SPATIAL CO-MANAGEMENT OF INDONESIAN FLOODPLAIN RIVER FISHERIES

An interdisciplinary investigation was nrade on the mechanisms underlying the development of local manugement systems for floodplain river fisheries in Sumatera. The five study sites varied in the hydro-rnorphology and sizes of their river systems and in their spatial relationships between waterbodies and villages. Interviews and formal meetings with Fisheries Department staff, fishernren, traders and village leaders clarified the relative objectives of the different players in the fisheries towards profit maximisation, resource sustainability or social benefits; and their ability to achieve their aims in different hydrological and social situations. Strong local managetnent reginres were found at all five sites, clearly adapted to local conditions. The site with the largest and most remote floodplains was managed for profit through an auction systetn. This contrasted with the other sites, where smaller, more locally accessible waterbodies were managed with more social objectives by the adjacent villages. Where perlnanent dry'season floodplain waterbodies were associated with villages, it was found that these were managed for the long term sustainability of local resources, especially using reserves. In contrast, main river habitats with migratory fish stocks more widely shared between conrtnunities were not nranaged in any way. Due to the complexity of factors affecting floodplain fisheries, and their variability between different locations, it is recommended that further consideration be given to a unified co-management strategy including elements of both local spatial control by communities for the resident floodplain 'black-fish' species, and a wider contribution by regional fisheries departments for the nrigratory riverine 'white-fish'. KEYWoRDS: ::::1T:::i::'*;;""'"nitv management' floodplain liver risheries,


INTRODUCTION
The rnanagement issues for floodplain frsheries can be reduced to two rnain components: withinseason rnanagetnent, relating to the seasonal trade-offs between the catches of different fishing gears, and the social groups that operate them; and between-season managernent, relating to the long-terrn survival and reproduction of different components of the stock the issue of sustainability.
The within-season distribution of fishing effort is particularly irnportant in floodplain fisheries because the expansion and contraction of the flooded area, the sharp seasonal increase in fish biomass and the migratory behavior of the fish mean that the profrtability of the fishery is highly sensitive to the type and tirning of frshing effort.
As in any fishery, the capture of young fish too early in the season prevents them frorn achieving their growth potential and so can reduce the total catch value.In floodplain fisheries, delaying capture rnav also be beneficial as a result of reduced costs as the fish are tnore easily caught in the drawdown and./or low water seasons.
The impact of within season gear interactions on potential incomes is recognised by fishermen, who compete for the best fishing locations in any unregulated environment.Often, far more frshing effort is used than needed to take the maximurn potential catch (Hoggarth & Kirkwood, 1996).These conflicts, and the resulting loss in profrtability, are best avoided through some form of spatial control.This can be achieved either by the leasing of discrete areas within the frshery, or by any other socially accepted mechanism, such as the allocation of fishing spots according to historical precedent or through lotteries The between season sustainability of floodplain fisheries depends on local hydrological conditions and the pattern and level offishing effort, Local hydrology affects fish survival as it determines which gears might be used and how successful they are likely to be.Most €iears can only be used within certain ranges of depth, river width and current speed.In some areas, shallower rivers and floodplain depressions can be frshed out almost completely every year, increasing the chances of recmitment overfishing.Elsewhere they can only ') Mfilirre Resorrrr,es Asscssrtrcrrt (irrlrrp L,trl (MltA(]),47 I'rinr:es (]ate, l,orrdon, SW? 2(lA, tlK ") R+rsean:lt lnstitutc firr I'jeshwrrter lrisheries D.D. Hoggarth, M.A. Thomas, A.S. Sarnita, and Ondara be frshed out occasionally, when water levels fall low enough, or not at all; here stocks are rnore likely to stuvive.The challenge for between-season managernent is to ensure that enough fish survive sornewhere in each Iocal area to sustain the fishery in following years.
