pearl products w1ll be degraded; 1t 1s said to 1mpart a sl1ght salt1ness and to affect the rheolog1cal properties of the starch. For th1s reason the pearl factor1es use only starch from the best factor1 es, wh1l e they must wash the crude starch purchased from the smaller, less fast1d1ous factor1es before thls can be m1Xed with the1r own for a f1nal wash.

Ownership of factories Seventeen sago factones were v1s1ted, 1nclud1ng a small wet-flour factory 1n the Muar suburbs wh1ch used manual paddl1ng to settle the starch 1n wooden tubs (s1nce fltted w1th motor-dr1ven blade st1rrer) and cheesecloth to separate the f1 brous res1 due (as dep1 cted 1 n Fa1 rweather and Yap 1937). They 1ncluded all the maJOr factor1es 1n the 01strict (Table 15), but excluded those 1n sago-meal product1on wh1ch are emerg1ng as backyard operations, l1ke the Chemor tap1oca-ch1p factor1es wh1ch compete 1n the same market - domestic p1 gfeeds - but 1t 1 s an up-and-com1ng small 1ndustry wh1ch has hopes of replacing the depressed hampas trade 1n this part of the pen1nsula. Seven factor1es were solely owned or 1n partnersh1p w1th sons, the other 10 were partnersh1ps, usually w1th other relatlVeS, that tend to be more substant1al and specialise in the h1gher-value product manufacture. A few were once partnersh1ps, but the Joint owners branched off on their own dun ng the boom years; th1s 1s 1n contrast to the current tap1oca 1ndustry where consol1dat1on 1s the trend because of the market1ng advantage and techn1cal 1nvestment. Although only three processors stated that they had 1nhented their establ1shments, the proport10n of 1nher1ted properties 1s probably h1gher, s1nce several current owners adm1tted to bu1ld1ng on the f1nanc1al or phys1cal structures of prewar fam1ly propert1es. F1fteen of the owners were Hokk1ens, about half belong1ng to 100

TABLE 15 Types of Sago Fac:f'ory Vlsl"tecl

Pearl-starch

2

Dry

ste~rch

5

Wet

t lour

5

Meal

the Tangkwa d1alect p1oneers, a group prom1nently 1n the perpetuating the p reduct 1on.

4

group; the other two were Teoch1 u. Among the of Tangkwa Hokk1ens, surnamed Kerk, f1gured 1ndustry, and the1 r sons and grandsons are bus1ness; one fam1ly monopol1sed pearl

N1ne were maJOr concerns, as SlZe goes 1n th1s 1ndustry, produc1ng the better quallty or more ref1ned products, each w1th da 1ly 1nput capaclty of not less than 60 logs, a few cons 1derably more. The maJOr1ty of them had the1r own plant1ngs. N1ne processors cla1med to be ent1rely dependent on external suppl1ers; the maJOr1tY were smaller operators. Most of the substantial processors belonged to the 11-member South Johore Sago Assoc1 at 10n {the 1933-34 Comm1 ss1 on recorded the ex1 stence of a Southern Sago Flour Assoc1at1on wh1ch had 15 members). Ten factorl es were estab 11 shed 1n the postwar rehab1l1tat 10n per1od up t1ll 1965. Another seven were less than a decade old, started or modernised du r1 ng the 1ast boom 1 n the early 1970s. A few were mod1f1cat1ons of an old establ1shment res1ted 1n a more conven1 ent l ocat 10n. Wart 1me damage to th1 s 1 ndustry was cons1derable dur1ng the Occupation, as 1n all 1ndustr1es 1n the country, but unl1ke them, the obsolescent technology was largely restored. Almost all factor1es used fam1ly labour l1ke h1red hands. Managers were 1nvar1ably the owner h1mself or a close member of the fam1ly where the owner had other 1nterests bes1des sago. One factory d1m1n1sh1ng 1ts scale of operat1ons ow1ng to var1ous tr1bulat1ons, such as recalc1trant labour, regularly employed over 30 labourers before the slump. The current largest labour force, full-t1me, exclud1ng fam1ly members, was 23 on the payroll, for a purported dally output of 3.5-4.2 tons dry starch. Most factor1 es tended to employ more youths and women than men

