Review of Aquaculture and Freshwater Fisheries in Papua New Guinea

This report was prepared during the course of the project identified on the title page. The conclusions and recommendations given in the report are those considered appropriate at the time of its preparation. They may be modified in the light of further knowledge gained at subsequent stages of the project.
The designations employed and the presentation of the material in this document do not imply the expression of any opinion whatsoever on the part of the United Nations or the Food and Agriculture Organisation of the United Nations concerning the legal or constitutional status of any country, territory or sea area, or concerning the delimitation of frontiers.

FOOD AND AGRICULTURE ORGANISATION OF THE UNITED NATIONS
Rome, 1989

Hyperlinks to non-FAO Internet sites do not imply any official endorsement of or responsibility for the opinions, ideas, data or products presented at these locations, or guarantee the validity of the information provided. The sole purpose of links to non-FAO sites is to indicate further information available on related topics.

This electronic document has been scanned using optical character recognition (OCR) software. FAO declines all responsibility for any discrepancies that may exist between the present document and its original printed version.

1. INTRODUCTION

This document serves to provide a brief and simplified outline of previous aquaculture and inland fisheries activities within Papua New Guinea (PNG). The present analysis of previous attempts to develop aquaculture and freshwater fisheries is made in order to identify the factors behind the success and failure of previous projects and fishery activities. This is highly relevant to the Sepik River Fish Stock Enhancement Project because previous experiences relate to the rationale for stocking and indicate problems and successes arising through alternative options (other than stocking) for developing this fishery.
This document discusses all previous inland fisheries and marine and freshwater aquaculture activities in PNG, including areas outside of the Sepik (and Ramu) River basins.

2. FRESHWATER FISHERIES

Freshwater fisheries are historically and presently dominated by subsistence fisheries in PNG. There have been a few semi-commercial and commercial developments inland which are mentioned below. For the present purposes the barramundi, Lates calcarifer, fishery is excluded from this document. Although a significant part of the barramundi catch is obtained in freshwater, this species is not included because of its migratory habit and the fishery having a long and distinct history, covered by Anon. (1989). In any case, barramundi are absent from the northern PNG rivers (Allen and Coates 1989). Also excluded from this document is the development of the Baimuru and Kikori fishery on the Purari River which is also covered elsewhere (Anon. 1989). These fisheries are dominated by barramundi but include a significant additional estuarine component. They have also been subject to considerable capital inputs through the International Fund for Agricultural Development (IFAD) project (see IFAD project and barramundi - in Anon. 1989).
Inland fisheries (including freshwater, estuarine and mangrove areas) were briefly reviewed by Coates (1986). The present document deals with freshwaters and, therefore, excludes estuarine/mangrove areas. The division between the two habitats is defined by Petr (1983) and refers to the extent of the Pandanus spp zone in rivers which extends to the limit of saltwater penetration into freshwater.
Freshwater habitats in PNG are extensive and a review of them is given by Osborne (1987). The major habitats are large river systems which have extensive floodplains at lower altitudes arising from networks of progressively smaller rivers and streams originating in the highlands. Almost all areas of land in PNG have at least some freshwaters nearby. Rivers in southern PNG, flowing towards the Gulf of Papua, have large deltas forming significant areas of estuaries with brackishwater, characterised by extensive mangrove systems. Rivers in the north, generally flowing into the Bismarck Sea, do not have deltas and limited, if not non-existent, estuarine/brackishwater zones inland. Consequently, mangrove systems in the north of the main island are negligible.
There are over 5,000 lakes in PNG. Most of these, however, are small except for Lake Murray (Strickland River) and Chambri Lake (Sepik River) both of which are considered extensions of river floodplain and vary seasonally in area. Most lakes within the Sepik/Ramu catchment are associated with floodplains. Non-floodplain lakes are limited in size and number.