As a fishery develops, both within and between season rnanagement usually become necessary.The priority given to the two lnanagement needs varies, however, among the different interest groups.Fishermen will thus agree to the restrictions on access arising from spatial control of their fishery, sornetimes because they have little choice, but also because they often recognise the costs to themselves of continual disputes.They are, however, less likely to rnoderate their catch of overexploited species unless thel'believe that their actions will encourage others to behave similarly.
Leaseholders, whose individual behavior influences only a fraction of the stock and whose uncertain tenure further underrnines their incentive to take a long terrn view are more interested in within-season rnanagelnent issues, even though they rnay be aware that collectively their actions are unsustainable.Fisheries departments or policy makers, whose rnandate extends to the entire stock, are Inore concerned with sustainabilitl' but frequently lack both the knowledge of local conditions and the resourees necessary to rnanage the fishery effectively.
Local cornrnunities, on the other haud, tend to be concerned with both sets of management issues.Within-season rnanagernent avoids both the dissipation of fisherrnen's incornes through excessive levels of effort and conflict.Between season rnanagernent can ensure that the flow of incornes to rnernbers of the community is sustained.Moreover, strong local participation in fisheries managernent decisions has a nurnber of distinct advantages.The distributional trade- offs implicit in alternative non Government controlled management regimes can be set against local socio-econornic priorities.
Cornmunities with high unernploytnent rnight thrrs prefer to widen the fishing opportunities available, even if this reduces the overall profitability of the fishery.AIso, local   comrnunities or fishermen's organiza tions often have the rneans to deal with the practical problerns of managetnent, for exarnple bv encouraging collective modifications in behavior, imposing social sanctions on rule breakers or acting as arbitrators in disputes.
Though well equipped for a management role, local communities can not, however, be the solution to all problems, particularly where the resources they exploit are shared with other communities.
Against this conceptual background, this study was undertaken to determine how river floodplain fisheries in southern Surnatera are currently managedby both formalmanagement agencies and local communities.Significant variability was fotrnd in the local management styles at five study sites.However, using an interdisciplinary approach to the assessment of the physical/ ecological and sociaVcultural features of the sites, the apparently complex management issues were clarifred.

METHODOLOGY
Comparative information was collected during freld visits in May 1996 to five case study sites in Jambi and South Sumatera provincesr) on the Islandof Sumatera (Figure 1).A fourman research team spent between one and frve study days at each of the field sites.During the freld trips, a mixture of informal interview s, sernistructured interviews, and formal rneetings were arranged with people connected with each fishery, in both field and offrce situations.The categories of respondents included Fisheries Extension Service (Dinas Perikanan) staff at both provincial and kabupaten levels; village leaders both administrative and traditional (adat); fishing leaseholders; frsh traders; and group and individual fishermen.
At each study site, largely qualitative information was collected on (1) the morphology of the fished waterbodies, (2) the spatial relationships between waterbodies and human settlements (3)  the types of fish species exploited, (4) the fishery management mechanisms in use, and (5) the resulting frshing practices.Particular attention was given to the spatial aspects of the allocation of use rights (e.g. by licensing), and the prevention of overfishing (e.g. by the use of reserves, or by localised gear restrictions).The ecological and rnanagement differences between the study sites are described in the following sections, and surunarised in Table

Floodplain Morphology and Human Settlement
The five study sites differed significantly in the size and rnorphology of their river floodplain systerns, and in their patterns of human settlement (see Table 1 for all results).
The first study site, Lempuing River, has one of the rnost productive inland fisheries in Indonesia, and has been the subject of a number of earlier studies (eg.Vaas et al., 1953; Giesen & Sokotjo,  1991; Hoggarth & IJtorno, 1994).Though the Lempuing is relatively small (15-40m wide), its high productivity derives from a 200krn!area of 'lebah' floodplain, containing nineteen'lebung' lakes, up to 390ha in size (Figure 1).