101

for work beyond the debark1ng stage, espec1ally those mak1ng starch and pearl. Th1s 1s not a s1mple matter of explo1tat1on. Many dry manufactur1ng procedures 1n a more rat10nal1sed setup wou 1d be done v1 a conveyor be 1ts, chutes and p1 pes, as 1n the tap1oca factor1es. However, here the conveyor belt 1s human w1th a lot of bucket work 1nvolved. Such labour 1s st1ll cheap enough to encourage a technolog1cally anachron1st1c enterpr1se, wh1ch 1n turn prov1des employment of some k1nd to th1s depr1ved rural populat1on. There were three groups of factor1es. The more endur1ng group 1s along the Batu Pahat-Yong Peng Road by Par1t B1lal, clustered below the br1dge over the Sunge1 S1mpang Kanan south of Tongkang Pechah, where the two pearl factor1es are also located. There 1 s a 1 so a scatten ng of starch factor1 es along the Panchor Road, wh1 ch has access to clean water from the hi 11 s nearby, and f1nally 1n M1nyak Beku. The starch-pearl factor1es are 1nvar1ably s1ted by a pant to fac1l1tate d1rect management of the raw mater1a l supply. The prewar 1ndust ry was dom1 nated by wet-flour product1on to supply the S1ngapore market, and th1s pers1sted unt1l the late 1950s. Sago-meal product1on is the 1atest deve 1opment, and there 1s a spr1 nk 11 ng of such enterpr1 ses over the 01 st r1 ct. They are not subJect to the econom1 cs of locat10n, s1nce the1r 1nput 1s low, not much more than a palm da1ly, wh1ch an owner-operator w1th fam1ly help can manage. Reg1mented by the s1ze of the open dry1ng yard and labour scarc1ty, a sheltered locat1on is preferable to an open one to make dry1ng more manageable dur1ng w1ncty spells. A wet-flour factory would requ1re a modest 1nvestment of about $10,000, although a small one changed hands between relat1ves for $3,000 a few years ago; a sago-meal factory would requ1re much less, s1nce 1t needs no wash1ng facil1t1es and sed1ment1ng space. The dry-starch factor1es were estabhshed 1n most cases at the pre-1nflat1on cost of $30,000-50,000 spread over a number of years, man1fest1ng the ad hoc cap1tal1sat1on typ1cal of Chwese fam1ly enterpnses; thls f1gure should be set aga1nst the per1od1c modern1Sat1on of equ1pment s1nce the 1950s. A pearl factory was bu1lt for about $100,000 in the early 1970s, whl ch however 1 nc 1uded a useless conveyor-dehydrator chute cost1ng $30,000 1ntended to reduce labour 1nput, speed up operat1ons and produce a more hyg1en1c grade of pearl. The core structure, exclud1ng mach1nery but includ1ng all cement works, such as water-storage and starch wash1ng-settl1ng tanks, cost $20,000 then. A modern sago factory of comparable capac1ty, comb1n1ng appropr1ate modern and trad1t1onal elements, would probably be casted at ten t1mes the above values. The technology of sago process1ng has undergone dramat1c changes 1n the last decade, although the prototypes of the 102

machlnery now w1dely used was 1nvented dunng the 1nterwar era (FaHweather and Yap 1937). Veterans of the 1 ndustry reca 11 ed that the parut was used up t1ll the early 1950s, and the hlghly enJoyable JOget- or ronggeng-styl e trampl1 ng as well as manual wash1ng and paddl1ng unt1l the early 1960s; wooden conta1ners and paddles were used 1n preference to other matenals as they would not d1scolour the starch. The perfect1on of the rotary sp1ked-grater, the pr1mary contnbutor to factory eff1c1ency, revolut10n1sed the 1ndustry. (ThlS pract1cal 1nvent1on 1s credlted to Mr Chua S1o Plow, now 1n hlS 60s and l1v1ng 1n Buk1t Pas1r 1n 1979. Even sago processors from Sumatra came to Batu Pahat to purchase 1t because the expert1se and raw matenals for dupl1cat1ng such mach1nery were 1ack 1 ng 1 n Indones1a.) A compact structure, studded w1 th sp1 kes about 1.5 1nch long rad1at1ng from the cyl1nder, 1t must be well mounted, us1ng su1table metal parts, otherw1se the ent1re grat1ng apparatus Wlll wobb 1e and the rasp1 ng w1ll be 1neff1 c1 ent, besides aggravat1ng wear and tear on the support1ng and mov1ng parts. The sp1ked d1sc, used 1n Sarawak, 1s absent 1n the 01str1ct. The fo 11 ow1 ng products are manufactured by the Batu Pahat sago processors, 1n 1ncreas1ng order of technolog1cal complexlty and market value: meal, wet flour, starch and pearl. It should be noted that sago and tap1oca process1ng are 1dent1cal, except for m1nor modif1cat1ons due to the d1fferent character of the raw matenals. On occas10ns, sago starch has had marg1nally hlgher value 1n local trade because 1t has a f1ner texture more su1table for 1ncorporation 1n noodles, b1scu1ts, and other local food products, whereas tap1 oca starch 1s the supen or 1aundry starch. But the palm 1s becom1ng less compet1t1ve as a starch source and the FOB value of tap1oca products 1s r1s1ng more rap1dly than for sago products.

