Civilisations Revue internationale d'anthropologie et de sciences humaines 44 | 1997 Les peuples des forêts tropicales

The human consequences of deforestation in the Moluccas

Roy Ellen

Electronic version URL: http://journals.openedition.org/civilisations/1628 DOI: 10.4000/civilisations.1628 ISSN: 2032-0442

Publisher Institut de sociologie de l'Université Libre de Bruxelles

Printed version Date of publication: 1 January 1997 Number of pages: 176-193 ISBN: 2-87263-122-4 ISSN: 0009-8140

Electronic reference Roy Ellen, « The human consequences of deforestation in the Moluccas », Civilisations [Online], 44 | 1997, Online since 29 June 2009, connection on 19 April 2019. URL : http:// journals.openedition.org/civilisations/1628 ; DOI : 10.4000/civilisations.1628

© Tous droits réservés THE HUMAN CONSEQUENCES OF DEFORESTATION IN THE MOLUCCAS

Roy ELLEN INTRODUCTION posing a danger t o existin g fores t an d fores t Compared with other part s of island sou­ peoples: swidden cultivation, plantatio n crop ­ theast Asia, little is known of either the forests of ping, commercial loggin g and migratory lan d the Moluccas (map 1) , o f indigenous patterns of settlement. Usin g as an example the Nuaulu of forest use , or of the threats pose d to both forest Seram, I illustrate ho w these factors interact in a and people by increasing rates of deforestation. In particular instance , as well a s the various phases this paper 1 attemp t to describe the effects of defo­ which typify a peoples exposure and response to, restation o n th e live s of th e loca l population , first, denudation, and then widespread degrada ­ using the small number of reports which are avai­ tion of the forest environment. I indicate that the lable. I begin by assessing the historical huma n phasing and character of these responses depend impact o n the forests of these islands, stressing very much on local perception s of government both th e varied patterns of sustainable accom­ policy and on the extent to which policy is inter­ modation reache d between people and forest, and preted by officials an d translated into action. The the fact that forest a s presently constituted i s the effects of policy vary between different part s of the outcome o f co­evolutionary processes of which Moluccas and different populatio n groups , but 1 humans themselves are an integral part . I then suggest that we can expect some convergence a s examine th e mai n factor s repeatedl y cite d a s the forested areas diminish i n size.

Map 1. The Molucca n island s of eastern Indonesia , illustrating principle place s mentioned i n th e text. B y the Moluccas , I understand here those islands lying within the boundary of the present­day Indonesian province of Maluku. The numbered locations ar e as follows : 1 . Wahai; 2. Pasahari; 3 - Bula; 4. Masiwan g area; 5. Mane o and Seti; 6. Tehoru; 7. Tamilouw an d Yalahatan; 8. Sep a area (includes villages of Rohua, Bunara, Watane, Hahuwalan, and formerly, Aihisuru); 9 Amahai ; 10 . Makariki; 11 . Ruata n transmigration are a (including Simalouw); 12 . Kairatu. The area on Seram bounded by a thick lin e shows the position of the Manusela National Park . HUMAN IMPACT ON THE FORESTS in upland central and west Seram. Indeed, from a [1] OF THE MOLUCCAS BEFORE 1900 scientific point of view, the Moluccas is one of the The ecology of insular southeas t Asia has bee n few places in the Indonesian archipelago where it dominated by rainforest fo r over 10,000 years, is possible to find a complete altitudinal sequence though i t has changed much historically an d is of vegetation, and there are few places elsewhere very varie d geographically . On e of th e mos t in the tropics which provide a comparable range immediately strikin g aspects of its variability i s [Edwards, 1993 : 3]. Although there have been a the significant decreas e in Dipterocarp species a s few surveys on Halmahera and Seram, there has we move east and their replacement by dominants been relatively little quantitative stud y of Moluc­ more typica l o f the Australo­Melanesian area . can rainforest [Edward s et al, 1993:6.3]. Thus, the forest biogeography of the Moluccas dif­ There are, however, many ecological simila­ fers from that associated with the classic Diptero­ rities between Moluccan forests and those further carp forests, of say Borneo or Sumatra, in several west in island southeast Asia. Not the least of these features of its structure and composition, resem­ has been the role played by human populations. bling much more Melanesia [Edwards et al, 1993; Forests have long been a focus of human subsis­ Edwards, 1993; Ellen, 1985 : 560-3]. I t is this tence extraction, an d huma n agenc y has had transitional (Wallacean ) character that makes it decisive consequence s for thei r ecology , for of special interest. On Seram, for example, there example, through the introduction an d hunting are possibly just two species of Dipterocarp [Sho­ of deer, the practice of small­scale swidde n culti­ rea selanica an d one other), compared with 300 vation, the extraction of palm sago and selective species on Borneo; there i s just on e Eucalyp t logging and collection for exchange [Ellen, 1985, (Eucalyptus deglupta) compare d with 450 in 1987]. The early history of Moluccan forests in Australia [Edwards , 1993 : 5]. I n addition , human term s i s poorly understood , with little although mos t of the primary lowlan d forest is of empirical researc h which would shed direct light the moist evergreen type, displaying little seasona­ on the subject. From work elsewher e in insular lity, in places (most prominently, th e west part of southeast Asia, the evidence for human impac t Yamdena and south Aru) we find semi­dry mon­ from 8000 B P onwards has been demonstrated, soon and savanna forest [SKEPHI, 1992 : 23; van and althoug h w e would not expect this time­ Steenis, 1958], an d patches of semi­evergreen depth fo r the Moluccas, we should anticipat e forest on other island s (especiall y Halmahera, chronologies in terms of thousands rather tha n north Buru and Seram). Along the coasts there hundreds of years. The sub­fossil and palynologi­ are some significant area s of mangrove (e.g. the cal evidence in question usually comprises signs Apu estuary on Bum, east and north Seram, Aru). of anthropogenic burnin g an d changing species In the low­lying valleys of the larger island s are composition reflectin g patterns of clearance, cul­ extensive areas of Metroxybn (sagopalm ) swamp tivation an d seed dispersal [Maloney , 1993]. N o forest, while montain forest is found on Bum and doubt similar data will eventually be forthcoming for the Moluccas, but despite prehistoric and his­ historically significant elsewhere; dry rice in parts toric modification, large tracts of Moluccan forest of Halmahera since around 1500 , and formerly have remained more or less intact until relatively Coix, Cenchrus, mille t (Setaria italica), an d