The other four study sites were located on relatively srnaller, less productive floodplain systems, though rnostly closer to large rnain rivers of several hundred rnetres in width.
At study site 2 in Kabrrpaten Musi Banyuasin or 'Mtrba' (Figure 1), a narrow secondary river, the 10-20m wide Lumpatin River, flows frorn rninor floodplains without permanent lakes or other dry-season habitats, directly into the broad Musi River, passing through the village of Sekayu.
Three study sites were also visited around the Batanghari River near Jambi (Figure 1).Study sites 3 and 4 were located on separate, -30m wide side channels of the Batanghari.In addition to the river waters of the Kurnpeh Channel River, Arang Arang village or 'deso'has exclusive control over a lake, Danau Arang Arang, with a dry season area around 20ha.Desa Jambi Kecil (study site 4), has fishing rights to part of the Berernbang Channel River, which retains permanent water in all dry seasons, particularly in its deep'Iubuh' pools.The frfth study site in Desa Muara Jambi lies alongside the rnain Batanghari River, downstrearn of Jambi Kecil.There are no significant floodplain or lake habitats close to Muara Jambi, but the village owns a lake called Danau Gerang, This lake, 17 krn away from the yillags (Figtue 1), was transferred frorn its original village owners in compensation for a murder in ancient tirnes.
The settlements of Sekaytr, Arang Arang and Jambi Kecil are thus all close to their main fishing grounds.The villages at the other two sites are less clearly associated with the fished waterbodies.Muara Jarnbi is rernote frorn its Danau Gerang for the above historical reasou, The Lempuing River has no permanent settlements among the waterbodies due to the depth of the floodplain waters and the absence of permanent roads.Most of the Lempuing fishermen live iu Pedamaran, downstream of the fishing grounds (Figure 1) and build temporary huts on the waterbodies they lease through the auction system (see later).

Fish Migrations
Floodplain fish species have adapted their behaviour to the hostile conditions in residual, dry-season waterbodies in dilferent ways.Some species, referred to as 'white-fish' in the classification of Welcomme (1985), return to the rnain river channels as the floods recede.Others, 'black-fish', have developed anatomically to deal with deoxygenated water and remain in the floodplain pools.
The whitefrsh species, such as the sihuid catfish Mystus netnurus should have 'authority' and the giant freshwater prawn Macrobrachium rosenbergii are the most 'shared' stocks, as their extensive rnigrations take thembetween the spatial fishing units of different fisherrnen (MRAG, 199D.Whitefish tend to be most vulnerable to capture during their migrations through the channels where the floodplains drain back into the main rivers.The black-frsh species, such as Helostoma temmincki and Osteochilus hasselti, ate more conlined within single waterbodies, or at least within a lirnited range, and are most vulnerable to capture in the dry season.
The Lernpuing and Jambi sites, with their mixture of rivers and floodplain lakes, produce catches with more than 30 different fish species Figure 1.Location, river morphologies and human settlements of the 6.ve case study sites in South- ern Surnatera, Indonesia the following five respects: revenue distribution; access to bidding; license pricing rnechanisrns; auction cornrnittee rnembership; and habitats licensed.
Licensing may raise substantial incornes for frshery adrninistrators.In Kabupaten OKI, which includes the Lernpuing study site, total ticense fees were Rp890m (-$380,000) in 1996.In OKI and Mtrba Kabu.patens, at least 6O% of such auction revenue was received by the habupaten, adrninistrations, contributing up tn 50% of the total habupaten incornes.Only 30% of the revenue was distributed to desa level adrninistrations in these sites, cornpared to 95-100% at the Jarnbi sites.An additional L\t'A 'retribusi'tax was also levied on Iicense fees in OKI.
At the Lernpuing, Muba and Muara Jarnbi sites, the auctions were open to all bidders, while at Arang Arang, bidding was limited to village rnernbers only.