Processing Methods

Meal To fac1l1tate meal (ch1 leow) product10n, logs stored 1n a pond are dned out for a day or two. A pr1me 3-ft log we1 ghs up to 150 kg, 1nclud1ng 18 kg of bark, and averages 42 kg of meal wlth over 70% starch at 13% mo1 stu re after 20% waste fibre has been s1eved out. Sagoworms 1nfest1ng a log are cons1dered benef1c1al to the meal as they upgrade 1ts prote1n-fat content. Generally 103

PLATE L

Prl_.y Sago Processing above:

If used as an animal feedstuff, sago meal Is dried in an open court, then sifted before bagging. Portable roofs made of corrugated Iron protect the meal during I nclement weather.

be I ow:

If used for starch product I on, sago mea I Is fed Into two-stage drums for leaching out the starch; the first stage removes the coarse fibre, and the second (lower drum) the finer residue. The

starch mI I k

seeps out of

below (lower left); sago product I on.

the

fIne mus I In

into the trough

large quantities of clean water are used

in

PLATE M

Sligo 1111511 Ing

left:

right:

The starch milk settles In troughs before further cleansing In the clrculer tub, which Is fItted wIth paddIes; such tubs may exceed 8 ft In d Iemeter. Sego st,rch Is washed and sedlmented severel times to settle out the finest Impurities: the top layer contelns scum and light particles, while the heavier contemlnants, such as sand, sink to the bottom. Semi-wet sterch blocks after the flnel sedimentation; the starch Is stored under weter If not lmmedletely used for pearl making, or It Is dried In the open or on a heated platform.

PLATE N Sago Process lng

left:

Semi-dry starch

right:

which are sieved to ens~re uniform qual lty before drying In revolving drums. Sago pearl emerges from the final heating drum. This factory was quite labour but was cleaner, being newer.

Is pulverised (by

left heap) and then tumbled

In a rotary drum to form small Intensive with

pearls of 1.5 mm diameter, Its processing procedures,

rneal processors are less discriminate over the quality of the 1ogs supp 1 i ed, tending to use younger or maggoty pa 1ms; they probably absorb the produce from less choice plantings. Poultry is said (by several informants) to be good layers when fed on sagoworms (and 1 certainly saw ample evidence of chickens relishing it). But at $9 per pikul, sagomeal is considered expensive and has not appreciably penetrated the pig or poultry feedstuff market. The debarked log is split into about eight stout battens. Each is pushed endwise against the revolving spiked roller and pulverised into a material resembling cream-coloured sawdust which soon changes to a deeper colour. The mash is spread out on an open court and takes a day to dry during the finest weather; during rainy spells, it is dried more slowly in a well-ventilated shed, which then requires less labour than if heaped under portable roofs. Bad weather degrades the quality of the meal. The meal is sifted to remove coarse fibres, then bagged; undoubtedly sagomeal thus produced has quite a high fibre content. The total production of sagomeal or its nutritive value as a direct feed is quite unknown; the economics of production remain also quite obscure. One small enterprise with two men at work produced 10 pikul of dry meal every three days, input averaging half a palm a day (bearing in mind that this tends to be less weighty than average).

Wet f1our or crude starch One factory in the District retained the prewar trampling method of producing crude starch, but mechanical methods had started to prevail during the Korean Boom. The initial stage is identical to that of meal production. A prime log could yield nearly 60 kg of wet sago, from mash containing about 20% starch; it used to retail at $7-8 per pikul-bag during the booms.l4 Fed into a huge horizontally-set revolving cylinder wrapped with fine-mesh nylon, and employing mildly the centrifugal principle, the mash is washed to remove most of the coarse fibrous residues, and the starch milk is fed to a smaller cylinder, where it is given a more thorough wash to remove finer fibres (tapioca processing does not use the second cylinder, and the one cylinder is usually hexagonal). This stage considerably reduces bulk and may be described as sieve-washing, to produce a flour that resembles aboriginal sago that was made directly into food in other parts of Southeast Asia; it is known as chor hoon or tham hoon, sagu basah, i.e. crude starch or wet flour. Factory production efficiency is determined at the rasping stage: the more dextrous the operator (who rasps a batten in 107