2.1 Freshwater fishes

PNG lies east of Wallace's Line which is a zoogeographic boundary separating Asia from Australasia. With one minor exception, PNG is devoid of native primary freshwater fishes. All fishes found in freshwater are either migratory, coming from or going to the sea to spawn (e.g. tarpon, sharks, barramundi, eels) or are permanent inhabitants of freshwater but belonging to or derived from marine families (e.g. fork-tailed catfishes, gudgeons, gobies, grunters, jacks). Practically all of the families of fish forming conspicuous components of catches from inland fisheries elsewhere in the world (e.g. Africa, Asia, S. America) do not naturally occur in PNG.
PNG is divided geologically by the central range of mountains into distinct northern and southern regions. Freshwater fishes living in the south are generally different species to those in the north unless they are capable of migrating through the sea. In general, southern rivers have a much higher diversity of fish species than northern rivers (Allen and Coates 1989). In particular, barramundi and several other commercially important species in southern rivers are absent from northern rivers. Details of the fish fauna of the Fly River (southern PNG) are provided by Roberts (1978) and for the Sepik River (northern PNG) by Allen and Coates (1989); the latter authors make a comparison of these two systems.

2.2 Previous and present exotic fish species introductions

Out of the 29 species recorded as being introduced into PNG (West and Glucksman 1976 - see Table 1) only two have since produced significant fishable populations inland. These are tilapia,Oreochromis mossambicus, in the Sepik River (Coates 1985) and elsewhere in PNG, and common carp, Cyprinus carpio, in the Sepik River (Coates 1984, Ulaiwi 1989). The government of PNG now attempts to control fish introductions in view of the cautionary approach recommended by the “Sepik River Fish Stock Enhancement Project”.
Present fish introductions are officially restricted to the importation of trout (Salmo gairdneri) eggs from Australia, for both aquaculture and stocking rivers in the Highlands (see later). In addition, PNG faces continuing problems with the unauthorised importation of aquarium fishes by local and expatriate aquarists. This activity is discourage by PNG authorities and importations are not allowed without a permit, which has normally been declined. However, frequent infringements of this procedure have occurred. At least five unauthorised importations of aquarium or pond fishes occurred in 1987. One stock was destroyed, whilst the other four could not be located. In addition, it is likely that other importations occurred, unknown to PNG authorities. In addition to those species listed in Table 1, the swordtail (Xiphophorus sp), an imported aquarium fish, is known to have established at least around the Port Moresby area. Two fish species are known to have entered PNG from Irian Jaya (Indonesia - the western half of the island of New Guinea). Anabas testudineus, the climbing perch, was recorded by Coates (unpublished) from rivers near Bensbach (southern PNG) in 1984 and has since spread into the Fly River system (Hortle, personal communication). The climbing perch is presumed to have been introduced into Irian Jaya from Java. It may be able to “walk” between catchments. In December, 1988, a previously unknown introduced fish species was reported by villagers near Amanab (West Sepik Province) to have entered local streams from Irian Jaya. At present, there is little more information and a specimen has not yet been collected for identification. It is not clear whether it occurs in the Sepik catchment or is restricted to streams discharging into Irian Jaya from PNG. Additional species are known to have been introduced into Irian Jaya and have the potential to enter PNG either through shared catchments or people carrying them when moving traditionally between the two countries.