License prices were set in OKI by an auction cornrnittee at last year's sale price plus a fixed percentage of up to 15% per year.Unsold license units were available to'sealed bids'after the rnain auction.At the other sites, Iicense prices were more flexible, with a guide price being set for the auction sornetimes at lower prices than previorrs years.
At the Lempuing site, the auction cornrnittees responsible for setting license guide prices included only fishery department staff and adrnirristrators at kabupaten and hecamatan levels.At the other sites, the auctions were organised at village level, and frshermen were also represented on the cormnittee to advise on the current status of resources and the feasibility of alt.ernative guide prices for each waterbody.
Finally, for the Lernpuing and Sekayu sites in the South Srunatera, the auction system allocated fishing rights to almost all waterbodies, including their floodplains.At the Jarnbi sites, only a sub.set ofthe dry season pools and river sections was Iicensed, leaving the floodplain areas open access for all fishers.
Due to the uncertainty of future ownership and the migrations of fish out of their waters, licensed fishermen at all sites exploited their resources heavily to maxirnise their catches.Licensing systems do not provide any incentives for long- terrn conservation of resources.The following two sections describe the additional 'technical' regulations used dt each site to control fishing activities and conserve frsh resources.

Reserves and Ceremonial Fishing
Reserves were used as conservation measures at three of the study sites.In the Lempuing River, the 126ha 'Teluk Rasau'Lake (Figure 1) was not included in the annual auction by the Fisheries Extension Service, The enforcement of the reserve was limited, though, and some frsherrnen were cynical about its effectiveness and value.Reserves were not used in the riverine Sekayu study site, though they had been recently implemented by the Kabupaten Muba Fisheries Extension Service in other nearbv lake waters.
In contrast, traditional reserves were strongly supported by the local communities at two of the Jambi villages studied, and supported by cerernonial fi shing practices.

Fishing on Danau Arang Arang reserve in
Jambi is traditionally restricted to one ceremonial day each year, the 'hari berharang'.This major comrnunal event, attended by local dignitaries, takes place during the low water period, on a Sunday, to allow the maxirnum partieipation of village qrembers.Gill nets and seines are forbidden, and the fish are driven into a capture chamber to be fished by relatively inefficient cast nets.Even with these limitations, catches are large, ranging from 2 to 10 tonnes.
In the nearby Desa Jambi Kecil, fishing is open access in all waters dwing the whole year, except in the three 'Iubuhpusa&o'or sacred pools of Sungei Berembang.These lubuh pusaka are restricted as dry season reserves, though fishing is again allowed on ceremonial'hari berharang' days.
Participation in the hari berharang is restricted to village members and higher district leaders.Only relatively ineffective spears and lift and jump nets are permitted on these days, and the net fishing positions are determined by the ancestral rights of cornmunity rnembers.
Both the Jambi hari berharang ceremonies have been practised since before living memory, and have some religious sigrrificance, with fish stocks being regarded as the gift of the ancestors.The annual decision to fish these reserves is based on their perceived stock sizes, since the catches must be large enough to demonstrate the high status of the village to the invited dignitaries.On average, the Jarnbi Kecil lubuhs were only fished approximately once every three years, and local chiefs estirnated that only a third of the frsh in the pools rnight actually be caught in the one afternoon of frshing.

Local Fishing R.egulations and Fishing
Gear Use Indonesian governrnent technical regulations on fishing gear use in inland waters are restricted to bans on the use of poisons, electric fishing and explosives, and a few rnesh size regulations.In these artisanal and rural fisheries, such regulations are diffrcult to enforce.There are no regulations which protect fish stocks against the many other highly effective gears such as barrier traps and dry season frsh drives.