recently on the larger islands : that is on Halma­ Sorghum [Visser , 1989]. Millet i s also important hera, Seram, Buru, Yamdena and Sula. This has in parts of the Kei islands, and Coix and dry rice been due to low indigenous population levels, the on a small scal e more widely [e.g . Ellen, 1973 : concentration of the existing population in more 460; Seran , 1922] . Musa (plantain s and bana­ accessible centres and along coasts, general eco­ nas) ar e grown almos t everywhere . Sinc e the nomic peripherality and low in­migration. seventeenth century, however, many of these cul­ At the present time, Moluccan populations tigens have been outstripped in importanc e by exhibit a variety of subsistence strategies focused introduced maiz e (particularl y in th e drie r on differing degrees of forest modification an d south), manioc (throughout , but especially on clearance. Though these patterns of extraction Kei), Xunthosoma (i n wette r areas) and swee t are often associated with separate types of people, potato. Rice is now grown more widely (particu­ linguistically, geneticall y an d in term s of eco­ larly by migrants in both the northern and cen­ nomy, the facts suggest that these distinctions are tral Moluccas), and in irrigated fields, but appa­ not hard and fast ones. At one end of the spec­ rently not with a great deal of success. trum of techniques are peoples such as the Tugu­ Apart from the impact of these modes of sub­ til o f central Halmaher a who ar e engaged in sistence, the main changes to Moluccan forest eco­ nomadic hunting an d gathering, but with some logy that we can be sure of historically are associa­ planting and reliance o n trad e [Martodirdjo , ted with the growth of regional exchange systems 1988 : 15]. On Seram there i s a wide variety of linked to outside trade in forest products. Dammar combinations of technique, ranging from mainly or copal (Agathis dammara) resi n has been extra­ hunting an d gathering with little cultivatio n cted on Seran [Ormeling, 1947], Morotai [Riem, (Huaulu, Maneo), through classic forms of swid­ 1913 (1909)], Halmahera [Giel, 1935-6], Bacan den agriculture [Ellen, 1978], to more intensive [Korn, 1916], Obi [Ham, 1911] an d elsewhere for forms o f permanent agriculture on the coast. The centuries, and involves little destruction of trees. common characteristic of all thes e is the pivotal Traditional dammar tapping has recently declined role played by the extraction of and dependence and been replaced by commercial exploitation i n on sago [Ellen, 1979,1988], which has the effect some areas [Edwards, 1993:8-9]. Much the same of minimizing the amount of rainforest cut. Crop may be said for the oil of Melaleuca (cajuputi = regimes vary partly i n relation t o the contribu­ leucodeudron) o n Buru and west Seram, repor­ tion made by sago. Tuberous starch staples such ted as early as 1855 [Schmid, 191 4 ; van der Crab, as yams and taro have probably been importan t 1862], production of which, however, continues in man y area s for thousand s of years, an d in to rise [Kantor Statistik Provins i Maluku, 1989]. some parts continue to be so. Grains have bee n Of lesse r importance ar e beeswax, kapok flos s (Ceiba petandra), charcoal , and gaharu resin with th e decreas e in deman d for spice s in th e (poss. Aquilatia) use d for incense and known to eighteenth century, the Moluccas became a com­ be collected in central Seram. But of the non­tim­ mercial backwater , and this afforde d som e pro­ ber plant products, the most commercially impor ­ tection to its forests. The nineteenth century saw tant in bulk tenms has been rattan [Kanto r Sta­ an upswing in the extraction of non­timber forest tistik Provinsi Maluku, 1989]. Timber itself has products for th e European and Asian markets , been extracted for export from before European and the first significant commercial loggin g acti­ arrival, mainly for boatbuilding an d fuel [Ellen, vity on Seram [Ellen, 198 5 : 584]. I t i s reported 1985, 198 7 : 40­1]. Fores t has been additionally that most of the forests of the Kei islands were modified through introductions, both of domesti­ clear­felled by a Dutch company in or before 1888 cates an d accidentals , throug h th e deliberat e [SKEPHI, 1992:25]. planting of non­endemic non­domesticates, such as Tectona grandis [Ellen , 1987], and through SWIDDEN CULTIVATION the inadvertent dispersal of seeds from such use­ Despite repeated assertions to the contrary ful tree s as Canarium indicum. However , the [e.g. Jones, 1976], forest­fallow swidde n cultiva­ most important singl e factor affecting Moluccan tion practised by indigenous populations is a pro­ forests during the early period was the spice trade. ductive us e of forests, more so than commercia l Early extraction may reasonably be presumed to logging in terms of size of population supported; have been of non­domesticated varieties of clove and on the whole it i s ecologically efficient i n its and nutmeg, and wild nutmeg ha s continued to rapid recycling of regrowth. Indeed, it ha s bee n be of significance in some parts of the Moluccas argued that the view of swidden agriculture a s a and coastal Irian Java. The sustained and gro­ destroyer of valuable resources and the creator of wing demand for spices, both in Europ e and in grass wastelands is simply a convenient ideology Asia, led to the appearance of the domesticate d to legitimat e commercial extractio n i n area s varieties of commerce and their systematic plan­ where the indigenous population ha s become an ting in particular area s [Ellen, 1979]: clove first obstacle [Dove, 1983b : 195], or a justification fo r on Ternate, Tidore and latterly Seram , Ambo n introducing large numbers of transmigrants with and the Lease islands, and nutmeg always focally their supposedly more «advanced» sedentary sys­ on Banda , but les s intensively elsewhere . Th e tems [Potter, 1991:188]. One of the virtues of the consequences of this development ar e taken up term «swidden» i s that it highlights th e crucial below. difference betwee n this kind of agricultural sys ­ The extraction of forest products for subsis­ tem an d unplanne d naiv e pionee r slas h an d tence and trade increased during the Dutch period, burn, with which it i s often confused. from the early sixteen­hundreds onwards. In the Although Nuaulu in central Seram expressed first plac e this reflected Dutch pressure to mono­ a preference for cutting mature forest during the polize and maximize spice production. However , early nineteen­seventies [Dove, 1983a: 133] , the amount cut during that period seldom exceede d PLANTATION CHOPPING 30 percent of forest cut in an y one year [Ellen, We have already noted that the establish­ 1978 : 83­4]. Overall, 90 percent of swidden land ment of nutmeg and clove groves occurred very in us e during the same period had been cut from early in Moluccan history. Eve n by the sixteenth secondary regrowt h o r bambo o scru b [Ellen , century this had resulted in sever e depletion of 1993a: 200]. This is high by comparative southeast rainforest on the small spice­producing islands of Asian standards, but probably a consequence of Ternate, Tidore and Banda. Heavy denudation on recent population growth , and well within the small islands led to the development of zones of parameters o f what i s sustainable. I n addition , inter­island trade in which small islands became technical constraint s on cutting (the us e of axe dependent on the forests of the larger islands for and bushknife) hav e effectively restricte d th e basic resources, such a s sago, fuel, timber an d amount o f forest conversion for subsistence far­ thatch [Ellen , 1979]. The rise in internationa l ming, leaving aside large trees such as Koompa- demand for spices led to the spread of production sia excelsa. Clearing for agriculture ha s in some in th e central Moluccas , and under th e Dutch areas (e.g around Wahai on Seram) left tracts of monopoly clov e productio n wa s focuse d o n Imperata cylindrica [Edwards , 1993 : 8], some­ Ambon­Lease, and native grove s planted on the times with Melaleuca leucodendron, bu t in most Hoamoal peninsula, western Seram and elsew- cases grassland succession is either due to histo­ here extirpated [Ellen , 1987 : 41­3]. Nutmeg pro­ ric clearance for plantations, logging , or a result duction after 162 1 focuse d on the Banda islands of reducing fallow periods as a means of intensi­ under a Dutch settler regime . The demise of the fying agriculture t o compensate for demographic Moluccan spice trade from the early eighteent h stress. The incursions of non­indigenous settlers century onwards mitigated the ecological conse­ [Dove, 1983 b : 90­3], entrepreneurial shiftin g quences of further clearance for plantations . cultivators, or what Secret t [Secrett , 1986 ] ha s In the present century there ha s been rene­ appositely called «shifted» rathe r than shifting wed clearance : for clove, nutmeg and other tree cultivators, ar e generally linked in the Moluccas crops, such as cocoanut, cacao and coffee, and to the expansion of transmigration settlemen t also for fast­growin g pul p trees[2]. Initially , this into surrounding areas, and the planting of cash was in the form of small groves adjacent existing crops [Dove, 1993 : 19; Vayda, 1981]. There is no subsistence plots , an d wa s sustainable in th e doubt that rapid forest conversation of this kind is context of an overall low­intensity agricultura l damaging, an d that long­standing sustainabl e regime. Indeed , groves continue t o be intercrop­ practices are being eroded by technological inno­ ped with various swidden species. However, since vation (e.g. chain­saws), population pressure and the seventies, and with government encourage­ market forces, but the inevitable conclusion that ment, indigenous swiddeners , and late transmi­ all indigenou s system s are to b e discouraged is grants, have cut increasingly large areas for plan­ ignorant and counter­productive. tations, by preference from primary forest [Ellen, 1993b; Grzimek , 1992]. U p until th e massiv e much dead wood, serious gully erosion, soil com­ influx of transmigrants over the last decade or so , paction, herbaceou s vegetation an d extensive this expansion in cutting represented the main areas of secondary regrowth trees such a s Maca- threat t o forest. Commercial estate plantation s ranga. I n this same area Shorea ha s bee n strip­ have recently become important i n som e areas, ped from ridges, replaced by the invasive grass such as on Seram (e.g. cacao in Wahai sub­dis­ Imperata, leavin g patches of mixed evergreen trict), o n Halmahera (cinnamon) an d on Yam ­ forest in th e valley s [Edwards, 1993:9]. Loggin g dena, where the government has granted conces­ is a particularly seriou s threat in th e area where sioners 30,000 hectares. the Manusel a National Par k meet s th e Sama l transmigration area , and failed transmigrant s COMMERCIAL LOGGING may move further into the forest [Ibid. ,11]. Log ­ Of th e standin g stock of major timber species ging road s in wes t Seram have opened­up the in non­plantation forest s throughout Indonesi a interior t o further settlemen t b y migrants an d only six percent is in the Moluccas, though this is many transmigrants i n the Pasahari area of north probably to underestimate th e timber potential . Seram rely on logging a s their main source of Indeed, Moluccan timber production ha s recentl y income. Remova l of forest (thoug h not entirely increased in importanc e [Potter , 199 1 : 179]. In because of logging) ha s also resulted in water June 198 9 there were 2 4 official forest concessions shortages in som e parts of the Wahai area, with throughout th e province, representing 2,593.000 knock­on health problems, and in seasonal floo­ hectares, the average size being 108,000 hectares. ding and water­shartages in th e transmigratio n Only the Kalimantan province s have more ply­ areas of Waihatu an d Waimital in wes t Seram. wood factories an d production capacit y [Ibid, Further east, south of Bula, in th e area of the 202,207]. Masiwang river, local transmigration ha s follo­ Logging is a threat t o forest and an indige­ wed logging an d som e replanting , mainl y i t nous way of lif e in a number o f part s o f th e would seem of cacao and som e timber trees . Log­ Moluccas, not least because foresTs contain good ging in this area continues. quality Shorea (meranti) [Edward s et. al. 1993: On Yamden a logging is posing serious pro­ 68]. What makes Moluccan forests more vulne­ blems fo r biodiversit y loss , threatenin g rar e rable i s the small surfac e are a of mos t o f th e orchids such as Dendrobium pbalaenopsis an d islands. Already, islands a s smal l a s Taliab u and several species of endemic bird. The mos t serious Mongole in the Sula group, and Morotai, Bacan, impacts, however , will be human, especially the Obi, Kasiruta , Mandiuli aroun d Halmahera have effect o n water supply . In 199 2 loggin g road s been opened­up to systematic felling. Eve n selec­ alone had cause d erosion in watershed areas and tive logging has bee n shown to cause considerable had contaminate d streams with silt. Socially, dis­ damage. After 15 year s forest in parts of north cen­ ruption such as th e destruction o f sacre d sites, has tral Sera m i s still lef t with a n ope n structure, not been compensated for by employment pros­ pects, the company recruiting onl y a few local requirements t o control fractious native s led to people. The Association of Tanimbar Village Lea­ the emptying of large areas of the highlands and ders has filed complaints to the local government interiors o f the main islands, in particular o n [SKEPHI, 199 2 : 24]. On Seram, timber extrac­ Seram. This threatened forest in coastal areas but tion has been perceived by some locals as having led to reduced hunting i n th e highlands. Redu­ beneficial effects: discarded sawn timber an d log ced hunting increase d animal densit y (particu­ ends are used a s fuel and in manufacturing, lum­ larly deer), with consequent severe browsing pres­ ber camp debris and leftover facilities provide a sure [Edwards, 1993 : 10; Ellen, 1993 a : 201]. range o f materials , whil e trackway s serv e t o Ironically, villagers in th e central highland s of enhance hunting an d communication [Ellen , Seram now register a reduction in the availability 1985; Ellen, 1993b; 133]. It is likely, though, that of gam e animals, cause d by disturbances to the the increasing scale of logging will modif y th e edge of the Manusela National Par k [Edwards, balance of advantages and disadvantages in th e 1993 : 11]. perceptions of local inhabitants (se e below). For many centuries the Moluccas have addi­ tionally been the destination for migrants moving POPULATION MOVEMENT AND TRANSMIGRATIONeast from south and southeast Sulawesi : Bugi­ Population movemen t ha s probably been nese, and more recently , and in large numbers, having an effect on patterns of deforestation i n Butonese. Over the last 1 5 years in­migration in the Moluccas for a s long a s these islands have certain area s has increased dramatically, partl y been inhabited b y humans. In th e cas e of the through direct government­sponsore d transmi­ smaller islands this would account for their early gration an d partly throug h increase d sponta­ depletion. Since 1600, however, we have clear evi­ neous migration mad e more attractive b y new dence fo r deforestatio n throug h relocation s infrastructures suc h a s roads, including thos e within the sam e island, inter­island migratio n created in the first place for government transmi­ within th e Molucca s an d in­migration fro m grants. without. Th e Moluccan­European wars of th e The Moluccas was first incorporated into the seventeenth century resulted (as we have seen) in national transmigration programm e («Transmi­ the extirpation of plantations, the wholesale depo­ grasi umum/nasional») as early as 1954, but was pulation o f certain area s and the movement of not an effective destination until the seventies. populations elsewhere. For example, Collins [Col­ Between 197 1 and 198 0 there were 4,300 sponso­ lins 1980 ; 1984 ] ha s provided linguistic evidence red transmigrants settlin g in the Moluccas. This for movements at this time which resulted in the increased t o 35,10 0 betwee n 198 0 an d 1985 . setting­up at Nuelitetu, along the south Seram lit­ Although only 2 percent of the total provincial toral, of a settlement of refugees from West Seram, population, they represented 1 7 percent of the speaking a Piru Bay (Wemale) language. During population increas e [Potter , 199 1 : 191] ­ The the nineteent h centur y Dutc h administrativ e greatest expansion took place between 198 2 and

• 18 3 « 1989, with 25,95 3 migrants fro m Java and other sometimes purchasin g lan d from indigenou s parts of the province settling special zones created peoples where government authority permit s (as on Seram (Pasahari and Banggai) and Halma­ in the case of the Nuaulu, see below), sometimes hera (Kao, Wasile and Ekor) [Kantor Statisti k simply taking it. Where the Departments of Trans­ Kabupaten Maluku Tengah , 198 4 : 114 ; 198 9 : migration an d Forestr y hav e recognize d th e 155]. Unde r the Fifth Fiv e Year Plan (1987­90 impracticality o f wet rice cultivation, the y have rates of transmigration hav e increased further, sometimes backed schemes dependent on indus­ and there are plans to resettle more. By 1992 there trial forest crops (Hutan Tanaman Industri). At were 13 settlements all told, 3­1 percent of the pro­ present, as far as the Moluccas is concerned, this vincial population : in excess of 23,042 transmi­ strategy appears only to have been used on Bum grants on Seram, 18,030 on Bum, 20,857 on Hal­ [Goss, 199 2 : 93]. The difficulty, though , is that mahera an d 17 4 o n Aru , a tota l o f 62,10 3 tree crops require the clearance of larger areas of individuals [Goss , 1992:89­90]. There has been forest than fo r rice . At this point th e problems a tendency for family siz e to increas e in recen t posed b y transmigratio n overla p wit h thos e years, and by 1994 transmigrant s ar e likely t o be already mentioned in relation to plantation crop­ 25 percent of the population of Am, 20 percent of ping. Bum, 7 percent of Halmahera and 8 percent of The consequences of transmigration for local Seram [Ibid, p.91]. These figures do not, however, peoples have been complex, but o n the whole include spontaneous migrants who follow later , negative. Among the benefits cited are improved which the World Bank estimates are often more markets and services [Ibid.. 95]. This is probably than doubl e th e numbe r o f official migrant s true for the south Seram area, where the influx of [Donner, 198 7 : 245]. transmigrants ha s been accompanied by upgra ­ It is generally reckoned that in Indonesi a a s ding of roads and bus services, and has resulted in a whole, transmigration an d its knock­on effects the development of local markets with new oppor­ have been more responsible than anythin g else tunities fo r sale of food products. Nuaulu cash for forest destruction, and certainly more dama­ incomes have increased, as we shall see, from the ging than either swidden cultivation o r logging sale of land. Maneo villages also benefit from trade [Donner, 1987:243; Potter, 199 1 : 210]. There is with transmigrants , throug h th e provisio n of no reaso n t o thin k tha t th e situatio n in th e schools in transmigration area s and in the oppor­ Moluccas is any different. The damage is not sim­ tunities afforde d to politically astut e individuals. ply that caused by initial clearance for new settle­ One of the main disadvantages , however, is that ments, but results from few of the schemes being clearance interferes with th e practice of indige­ economically self­contained , ofte n involvin g nous patterns of subsistence. Fo r example, between inappropriate farming models. To compensate , 1982 and 1985,5000 hectares of forest cleared in settlers extract from nearby forest, and seek to central Halmahera for transmigration cu t across extend their land by slash­and­burn techniques, pre­existing lugutil zone s of extraction and sago palms [Martodirdjo , 1988: 4]. Moreover, there is THE MUHULU CASE generally little recognition tha t the viability o f Historical background indigenous sustainabl e swiddenin g require s a The Nuaulu ar e an ethnic grou p of south constant ratio of current garden s to forest fallow, Seram widely known in the central Moluccas for and that therefore forest not being used by local the tenacity an d success with which they have farmers i s no t surplu s t o requirements . Th e clung to a traditional animis t wa y of life. What government has in places expropriated disputed distinguishes them from other similar group s is territory