Table 1. Introductions and distribution of exotic fish in Papua New Guinea according to West and Glucksman (1976).
Species	Date of first introduction	Present distribution	Comments
Gambusia affinis	1930	Very widespread	Of little value in malaria control, and possibly harmful to native fish.
Salmo trutta	1949	S.H.P. Chimbu, E.H.P.	Moderate potential for localized food source, aquaculture, and tourist incentive.
Salmo gairdneri	1952	E.H.P., S.H.P., W.H.P., Enga, Chimbu, W.S.P.	Similar, but better potential than S. trutta.
Tilapia mossambica	1954	Very widespread	Important as food, especially in Sepik and Central Provinces. Considerable potential.
Trichogaster pectoralis	1957	Very limited distribution; Central & Gulf Provinces	Of no value at present; possible potential in sewerage ponds and manure-methane digester systems.
Ospbronemus goramy	1957	Small, widely scattered populations	Of little value, and low potential.
Cyprinus carpio	1959	In most suitable highland streams and lakes; isolated populations in lowlands	Of moderate value, potential probably realized.
Hypophthalmichthys molitrix	1961	Nil	Unsuccessful in pond culture trials.
Bidyanus bidyanus	1962	Nil	Intended for stocking in natural waters, but did not survive.
Tandanus tandanus	1963	Nil	Pond culture trials discontinued and stocks destroyed.
Carassiops klunzingeri	1963	Nil	Intended as forage fish, but did not survive.
Retropinna semoni	1963	Nil	Intended as forage fish for trout, but did not survive.
Ctenopharyngodon dellus	1961	Nil	Unsuccessful in pond culture trials. Stocks destroyed.
Plectroplites ambiguus	1966	Nil	Released in natural waters but did not survive.
Trachystoma petardi	1966	Nil	No information available.
Puntius gonionotus	1970	Experimental ponds. Aiyura, E.H.P.	Not to be distributed.
Poecilia reticulata	1967?	Port Moresby area	Escaped aquarium fish, possibly harmful.
Trichogaster trichopterus	1970?	Port Moresby area	Escaped aquarium fish, possibly harmful.
Salvelinus fontinalis	1974	Some fry released in streams in S.H.P.	Considerable mortality of ova during hatching.

2.3 Previous freshwater fisheries activities

There have been a large number of reports and projects dealing with freshwater fisheries, collated by Lock and Waites (1985). Many of these, however, give only incidental information of limited value for planning purposes. Only two inland fisheries activities (excluding aquaculture) have involved attempts at increasing production to a commercial level based on freshwater species:

2.3.1 freshwater crayfish (yabbies)

Freshwater crayfish or yabbies (Cherax spp - mainly C. albertisii) occur from the Irian Jaya border eastwards to the Binaturi River and north to the Middle Fly. They are also reported from the Bamu/Aramia River systems. They are restricted to southern rivers and do not occur in the Sepik/Ramu. There is limited information on a previous yabbie fishery from freshwater swamps in Western Province. Anon. (1980) reports a commercial production of 3.7 t of yabbies from Sigabaduru village in 1975 with catches as high as 0.5 t per day at peak season. Catches from three villages in this area totalled 11.3 t in 1976. Production in 1977 was less than 100 kg. Variations in catches appeared to be due to natural factors, especially seasonal differences in water levels between years. Sampling at Ali village (river systems east of Daru) in 1978 produced catches of 50 yabbies per trap per night. In 1978, a FAO funded project sent fisheries research vessel Cee B to the Aramia River, together with a yabbie project manager, in order to assess the potential of this resource. Unfortunately, this took place too late in the season and the results were inconclusive.
FAO (1982) provided recommendations on developing the crayfish industry in the Fly River system including results of the 1978/9 survey. Yabbies are caught in commercial quantities only in the dry season and the fishery is, therefore, highly seasonal. No areas that were surveyed could support a fishery dependent entirely on yabbies. In some areas, the production of yabbies and freshwater prawns could supplement the production of fin-fish. The potential for yabbie farming is good according to FAO (1982). There may be potential here if suitable waste feeds can be identified (e.g. waste from barramundi filleting) - but note general comments on socio-economic problems experienced with aquaculture in PNG outlined later in the present document.
Since the late 1970s yabbie production has ceased. Quite why this is so is not known. It appears that the highly seasonal nature of the fishery and the variations in catches between years and locations make it difficult to develop this resource. However, a production of 11.3 t of yabbies from a small area, and 0.5 t per day from a single village, indicates a substantial, if variable, resource.