20-30 seconds on average, depend1ng on 1ts s1ze and qual1ty), the h1gher the 1nput of logs, hence the greater the output of mash or meal, wh1ch controls the volume of wet flour product1on; the other cr1t1cal factor 1s sed1ment1ng and dry1ng space. The process1ng equ1pment used by sago factor1es d1ffers a llttle from that for tap1oca 1n that the ends of the cyl1nder are open and 1t 1s sllghtly t1lted to enable hampas to be d1scharged cont1nuously at the other end. In late 1980, the f1 rst s1multaneous rasper-debarker was 1nstalled 1n one of the most reputable starch factones: the battens are fed manually, ass1sted by a clamp, and the bark emerges cleaned of a 11 plth; 1og 1nput 1s at 1east doubled 1n terms of t1me, but the real sav1ngs 1s 1n labour. Debark1ng takes 1-2 m1nutes per log, manual rasp1ng another 3-5 m1nutes 1n the hands of a sk1lled worker; mechan1cal rasp1ng-debark1ng lasts about the same t1me as manual rasp1ng alone. S1nce sk1lled labour is 1ncreas1ngly d1ff1cult to come by, th1s labour-sav1ng 1nnovat1on 1s a boon to the 1ndustry. But there 1s no escape from the waste d1sposal problem because each palm generates some 8 p1kul of hampas, 1.e. almost half log 1nput becomes waste output; a daily 10-pa 1m 1nput wou 1d produce 30 tons of wet res1due weekly. Unless the log1st1cs of res1due removal are good, a mount1ng h1ll of hampas becomes a nu1sance and w1th1n a week could dwarf a small factory, apart from the unwholesomeness of 1ts putrefact1on. As 1n Chemor, one can be led by the nose to Parlt B1lal, although sago 1s not so strong as tap1oca 1n th1s respect. How and where to dump hampas econom1cally, when labour 1s so scarce, 1s another problem fac1ng the 1ndustry s1nce the slump began (by late 1980, some 1nstalled e 1evated outhouses w1 th a bucket system for hampas, where prev1 ously 1t was manually collected at floor 1eve 1 ) • Thus some factor1 es 1et an1 ma 1-rearers take away sackfuls of the hampas for a plttance, 20-50¢ per 1-2 p1kul bag, g1ve 1t away, or contract others to remove 1t w1 thout charge who 1n turn dry 1t for sa 1e as p1gfeed. Hampas, s1agor tow, s1agor poh, was once popular as a p1gfeed, and shortly before the slump reta1led at $60-70 per 2.5 ton 1nclud1ng transport; much of 1t was exported to S1ngapore. Two maJor processors found hampas d1sposal such a costly problem that they dec1ded theu enterpr1se would not outl1ve the slump, and were phas1ng out product10n, v1z. one 1n Muar D1stnct expected to cease ope rat 1on when the Town Council 1mp 1ements 1ts conservancy laws, s1nce the factory is located 1n a suburb and would be unwelcome when hous.1ng denslty 1ncreases (but 1t revlVed v1s1bly in early 1981, and a mechan1cal tub st1rrer was 1nstalled), wh1le the other, near M1nyak Beku, nelther replaced labourers who left nor modern1sed 1ts mach1nery. A large wet flour factory requ1res at least six male labourers to handle, debark, spllt and grate the logs, clear the hampas, and bag the wet flour, and some factones have great d1ff1culty ma1nta1mng 108

even th1s labour force. Wet flour factones employ women 1f they are capable of work1ng l1ke men 1n v1ew of the arduous nature of th1s work.

Dry starch Starch producers may supplement thelr own crude starch product1on w1th starch purchased from small factor1es, wh1ch they reprocess after a san1 tary r1 nse or two together w1 th the1r own. The crude starch 1s r1 nsed severa 1 t 1mes : the fl rst two washes are cons1dered the f1rst wash-sed1mentat1on, to produce the med1um-grade starch, chong hoon; up to three further r1ns1ngs are requ1red for the top grade, s1ong hoon. In the f1rst nnse, 1.e. after the mash 1s reduced 1n bulk to y1eld char hoon, a fine hampas results wh1ch 1s more nutr1t1ous than the pr1mary res1due and far less bulky. It lS read1ly sold m1xed w1th sed1mentat1on scum and contam1 nated starch to 11 vestock farmers. More than any other type of sago factory, dry starch producers need access to a clear water source, for 1t 1s one of the very few agr1cultural process1ng 1ndustr1es to requ1re enormous volumes of water to obta1n 1ts products. After the f1nal sed1mentat1on, blocks of wet starch are spread out for dry1 ng. In the case of covered courts, the flour 1s heated from below (by bo1ler steam, unl1ke wood f1bre 1n tap1oca factor1es), whlle 1n open courts slid1ng roofs shelter the flour during 1nclement weather (the latter was not seen 1n the Chemor tap1oca factor1es); strangely, bark 1s not used as a fuel source at all. The open-dry1ng system has the protect1on of portable roofs under wh1ch the flour 1s mounded, a supplementary fac1l1ty resorted to when the weather 1s expected to be calm and sunny, w1ndy weather creat1ng loss and nuisance. The mutually benef1c1al relat1onsh1p between the crude starch and dry starch producers 1 s due to two factors. Wet sago product10n has a much longer h1story than any of the other starch products, the factor1es of the early decades of this century excluslVely produc1ng th1s for the S1ngapore manufacturers, to whom wet sago was sent 1n sacks, the very nature of the product and its t1ght pack1ng coinc1dentally ensuring that the flour would not deteriorate aunng the short boat JOurney. Towards the late 1950s, dry starch product1on in the D1str1ct emerged as an 1mportant part of the 1ndustry, after the S1ngapore ref1nenes, wh1ch formerly monopol1sed th1s type of manufactur1ng, largely ceased operat1on, and now 1s the dom1nant component in the Dl Strl ct. Sl nee the better qua llty starch requ 1 res abundant supp 11 es 109