an d no t compensate d owner s [Goss , their demographic strength and the compromises 1992 : 94]. Th e potential fo r conflict i s conside­ they have historically mad e which have enhan­ rable, and even where the indigenous population ced their survival a s a discrete group. My own has in practice receive d certain lega l protections Nuaulu fieldwork ha s spanned the period 1970 ­ (as among the Nuaulu), conflict with transmi ­ 1990, during which time important change s have grants has escalated to an alarming level. taken place in terms of Nuaulu relationship with In parts of Seram local peoples have sponta­ forest. However , there is no particular reaso n to neously and voluntarily settled in transmigratio n assume that before 1970 significant changes were areas (e.g. Maneo). Others have been assimilated not already underway. willy­nilly a s these areas have expanded to incor­ By the latter par t of the nineteenth century porate them (Seti), in some cases turning indige­ Nuaulu inhabited dispersed patrician hamlets on nous villages into «reservations » within large r the souther n sid e of th e watershe d of centra l environmentally­depleted an d immigrant­domi ­ Seram, focused on the Nua and Ruatan valleys . nated zones. In a number of places there have In this schem e of things, the term wasi referre d been attempt s t o incorporat e loca l «tribal » to all cultivated clan­hamlet land , in contrast to peoples into transmigratio n schemes : on Burn wesie, uncut forest. It was not necessary to assert [Goss, 1992 : 95], Sera m [Ellen , 1993b, and clan ownership of particular plots, since all terri­ below] an d Halmahera. The Halmahera scheme tory was clan territory. Bu t also, the idea of col­ involved nomadic forest collectors (Tugutil) and lectively identified Nuaulu territory , a s opposed was reportedly not a success [Martodirdjo, 198 8 : to th e land of individual clans, remained no more 2,22]. Incorporation of Nuaulu into th e Ruatan than a vague abstraction until faced with the poli­ scheme has been a mixed success: some Nuaulu tical realities imposed , first by the conditions of have moved into the area permanently, som e on coastal settlement, an d then b y the Indonesian a temporary basis; the scheme has provided good state in th e nineteen­eighties. I n the eighteen­ access to traditional area s of extraction, but led to eighties Nuaulu began to settle around Sepa, a conflict amongs t Nuaulu, between Nuaulu and polity with which they have recorded traditional other indigenous peoples in the area, and between relations of amity goin g back to the seventeenth Nuaulu and settlers. W e can now turn to this case century. This led to changes in land tenure arran­ and examine it in greater detail. gements upo n which I hav e already reporte d [Ellen, 1977; 1978:81­107] . Here, I draw o n thi s the head of th e cla n which claims the land a s par t earlier work, and present only enough informa­ of it s ow n traditiona l territory. It i s th e head of this tion to clarify the main issues being discussed. clan or his proxy who mus t b e presen t when forest is first cut, in order to seek permission from the Land tenure circa 1970: normative arrangements ancestors of th e relevan t clan, an d t o offer a com­ The cla n Matoke is held to b e primus inter pensatory sacrifice. pares wit h respect to many ritual matters, inclu­ The cuttin g of mature fores t (wesie) yields ding thos e relating to th e utilization o f forest wasi, a term applied to all lan d which ha s bee n resources. Th e exten t to which this wa s th e cas e humanly­altered through clear­felling an d over prior to 188 0 is unclear, but it wa s a custo m firmly which direct (rathe r tha n residual ) right s ar e embedded b y th e time Nuaulu clans were moving maintained. The term may refer to cleared land into the area of Sepa jurisdiction. The relation­ in general or to individual jural units ; to both a ship which Matoke have with the land is perhaps vegetational typ e and to a legal idea. Individual best described as guardianship, though by 197 0 plots of wasi cultivated in any one year, or their this wa s routinel y being translated into Ambonese productive ecological successors ar e known as Malay a s «ownership » (punya, to possess; milik, nisi, which w e ma y translate a s «garden» . Speci­ property). The role is personified in the «lord of fic rights ar e thereb y conferred, through the exer­ the land», the ia onate Matoke , who is ultima­ cise of labour, on individuals, domestic groups, tely responsibl e fo r th e ritua l supervisio n of «houses» and clans who obtain acces s and us e Nuaulu relations with their environment. By 1970 for a s lon g a s th e claim i s effectivel y perpetuated. day­to­day responsibility had been delegated to The right s s o conferre d ar e serial , in that they are Matoke sub­clans (Matoke-hanaie i n Niamonai, simultaneously individual, household, clan, or Matoke­pina i n Rohua ) and, in the case of the whatever; the level of emphasi s depending wholly village of Bunara, to the clan Sonawe-aipura (i n on context . Such land is inherited through the the perso n of th e so­calle d ia onate Matok e male line , though ca n b e transferred between Sonawe [Ellen , 1977; 57 , n.8]. Th e cla n Matoke clans through marriage. has n o practica l jurisdiction over gardens or plan­ tations whic h ar e not regarde d a s traditional Changes in land Tenure. 18B0-1970 Nuaulu territory. If a domestic group or indivi­ The abov e highly­compresse d summar y dual wishes to cut forest from an inland area over represents what w e migh t regar d a s th e «traditio­ which i t i s generally agree d the Nuaulu have nal* an d normative arrangement s a s they exis­ jurisdiction, authorit y mus t theoreticall y b e ted in 1970 . Since first concentrating o n th e coast sought from the ia onate Matoke , or his proxy. important structura l change s have been taking Although the Matoke headman may rule or advise place. The factor s involved ca n b e groupe d under that certain areas cannot b e cultivate d or extrac­ five headings ; 1 . the creation of multi­clan sett­ ted from, in practice the immediate authority i s lements, 2 . cash­cropping , 3. land scarcity, 4 . sal e of land, 5. and market individualism. All of them settlement mad e possible. This has led to rules are discussed in detail in Ellen, 1977, an d the relating t o land increasingly resembling those issue of market individualism i s taken further in relating to other kinds of property. Traditionally, Ellen, 1993b : 131. Her e I mention jus t a few swiddens were cleared, cultivated and re­absorbed aspects for each heading : into the forest, leaving traces only in the form of 1. with th e congregation of the first clans on a small groves of valued trees, such as Areca palms. confined area immediately to the west of Sepa, in But the relationship between people and land was what is now Nuaulu Lama (or, in Nuaulu, Nia- still regarded