The choice of fishing gear use at the fully licensed Lernpuing River and Sekayu study sites is thus the sole responsibility of the licence- holding fi.shermen.At these sites, seven rnain classes of fishing gears are used (Hoggarth &  Kirkwood, 1996).In the ebb season, well frnanced groups of fishermen usually use expensive net and flume traps mounted in heavy wooden barriers, spanning the full widths of their rivers and channels (Hoggarth & Utomo, 1994).In the dry season, if water levels fall low enough, the same group fishermen also use active fish drives to catch as many of the rernaining fish as possible.
Individual fishermen, sub-licensed by the groups, use smaller, cheaper gears including various forms of srnall lift nets, gill nets, cast nets, portable traps and hooks.
In the smaller Jarnbi sites, fishing communities sornetirnes rnahe their own local gear regulations to conserve their stocks, These are particularly associated with the reserves and'hari berharang' ceremonies described above, but also serve to reduce conflicts in open access areas.
At the Arang Arang site in Jambi, fishing rights are auctioned for the three inflow channels of the reserved Danau Arang Arang, while its outflow channel to the Kurnpeh River (Figure 1) is open access.The inflowing river sections are intensely exploited by their license holders, using the effective barrier traps and seines found in the Lernpuing.These gears are, however, banned on the unlicensed outflow channel, which is shared by up to 50 village fishermen using lift nets and portable traps.

Interpretation of Results
This discussion atternpts to interpret the rnanagement systems and exploitation practices used at the five sites, based on their physical and social characteristics.The greatest differences in rnanagement styles were observed between the rnost profrt-oriented IFR Jou.rnal Vol.IV No.I, I ggg Lempuing River site and the more communitv- oriented Jambi sites.
In the Lemptring River, the spatial licensing systern is organised by the habupaten and lucamalon level administrations.The &obupatenhave a high stake in maintaining the auction system, due to their dependerrce on revenues obtained from it.Several key points about the structure ofthe auction system confirm its focus on revenue generation.Firstly, the auction is open, allowing the possibility of high bids from rich traders outside the local area.Secondly, the gtride prices are set by an administrative commjttee, with no fishermen's representative, and are increased on a regrlar basis.Thirdly, additional firnds are raised by the lSYo 'retri.busi'tax,imlrosed in OKI since 1983, despite local objections.
In addition to raising local government revenues, the Lempuing licensing system also has some clear within-season benefits for the fishermen.It delegates authority for the regulation of frshing effort to the license holders for the whole year.This reduces the potential for conflict and enables fishermen to use the productive barrier and driving gears to catch fish in the drawdown and dry seasons.In an open access situation, such gears rnay be unproductive if too rnuch of the catch is taken before their operating seasons.
The Lempuing rnanagement system thus has econornic and financial benefits, and prevents conllicts for both the administration and the fishing comrnunity.Apart from the Teluk Rasau reserve, though, the system does little to guarantee the sustainability of the frshery.Whitefish are caught in large munbers in the barrier traps and black- 6sh are targeted in the dry season fish drives.Total catches are stillhigh at the present time (Hoggarth  & Kirkwood, 1996), but many of the most valuable fish species are becoming increasingly uncorrunon (MRAG, 1997).
In the Jambi study sites, the management styles observed suggest a greater focus on betweenseason sustainability issues for the benefit of local communities.