of c 1ean water, a wet sago factory must re 1ocate when a r1 ver or dra1n 1S no longer as clean as before (the s1tuat1on w1ll worsen when palm 011 factones abound). Yet to 1nvest 1n water p1pes and storage tanks requ1res capital 1nvestment beyond the means of most small entrepreneurs. Some starch processors would prefer not to purchase wet flour from smaller enterpr1ses because 1t 1s unclean, but they cannot produce enough of the1 r own and must rely on the latter to overcome the shortfall. Second, some processors cannot expand, s1nce a factory produc1 ng dry starch requ1res not only tw1 ce as much cap1 tal as a crude flour establ1shment, but also a cons1derable reserve of palmer1es. Even where cap1tal 1s ava1lable, a processor may be deterred from 1nvest1ng 1n 1ncreased capaclty because of other cons1derat10ns, such as labour problems, l1m1ted factory space, 1nsuff1c1ent land for plant1ng or 1nsecur1ty about external suppl1es. By the very nature of the enterpr1se, there 1s a l1m1t to expans 10n because the eff1 c1 ency and capaclty of the type of mach1nery used are severely l1m1ted, and a factory could grow to an unhea 1thy phys 1ca 1 sHe and shape, wh1l e dependent on cottage-type low-eff1c1ency equ1pment. The larger starch producers who manufacture a h1gh-qual1ty starch have access to "h1ll water", and have amenit1 es such as tubs and sed1 ment 1ng tanks 1n wh1 ch the starch m1l k 1 s sett 1ed and the 1mpur1t1es dra1ned away. Produc1ng starch w1th Ch1nese technology 1s a labor1ous, cumbersome procedure, and the tanks take up much space w1th1n the factory, but 1t enables the production of var1ous grades of starch. The coarsest is sed 1men ted 1n the f1 rst tank or at the 1owest 1eve l and the f1nest 1n the last sed1mentat1on or topmost layer after the scum 1s scraped off. Automated mach1 nery 1 s ava1l ab 1e to d1 spense w1 th the wash1ng, paddl1ng and sed1ment1ng, so that a modermsed factory could do w1thout any sed1ment1ng tank. This more product1ve method was a 1 ready adopted at the second and subsequent separat1on stage by several Chemor tap1oca factor1es by the early 1970s (one of wh1ch d1spensed w1th all such tanks), but in th1s D1str1ct was apparently unknown even 1n 1979. Th1s would 1nd1cate that the two groups of d1screte starch producers 1n the pen1nsula operate lndependently, except for some absorpt1on of Batu Pahat sago starch for spec1al customers by the Chemor-assoc 1a ted Penang export 1ng agents (wlth Tha1 contacts), some of whom are dlrectly 1nvolved 1n tap1oca product1on as well and whose assoc1at1on w1th Perak 1s far more act1ve and direct. By contrast, the Batu Pahat sago producers have been left bereft of s1m1lar st1mul1 by the economic and po1Tt1cal separation from the1r port of trade, S1ngapore. 110

Pearl Pellet1ng 1s the most labour-1ntens1ve process 1n the sago manufactun ng system, and the methods adopted seem 1ess eff1 c1 ent than the same process 1n the tap10ca 1ndustry. Sem1wet starch 1s mllled and fed 1nto tw1n revolvmg drums, dur1ng wh1ch a mass of Hregularly-slZed small pearls are formed. These are s1eved 1n a gyrat1ng screen, after wh1ch the standard-s1zed pearls are fed 1nto a heated open cyl1nder; these pearls are graded aga1n to ensure un1form qual1ty (1.5 mm) and further dned, 1n trays 1n a heated chamber (1n late 1980, one factory replaced the mult1-tray chamber-heater w1 th another set of heated cyl1 nders to speed up product10n). After each of the heat1ng stages, a s1zeable amount of reJected mater1 a 1 has to be rem1ll ed and reprocessed. The wastefulness 1s str1k1ng.