as highly personal, being likened to monai, «ol d village»), Nuaulu settlements for­ that between a father and children. As a father is feited their genealogica l exclusiveness; formal responsible to the ancestors for his children, so is kinship group s were no longer coterminus with the group for the land. With permanent or semi­ local groups . Some attempt was made to assert permanent groves , the human effort goin g into clan autonom y throug h th e establishmen t o f the transformation of the forest and maintenance separate settlements, first at Aihisuru, then Hahu­ of the land became continuous and hence the walan, Bunara, and finally a t Rohua. But all of particular relationshi p becam e more enduring these ­ with the exception of Hahuwalan ­ even­ and intensive; tually becam e multi­clan villages . This has had 3. prior claims b y Sepa, and by other long­esta­ radical implications for land relations in general blished coastal villages, to accessible garden land, and has given rise to some contradiction in inter­ and increasing demand for land arising from preting the rules, aggravated by pressure on land cash­cropping and population growth had, by the [Ellen, 1977 : 59]. There has been, therefore, a early seventies, for the first time, turned land into greater consciou s identity generate d between a scarce resource. Scarcity provided an incentive groups and individual plots . In coming t o the to maintain relation s with land over long periods coast much ancestral land was neglected, except of time, which resulted in a greater emphasis on that of those clans with traditional claims to land clear­cut, unanbiguous, jural relations with land on which most Nuaulu gardens in the Sepa area held in perpetuity. This ha s inevitably becom e a are now situated. There was a dislocation in the source of conflict between clans, villages, and par­ hitherto enduring connection between clan and ticularly betwee n Nuaulu and non­Nuaulu in land. This situation, together with the movement Sepa and Tamilou. Disputes increased during the of land between clans following marriage , has seventies and eighties, a s pressure on land and meant that the lineal continuity of association other resources has become more acute, due to yet between a clan and a particular area of land was more cash­cropping , indigenou s populatio n broken; growth, in­migration and land sale. The situation 2. cash­cropping, particularly of cloves and coco­ has been further exacerbate d by the lag in revi­ nut, but als o of coffee, quickly followe d daily sing rules and practices relating t o land tenure, contact with th e market econom y which coastal amongst the most important o f which ha s bee n failure t o establis h clea r boundarie s betwee n transactions still involve a customary element . wasi.; Thus, in 199 0 during m y visit to Simalouw, the 4. in th e early seventies sale of land was still a main Nuaulu settlement in the Ruatan transmi­ relatively novel concept [Ellen, 1977 : 63]. No one gration zone, Merpati was engaged in setting­out could remember a Nuaulu ever having acquired the terms of a sale for some new settlers at Kilo 7 land from Sepa since the first gift of land made for approva l b y the loca l Distric t Officer . The by Raja Kamari Kaihatu Tihurua around 1870 , asking «price» on this occasion was : five piruna which established the physical villages of Watane, batu (lit . «ston e plates»; that is old porcelain, Aihisuru, Bunara and Hahuwalan; and only one though no t necessarily oriental in origin), five other instance of sale of land in Rohua could b e meters of red cloth an d 10,000 rupiahs for each recalled. However , in 196 8 Merpat i Sonaw e of household head. This is a global payment to the Watane had bought some garden land from Sep a Nuaulu negotiated on their behalf by Merpati at for 9,00 0 Indonesia n rupiahs , an d durin g the time of my visit. February 1970 , Utapin a Kamam a o f Bunar a bought some land from Sepa a s a means of obtai­ Population growth and transmigration, 1970­1990 ning some level ground for a coconut grove. But During the period covered by my own field­ despite the rarity o f actual sale, the concept of work Nuaulu population ha s continued to grow land a s an exchangeable commodity wa s well­ dramatically (se e table), despite some fall­away due established by 1970 [Ibid., 63­4]; to religious conversion. This has led to greater pres­ 5. although attitude s t o land are being increa­ sure on existing land, intensified b y competition singly moulded b y a market model, individua l along the south Seram littoral with people from tra­

Nuaulu populatio n Census Year Robua Total Total Sepa Tot. Amabai growth i n relation mmm mm (kecamatan) sub­district population . ШШШШШЙШ ШЁЛ mmm ш т ш 496 2667 18,538 1971­19903 ШШШЕЕШ mm 196 544 1975 207 575 шкшкмшш ••• 1 •НЯВЯ В т мш т т т ш •••••i 1 25,207 1980 30,82 0 1981 •MB • • 747 •••••• •»31.023 ш ш ш ш яш ш •Ж 35,30 6 1986 268+ 744 Шт Ш 597 6 1988 608 1

Note For the basis of figures liste d in columns 2 and 3 see footnote 2 . The Sepa figure for 197 1 (base d on 197 0 data) wa s provided by the Kantor Sensus dan Statisik, Dati II. Maluku Tengah. Masohi. The remaining sources are: Amahai 1971 1 Kantor Sensus dan Statistik Propinsi Maluku. 1972]: Amaha i 1978 . Sepa 1978 [Kantor Sensu s dan Statistik Propins i Maluku, 1980] : Amaha i 197 9 1981.1983 , Sepa 1983 [Kantor Kecamatan Amahai, 1983] ; Amahai 198 0 [Kanto r Sensu s Propinsi Maluku, 1980] ; Sepa 1986 , 198 8 [unpublishe d figures in Kanto r Camat. Amahai, 1990]. ditional non­Nuaul u villages, and due to unplan­ cally contrary to any locally­asserted principles of ned immigration, mainl y of Butonese. Prior to this indigenous ecologica l wisdo m [Ellen , 1986] . there had been the arrival of a few Chinese and However, there ha s been increased conflict with Buginese traders, but these have hardly amounted other autochthonous villages over rights t o land to much; and some re­settlement fro m Ambon­ and, since 1990, disenchantement with the effects Lease, internal relocations (Yalahatan, Rutah), and of logging and serious conflict with settlers resul­ of course some growth of the administrative pos t at ting in conviction s fo r murde r bein g brough t Amahai. Growth along the south coast was facili­ against three residents of Rohua. Thus, the possi­ tated by extension of a metalled roa d during the bilities for re­creating some aspects of traditional early eighties. social life and intensifying others , paradoxically At abou t th e sam e tim e th e governmen t through sale of land and other resources to outsi­ began t o establish transmigratio n settlement s ders, is undermining the very system the protago­ along the Ruatan valley [4] Th e government reco­ nists seek to preserve. gnized uncut forest in the vicinity as belonging to the Nuaulu an d encouraged them t o move into DISCUSSION one of the new settlements a t Simalouw. Many In this paper I have