The organization of the Jambi auctions at desa level, and the restriction of bidding to village residents may reduee total revenues by preventing high bids frorn rich outsiders.However, they have the advantage that aII the incornes from the frshery stay in t}re desa.Conflicts are also reduced by the D.D. Hoggarth, M.A. Tltontas, A.S, Sarnita, and Ondnra The Jambi Kecil lubuk pusaka and the Arang Arang reserve lake are clearly managed by these villages for the purpose of long term strstainability.The hari berharang in these waters take a lirnited catch in an effrcient manner, and with social equity prornoted by the ceremonial participation of all the village households.By lirniting the type of fishing and its duration to one day or less per year, the villages reduce the chance that their local black-fish species will be overexploited.According to the village elders, fisheries within the local area continued to be healthy because their dry season stocks were only fished for one season out of three.This difference in approaches between these three study sites rnay be simply explained by the fact that the Jarnbi Kecil and Arang Arang fisherrnen live permanently among their small waterbodies, while the Lernptring River fisherrnen are only ternporarily associated with their waters in the years of their licenses.As described by Ostrorn (1990), comrnunities lnay only be expected to take responsibility for those natrual resoruces for which they have exclusive ownership within clearly defined boundaries.In this regard, the Lernpuing River lake district rnay be seen as too large for individual managernent by its rnany fishers, necessitatirrg the rnanagement role of a higher authority.The high potential income from this major fishery then appears to have directed the managernent style of that arrthority rnore towards profi t rnaxirnisa tion than resoruce conservation.This spatial viewpoint is supported by the abserrce of reserve waterbodies and ceremonial fishing in the Muara Jarnbi study site.Due to the rernoteness of Muara Jarnbi's only lake, Danau Gerang, the status of such a reserve rnay be irnpossible to enforce bv the Muara Jarnbi corrununity.Alternatively it rnay sirnply not be regarded as one oftheir ancestral resources or local responsibilities.
A spatial viewpoint is further supported by the attitudes of the frshing corrumrnities towards the relativelv inrmobile blach-fish species rnost resident in their waters, and the riverine white-fish species, rnore shared with other cornrnunities in the catchrnent.The Arang Arang and Jambi Kecil Villages thus protect the blackfish species in their reserves, but take no steps to conserve the riverine white-frsh of the secondary Sungei Krunpeh or the rnain Batanghari River.In Sekayu, sirnilarly, there are no fishing restrictiorrs for the rnostly white-fish species carrght in the village's waters, even though the village straddles the Lurnpatin River.In stumnary, a comrnunity's incentive for the sustainable rnanagement of riverine fisheries appears to be strongly but simply influenced by the degree of control that it has over its resources.
Where communities have traditional associations with local waterbodies, they have been found to rnanage them for the sustainable benefits of community members.Where resource control is shared with other cornmunities.either for whitefrsh species or in large or remote waterbodies, cornmunities rnanaged primarily for short-term financial gain rather than long-term stability.Wbile it is clear that srnall communities may be encouraged to take responsibility for their local fish stocks, the rnore shared resoruces may only be conserved by concerted joint actions, under the regional coordination of higher'catchment management' authorities (Caddy, 1982).

CONCLUSIONS
Using a case study approach, this research has shown that the management of Sumatera river fisheries is dependent on the detailed local relationships between river morphologies, the behaviours of fish that inhabit them, and the distribution of local communities among the waterbodies.
In view of the complexity of factors affecting floodplain fisheries, and their variability between different locations, it is recommended that a unified inland fisheries management strategy should include elements of both spatial control by local communities for their resident floodplain black-fish species, and a wider contribution by regional Fisheries Departments for the migratory riverine white-frsh.Such a co-management strategy should recognise and take advantage of the local rnanagernent already present in traditional cultures.A shared approach would also remove sorne of the difficulties facedby the Fisheries Extension Service trying to irnpose new administrative regimes from their centralised offices, get constrained by limited resources for field-based monitoring and errforcernent (Bailey & Zerner, 1992).
The actual contribution to be made by the Fisheries Extension Services for the rnanagernent of the riverine white-fish is not yet clear.Until effective conservation rneasrues can be found, which have the support of local communities, the highest priority shorrldbe in rnonitoring stocks to determjne when active rnanagement is necessary.Whatever lneasures rnay eventually be recomrnended, these should be applied in a spirit of co-management, and should not be at the expense of the existing and successful strategies used by communities to conserve their local black-fish stocks.
. Hoggarth, M.A, Thonr,as, A.S. Sornita, and, Ondara RESULTS AND DISCUSSION accountability of local people, and the Arang Arang cornmunity appears happy with the current balance of reserved, open access and licensed waters. 30 1.