Processing Efficiency A veteran of the sago 1ndustry est1mated that the monthly max1mum process1ng capac1ty of the D1str1ct's rema1n1ng factor1es 1s under 1,000 tons of dry starch at max1mum eff1c1ency; were they to operate on 24-hour sh1 fts, th1 s capaclty m1 ght be 1ncreased by 50% at the most, largely because of llmlted sed1ment1ng-dry1ng fac1l1t1es and the stam1na of the ava1lable labour force. In the heated platform system, 50 p1kuls (3 tons) wet starch normally take 2-3 days to dry and clear away. The largest factory 1s sa1d to be 1ncapable of produc1ng more than 150 tons of dry starch a month, 1f that. Iron1cally, the general compla1nt was that the processors found d1ff1culty 1n market1ng even the1r l1m1ted volume of product1on, and they were baffled by 1t. The s1mple explanat1on 1s that th1s volume 1s too small to ent1ce a fore1gn 1mporter who has to keep hlghly cap1tal1sed equ1pment work1ng at a very h1gh throughput of raw matenals; thus, several years ago, a Japanese scout1ng team fa1led to acqu1re the1r m1n1mum quota of 4,000 tons of starch a month from the Batu Pahat sago producers. The paradox 1 s that the 1ndust ry must expand product 10n cons1derably desp1te the poor marketab1l1ty of current product 1on, 1 f 1t 1s to make any 1n roads 1n the 1nternat 10na 1 market; otherw1se 1t w1ll revert to be1ng a cottage 1ndustry. Accord1ng to the processors, 1962-65 and 1971-74 were the best years for the 1ndustry (Table 16). Bes1des the low stem pnces, they have had to contend w1th ns1ng labour costs s1nce then. Wages, once $4 a day, had r1 sen by 1ate 1979 to over $10, and up to $15 for harvesters, the most pr1zed of all workers. Wh1le most processors

Vl

ewed the current slump as the worst

111

TABLE 16 Cost of MIJ1"erlals In Sago Factories In Mlllaysla Bara1" • 1968-1973

Component ($/tonne)

1968

1969

1970

1972

1973

Stem

31.25

29.74

30.24

22.42

33.65

Wet sago

77.95

63.00

90.22

76.71

98.65

201.77

227.14

236.21

230.46

279.13

Input

Output Sago flour

NOTE.

SOURCE:

1968-72 values are computed from prices given In pi kul; 1973 Is as pub I I shed. Dep. Statistics. ~and Survey of Manufacturing Industries (relevant Issues). Cf Appendix VI I I.

1n l1v1ng memory, some bel1eved that, g1ven past trends, the market should perk up. But 1t 1s a race against t1me, s1nce other crops are now far more attract1ve. Thus the decl1ne 1n the sago 1ndustry may be 1rrevers1ble. When I rev1s1ted the 01str1ct towards the end of 1979, the sago market had started to recover, and one processor feared that he m1ght not f1nd enough palms; by early 1981, 1t had become v1s1bly rev1tal1sed. The pr1ce of the palm had 1mproved stead1ly and was cons1dered rather h1gh because of the general fear that stem suppl1es would become even more compet1t1ve when more farmers remove the1r sago after harvest1ng the mature stems, to make way for the o1lpalm. Only loyal sago farmers and those w1th land st1ll l1able to pers1stent 1nundat10n and brack 1sh 1nt rus 1on cou 1d assure factory operators of future supp 11 es. If these farmers mal nta1 n the1r 1nvestment, they w1ll be the only supplementary source of palms left when the market recovers fully. In a way, the slump w1ll compel the process1ng sector to con so 11date 1nto a core of more eff1 c1 ent enterpr1 ses, 1f 1t does not pers1st for too long, and would be the f1rst step towards techn1cal 1mprovements 1n process1ng and coord1nated market 1ng. each

Accord1ng to the larger starch factones 1n the D1str1ct, uses 80-100 logs da1ly; th1s 1nput perta1ns 1n a slump