tried to make sense of the saw thi s a s a return t o traditiona l lan d an d small amount o f data available on the ecological although by 199 0 only the villages of Watane and and human consequence s of deforestation i n the Aihisuru had moved permanently, many Nuaulu Moluccas, and human response s to this. I have established temporary dwellings, used the impro­ supplemented th e brief reports we have from most ved transport facilitie s to reac h ancestral sago places with one case­study, that of the Nuaulu, areas and began to cut land for cash crop planta­ which summarizes what I hav e published elsew ­ tions. Implicit government recognition of Nuaulu here. The Nuaulu case may well be atypical, but it preferential right s to over one­and­a­half thou­ is at least indicative in several general ways. sand square kilometres [5] enable d them t o sell Firstly, it enable s us to look a t a case where land to other incomers. This alleviated the gro­ deforestation arise s from a number of interactin g wing pressure on Nuaulu land generally and per­ forces: intensification o f subsistence agriculture, mitted them als o to sel l lan d along th e mor e cash­cropping, fores t extraction , loggin g an d crowded south coast around Sepa, most of which transmigration. The commercial lumber industry has gone to Sepa itself and to incoming Butonese. has grown remarkably over the last three decades, As I have argued elsewhere [Ellen, 1993b], this but th e main threat t o the livelihoods of those created a rarely reported situation whereb y an people dependent on the forest, and to the future of indigenous forest people appeared to be endorsing the forest itself, comes from Indonesian govern ­ further fores t destruction, by themselves and by ment transmigration policy . others, for short­term gain . Moreover, the prac­ Secondly, the Nuaulu case shows how the tices which accompanied this were not dramati ­ form, rate s an d consequences of deforestatio n change over time; and that when we assess the tain compensation. But a s we have seen, in th e advantages and disadvantages to local popula­ Ruatan scheme , Nuaulu an d other indigene s tions, these must b e related to different stage s in were permitted t o benefit by selling land to trans­ a process. I note that the initial response to some migrants, an d by releasing land to th e govern­ forms of forest destruction an d consequent land ment in exchange for houses and other facilities. settlement ma y often be viewed positively by indi­ The Basic Forestry Law of 1967 recognize s social, genous peoples, the situation offering opportuni­ traditional and individual rights , but doe s not ties to sustain, and indeed intensify, existing pat­ permit their expression to interfere with the goals terns of subsistence and other cultural practices. of the law, one of which is the production of forest I furthe r sugges t that an eventual realization of commodities. Other special forestry laws override the follow­throug h consequence s lead s t o a the Basi c Agrarian Law , criminalizing certai n middle phas e of uncertainty, which may even­ kinds o f forest us e by indigenous population s tually translate into intense hostility t o incomers [Colchester, 1993 :75]. Thus , in 1971 th e Fores­ and to any additional destructio n o f forest. This try Department designate d Yamdena a protected final stage is fast approaching for the Nuaulu, area, and ten years later created a 60,000 hectares and in the Maneo area it should be expected that nature reserv e with UNDP/FA O backing, only to villages might ac t decisively to defend territorial later issue a decree establishing a logging conces­ interests were they to be faced with comparabl e sion in the area of the reserve. Such overlapping challenges. Of course, not all local groups are in of decree s is common in Indonesi a [SKEPHI , a position to resort to such measures, and locally 1992 : 24; see also Hurst, 1990 : 10-2]. In prac­ patterns of response must b e expected to vary. tice, whatever the legal position, the state has the

Thirdly, th e Nuaulu cas e reminds u s that authority t o regulate and implement th e alloca­ interpretations o f the law vary depending on local tion, use, supply, and care of all resources , pla ­ considerations, a s well a s political an d bureau­ cing the national interes t abov e that of the indi­ cratic purposes. In theory, the Indonesian govern­ vidual [Hardjono , 1991:9]- But in certain areas ment continues to maintain a fundamental asser­ traditional patterns of acces s still prevail. Occa­ tion of the Basic Agrarian Law (Undang­Undang sionally, compromise has been possible, but even Pokok Agraria) of 1 960, namel y that state law is if initia l prospect s seem promising conflic t o f based on adat (customary) la w [SKEPH I & Kid­ some kind is predictable eventually. dell­Monroe, 1993 : 236-7]. But the law is inter­ Fourthly, the Nuaulu case illustrates clearly nally contradictory, a s are the goals and interests that the increased cutting of mature fores t was of different governmen t departments. Th e result only possible through ne w government assisted is confusion. Officially, adat claim s t o historic infrastructures an d incentives, and that land cut areas of forest have no bearing on the selection of is for the establishment of plantation crop s not to transmigration area s [MacAndrews, 1986], and supplement subsistenc e swiddenin g needs . certainly the government does not officially enter­ Nuaulu sell land because settlers as k for it, and because the government expects them t o do so . The greater the degree of isolation, the more effec­ There is certainly a financial inducement, though tive control over resources. Both isolation and low as we have seen, land transfers are still partly a population ar e properties of a system which has matter of customary ritual compensation. The the mechanical effect of making regulation sim­ material spin­off ha s so far proved to b e fairly pler and more reliable; relative autonomy increa­ short­lived, and its distribution the cause of inter­ sing th e probabilit y o f effectiv e regulatio n ­ nal disputes. The benefits to the Nuaulu from sel­ consciously or inadvertently. The varied ecologies ling non­timber fores t products is minimal, rai­ of different Molucca n peoples well illustrate these sing littl e hop e fo r sustaine d non­destructive features in relation t o an interconnected rainfo­ extraction and income­raising [Dove , 199 3 : 17, rest and maritime system. The Moluccas also pro­ citing Shaw et al.]