112

market and 1s probably about half of factory capac1ty; Wahby et al. (1970) assessed the average capaclty of the Sarawak factor1es at 7.5 palms/day (cf. 75 tons tap1 oca tubers 1n early 1981 1n the largest Chemor factory). Even 1n the marg1nally 1mproved Batu Pahat factones (1n compar1son w1th the Sarawak1an), 1nput-output rat1os are low and should be set aga1nst the ad hoc nature of Ch 1nese commerc 1a 1 enterpr1 ses of thl s type, 1nc 1ud1 ng the 1ack of qual1ty control and the use of ant1quated mach1nery and vanable raw matenals. But econom1c eff1c1ency should be real1st1cally set aga1nst other factors of product1on, pnnc1pally labour shortage, transport problems and the need to tap the peasantry for raw matenal suppl1es. Wlth centr1fug1ng, product10n could be doubled Wlthout 1ncreas1ng manpower or the huge area requ1red under current methods of sed1ment1ng and dry1ng; s1nce flash-dry1ng 1s vert1cal, much of the space already ava1lable becomes redundant. Some examples of output g1ven are set out 1n Table 17, w1th y1elds based on actual operat1ons, and assum1ng 10 logs per 1-tonne palm; wet sago (Grade III flour) dr1es down to 28-36% dry commerc1al starch (assumed 13% mo1sture), w1th extract1on rate at 108.9 kg/palm (1.8 p1kul) of commerc1al starch, except1ng {4) where product1v1ty 1s almost double. The sol1tary sago factory 1n Perak I v1s1ted 1n 1971 obta1ned 1.3 plkuljpalm of commerc1al starch, 1.e. the waterfree starch equ1valent 1s 78.6 kg/palm (Tan 1973). The fo 11 ow1 ng assessment of the qua 11 ty of Batu Pahat sago product1on 1s based on factory product1on publ1shed by the Department of Stat1st1cs {Table 18), Whlch 1n 1973 recorded an 1nput of 28,773 palm-tonnes and 7,437 tonnes of wet sago, w1th dry (commerc1al) starch output of 5,785 tonnes. Total 1nput equ1valent therefore was 47,366-53,563 palm-tonnes, based on a wet sago content of 30-40% (calculated from Table 17), as follows: (1)

output of 30% wet sago content would be from 53,563 palm-tonnes: average 94 kg/palm of waterfree starch;

(2)

output of 40% wet sago content would be from 4 7, 366 palm-tonnes: average 106 kg/palm of waterfree starch.

S1 nee factory output was equ 1va 1ent to 5, 033 tonnes ( 5, 785 tonnes at 13% mo1sture), the waterfree starch content of the Batu Pahat sagopalm 1n 1973 was 94-106 kg (cf. 1-2 plkul/log (Wahby et al. 1970), and 0.1-0.2 tonne/palm {Flach 1971) 1n Sarawak). The Department of Stat1st1cs elsewhere recorded nat1onal product10n as 5,920 tonnes dry starch and 929 tonnes of exported pearl (0% mo1sture) (Table 13; Appendlx VIII); thls 1s equlValent 113

TABLE 17 ExtrectiCIII Efficiencies In Sago Fac:tarles

Wet Flour Dry Starch plkul/palm

Starch Grade

111 II

5-6.5 3,5-4,5

2,2-2,7 1,8

3

Starch Grade

Ill

Wet Dry $/pikul (b)

Description of Manufacture

Wet sago or crude starch, sagu basah, char hoon, tham hoon, one sedimentatlon, but It provides an Inkling of the relative operations of the larger enterprises of both Industries covered by the Census. v. Table 14.

* SOURCE:

- 1,938 -

Cf. Note 5. Malaysia, Department of Stat1stlcs. Industries 1973; Kuala Lumpur,

115

Census of Manufacturing

to 6,079 tonnes of waterfree starch. Extract1on from the above palm-tonnage at th1s product1on volume based on 30% wet sago content y1elds 113.5 kg, and on 40% y1elds 128.3 kg, waterfree starch. If factory 1nput-output were based on wet sago-tonnage calculat1ons, the starch content range 1s 94-126 kg/palm; the results correspond closely w1 th the above output of 5,033 tonnes and production of 5,920 tonnes of waterfree starch (wet sago dr1es down to 31.3% waterfree starch, based on the more eff1c1ent rate of 36% commerc1al starch), as follows: ( 1)

output of 30% wet sago content y1 e 1ds 5, 030 tonnes of waterfree starch: 94 kg/palm;

(2)

output of 40% wet sago content y1elds 5,930 tonnes of waterfree starch: 126 kg/palm.