. This is consistent with Dove's vide us with a conveniently long historical time­ observation [Ibid p.2 1 ] that forest peoples do not depth in which to observe the break­down and appear to degrade forests because the y are poor , transformation of locally autonomous patterns of but because they are impoverished by the degra­ subsistence. Developments over the las t 2 0 years, dation of their forest s by external forces which however, have accelerated exponentially long tenn they are too weak to control. processes in a way which endanger the survival of We have long known that ideology and cos­ both remaining sustainabl e extractive regimes mology are poor indicators of practice, and now and the forest itself. know that claims fo r the existence of balanced ecologically self­sustaining rainforest economies are ethnographically difficult t o demonstrate . Some peoples, measurably, do not degrade their environment in any obvious way; but this is often less «adaptation» (in the sense of an outcome of various selectiv e pressures, or o f a particula r ethos) tha n a benign consequence a specified social organization, demographic structure and pattern o f subsistence geared to investment in environmental resource s which replenish them­ selves through relatively short cycles. What is cri­ tical is to maintain population a t a level which never threatens the carrying capacity of a system, even if i t alter s it. To this end, the us e of broad spectrum subsistence strategies reduces stress on particular resources , patches and time­phases, and the degree of isolation from other systems. ACHNOWLEDGEMENTS

The Nuaul u fieldwork o n which par t of this paper is based was conducte d under the auspices of the Indonesia n Academy of Sciences in 1969­71 , 1973,1975,1981,1986 and 1990 ; a period of 2 4 months al l told . The researc h has been supported, on different occasions by the ESRC, the Hayter Travel Scheme, and the University of Kent. My personal indebtedness to the Nuaulu in relation to the particular matter s described here is recorded in full in Ellen 1993b . I am additionall y grateful t o the following individuals for augmenting m y knowledge of current development s (especiall y in other parts of the Moluccas): Rosemary Bolton, Ian Edwards, Jon Goss, Jim Hagen , Christopher Healey, Alistair Macdonal d and Leontine Visser. Of course, I alone am responsible for the final views expressed, and for the accuracy of the information a s printed. Writing support was made possible by receipt of an ESR C research grant (R.000.23.3088) for the period 1991­94. ENDNOTES in particular, th e common practic e of some two kilometres up­rive r from name­changing an d shifts in Makariki wa s beginning to receive its residence between villages. The first official settlers (SO household s in In this and subsequent sections I Rohua figures are designed to 1976, and the same again in 1977) . have taken a n essentially include all ethnic Nuaulu, including In 197 9 the government moved in a phytocentric vie w of rainforest, and Christian and Muslim converts massive 1,17 5 household s into the have little to say about fauna l who remain in the village, any Wae Pi a area, and in 198 2 anothe r depletion or modification. I t is, non­Nuaulu spouses, and their joint 90­150 households [Kantor Statisti k however, widely recognize d that offspring. The y do not include other Kabupaten Maluku Tengah. Moluccas avifauna in particular i s in­migrants. I n 197 1 th e populatio n 1984:115]. However, the most under threat [se e Ellen, 1993b: 199 , of Rohu a was 36 percent of the spectacular project, if not the largest 201]. I have also ignored the impac t Nuaulu total and I have assumed in term s of actual numbers, was the of industria l development . This is for the purpose of calculating th e wholesale removal of the sub­district because, at present, such factors are total figures for 197 3 throug h of Teon, Nila and Serua (a small of limited relevance. However, crude to 199 0 that this has continued to be group of islands some 300 kilometre s oil ha s been extracted from wells on approximately th e case. There are no to the south) t o the Wae Siru area on Seram since 1917 . mainly in th e separate official statistics for ethnic the Nua, which in 198 3 (just before vicinity of Bula. The area around Nuaulu, and occasionally available resettlement) ha d a population Bula has, consequently, long been an figures fo r animists in the sub­district of 6,42 1 [Kanto r Statistik Kabupate n environmental mess . Much more as a whole have to be treated with Maluku Tengah, 1983:24]. recently, work has begun on a factory extreme caution; electoral figure s along the Wae Lau in the Tehoru available for 198 6 refe r only to adults 5 sub­district, th e purpose of which will and do not discriminate betwee n The official land area of Amahai be to produce klinker fo r cement Nuaulu and non­Nuaulu. sub­district i s 2,070.28, and the manufacture. Thi s is likely t o require The Sep a figure for 1971 (base d on shoreline in excess of 80 kilometres . large quantities of water, generate 1970 data) wa s provided by the The figure provided i s obtained i f we alkaline dus t and pollute stream s Kantor Sensus dan Statistik, Dat i II. multiply the length of coastline by with consequence s for riversid e and Maluku Tengah , Masohi. the 5 kilometres accepte d by Nuaulu estuarine biota. It will also involve The remainin g source s are : Amahai and government alike a s that removal of timber fo r construction 1971 [Kanto r Sensus dan Statisti k distance from the coast beyond which and roads, and make claims on Propinsi Maluku, 1972]; Amahai all lan d must be regarde d as the legal indigenous territories fo r mines. The 1978, Sepa 197 8 [Kanto r Sensu s dan entitlement o f the Nuaulu , and exact human an d environmental Statistik Propins i Maluku, 1980]; subtract this from the official area. results of this are presently unclear, Amahai 1979,1981.19% Sepa 1983 The are a left includes part of the though the prognosis is depressing. [Kantor Kecamata n Amahai, 1983]; Mauusela (Wae Mual) Nationa l Amahai 1980 [Kanto r Sensus Park, a forest reserve of some 2 Propinsi Maluku,l980]; Sep a 1986 , 186,000 hectares and in which Agathis ha s even bee n planted in 1988 [unpublished figures in Kantor settlement i s officially prohibited . places for the commercial extractio n Camat. Amahai, 1990]. There is evidence for a similar of dammar [Ormeling , 1947]. benign government attitude towards 4 indigenous rights from Seti and some closeby villages along the north coast 3 Seram did not effectively featur e in In table 1 , column 1 provide s the the transmigration programm e of Seram, where people reportedly years during which I conducted my until the second half of the seventies enjoy «exclusive» right s to ancestral land. own initial census and subsequent (see above), with the arrival of updates. The 197 0 census was 400 families in the Kairatu area. complete an d reliable, but on The first official (a s opposed to subsequent visits I have only updated Independent or spontaneous) census data for Rohua. Even here the settlement o f this kind in South figures must b e understood a s Seram was at Letwaru in 1964, and provisional and errors may occur due comprised some 60 households from to under-reporting of infant mortality Serua By 1976, the Ruatan valley