have est1mated are That factory output eff1c1enc1es exceeded by nat1onal product1on rates 1s expected, cons1dering the 1ncomp l ete coverage wh1 ch 1gnored the smaller, 1ess eff1 c1 ent enterpr1ses. While starch product1on may be presumed to be comprehens1vely covered, s1nce th1s 1s only produced by the larger factones, wet sago product1on cannot be since 1t also comes from the smaller enterpr1ses, and these have escaped the Survey of Manufactur1ng Industr1es; hence nat1onal product1on data would ind1cate h1gher-eff1ciency product1on. Accord1ng to the market s1tuat1on (Table 16), commerc1al starch 1s almost thr1ce the value of wet sago, wh1le stem 1s 10-16% of starch value at a favourable per1od of the 1ndustry; rate of sago extract 10n has to be h1 gher than these values to be econom1 c. The Sarawak rate of extractiOn 1s sa1d to be 40% lementa. The d1 fference 1n the computat 1on suggests that the factory extraction rates (given 1n Table 17) could not be extrapolated to overall product10n in the 1ndustry; 1t 1s diff1cult to weight the data, but on balance, the higher eff1c1ency values computed ought to 1nd1 cate the 1ower rather than the correspond1 ng extract 10n eff1c1enc1es 1n the sago 1ndustry. Furthermore, the follow1ng factor should be cons1dered, d1scount1ng any gross 1rregular1ty 1n mon1tor1ng product1on volumes: there was cons1derable wastage 1n the Industry, espec1ally 1n smaller factor1es due to manual handling 1n the 1ntermed1ary stages of manufacturtng, but part of 1t was deliberate 1n order to produce a starch1er hampas as a h1ghly valued animal feed then worth $20/tonne locally and $19/tonne exported (Append1x VIII). If 128.3 kg/palm was the best extract1on (at 40% wet sago) apparently obta1nable under pract1cal cond1tions, thiS rate could be raised by up to 25% to over 160 kg/palm, assum1ng that the eff1 ci ency of extract 1on normally d1 d not exceed 75% (as seemed the case at a parallel stage of technology 1n the tapioca 116

1ndust ry), 1f measures are taken to reduce waste to a negl1 g1 b1e level. I noted that most small wet sago factones gave product1on volumes that suggested less than 10% starch was obta1ned from palm 1nput, compared w1th the 13% average of a pearl enterpr1se. One m1ght assume that unl1ke the larger factor1es wh1ch are more spec1al1sed, the smaller factor1es d1v1de the1r energ1es to produc1ng e1ther wet sago or hampas more eff1c1ently depend1ng on the1r relat1ve pr1ces, so that the1r starch extract1on eff1c1ency tends to be even lower. In an organ1sed plantat1on system where the harvest1ng cycles can be log1st1cally and b1olog1cally controlled, and where mechan1sed operat10ns could reduce wastage, 1t m1ght be poss1ble to almost obtaln the 185 kg/palm Flach (1977) cons1dered atta1nable based on h1s expert observat10ns 1n the D1stnct in the early 1970s. Thus, wh1le 1nput (4) 1s apparently anomalous (Table 17), 1t 1s obta1ned 1n a large, probably more eff1c1ent factory ab 1e to cut wastage by 50%. A fortultous comb1 nation of two cruc1al factors of product1on cannot be ruled out e1ther, v1z. a regular supply of larger or starch1er palms from cho1ce plant1ngs and more eff1c1ent or better ma1nta1ned mach1nery enabl1ng thorough removal of the starchy matter 1n these palms. If 544.3 kg/palm (9 p1kul) 1s 40% wet sago content of a palm, pa 1m we1 ght would be nearly 1.4 tonnes; thus, waterfree starch extract1on at 31.3% of wet sago 1s 170.4 kg/palm, or 12.5% of palm we1ght. The above computat 1ons prov1 de useful bas 1c 1nformat 10n: the extant p1ant at 10n genet 1c materld 1 and techno 1ogy 1n Batu Pahat 01 st r1 ct generally y1 e 1d 13% waterfree starch by we1 ght from the sagopalrn. The two methods of calculat1on help to establlsh the lower but not the upper l1m1ts of palm productiv1ty 1n Batu Pahat, and afford some measure of the techn1cal eff1c1ency of the plantat1on 1ndustry here. H1gher product1v1ty can be expected w1 th 1mproved product 10n techm ques, even JUSt us 1ng the ava1lable plant1ng mater1al.

NOTES 11

Salinity effects on crops are not considered by farmers to be lnvarla~ly I nv I d I ous • coconut growers In Ser I Men ant I (near Muar) a I I owed sel;lwater Into their gardens but not overnight, according to a well-tried sch!!.dulehlnged on tidal levels; this was believed to Increase nut yields, but a local official scotted at It CAO JN 109/53), ot course. In the Philippines, sa lin lty Increased In land coconut yields, due to the chlorine enabling the

117

palm

to

absorb

12

nutrients

1110re

Agric. J.12/11, 1977:

Husband.

Nonetheless, the raw material

readily,

thus

speeding

up

growth

(Animal

10). supply

IS

not as casual

or speculative as the

team of FAO e>