2.3.2 the Sepik salted tilapia (solpis) project

Tilapia, Oreochromis mossambicus, were brought to PNG for culture in ponds (see Aquaculture Section) where they were never a success in terms of fish production. They “escaped” from ponds near the Screw River, a Sepik tributary, in the mid-1960s. Once on the floodplain they quickly proliferated in both numbers and distribution. By the mid-1970s they were considered abundant.
Various calculations were made to estimate the size of the Sepik tilapia resource. These were invariably based on direct comparisons with African rivers and the yields achieved there. Hypothetical figures of yield of between 20,000 and 50,000 tonnes per year were obtained on this basis. Quite who was responsible for these estimates is not known but they were extensively quoted at the time (e.g. Wilson, undated; Bardach 1977; Asian Development Bank reports quoted by Takendu and Sadler 1980). The estimates were fundamental to plans in the mid-1970s to develop the Sepik tilapia fishery. Mitchell (1972), for example, estimated the amount of tilapia to be 128,000,000 lbs (57,000 t) in 2,000 square miles based on an assumption of a yield of 100 lbs per acre; resulting in the Sepik being considered the “big potential”. Wilson (undated) might also be consulted for further information on the approach to Sepik tilapia at the time. It is worth mentioning, as an indication of things to come, that up until this time nobody had actually fished the river system in order to obtain accurate data on yields.
The “East Sepik Rural Development Project” (ESRDP), funded by the Asian Development Bank was initiated 1978 and included as a sub-project the exploitation of tilapia based on salting and drying fillets, later becoming known as the “SOLPIS” project. It is not clear why the project decided to salt and dry tilapia which meant the introduction of new technology and high infrastructure costs when Sepik people traditionally smoked their fish, and usually quite effectively.
In the original Appraisal Document (1977) for the project, the total fish resource of the Sepik was regarded as “30,000 t per annum of which 21,000 t was tilapia with 9,000 t being able to be exploited for solpis” (quoted by Takendu and Sadler 1980). Based on these production estimates a loan of about K770,000 (presently about US$850,000), to be repaid over ten years, was obtained specifically for this sub-project. The production estimates of the project were based on the previously mentioned yield estimates. During the first five years of the project, projected yield was to increase to 357.7 tonnes (dry fish weight) in 1981 (Rodwell 1982). The progress of the project was reported by Thorpe (1977, 1980). Production never reached projected figures. The best year was 1979 with 40.3 t production (only 30.2% of the projected figure). Since then production has declined. At year five (1981) production was 10.3 t (2.9% of projected figure).
Rodwell (1982) noted that by 1981 the infrastructure of the project had been set up in anticipation of much higher outputs and, therefore, problems emerged in relation to the administration of the project from the under-utilisation of project capital. Rodwell also includes data that shows the cost of production in 1981 was about K145,000 or about K14.00 per kg of saltfish (it retailed at that time for about K1.40 per kg !). Recent production (1987/8) has been negligible. Takendu and Sadler (1980) and Rodwell (1982) should be consulted for further information on economic aspects of this project.
During the first five years of this project there was an outbreak of the floating weed Salvinia in the region (Mitchell et al. 1980). This was variously blamed for the failure of the solpis project. However, Coates (1985) showed that production was still low in the major solpis producing areas where this weed did not occur. The weed undoubtedly had a negative impact on fishing in many areas but cannot be used solely as an excuse for the economic failure of the project. Salvinia in the Sepik has since been satisfactorily controlled by the beetle Cyrtobagous salvinae under a joint FAO/PNG project (Room and Thomas 1985). However, despite the success of the control of this weed, solpis production has not significantly increased.
Coates (1985) made yield estimates for the Sepik River based on analyses of solpis data, market landings and very extensive catch data from gillnets and local fishermen. The yield of the fishery in floodplain regions was estimated to be between 3,000 and 5,000 t per year (averaged for present purposes to 4,000 t per year). Coates (1985) also noted that his yield estimates were only about 10% of the yield obtained from “similar” African rivers; hence the previous estimates in the region of 40,000 t per year undertaken by coefficients of comparison with Africa. A major reason for the low yield was suggested to be due to the differences in the fish faunas between PNG and other zoogeographic regions. Coates (1985) estimated that tilapia accounts for about half of the present yield. Although locally successful, it is only a single species whereas a river of this size might be expected (in other regions) to have many more fish species contributing to commercial catches.
The reasons for the failings of the solpis project are discussed at length by Takendu and Sadler (1980) and Rodwell (1982). These should be consulted for further information.
The following factors are highlighted with respect to experiences with the solpis project:

the project concept was based on assumptions on yield derived by direct comparisons with African rivers. Although the information of Coates (1985) was not available at the project planning stage, it has been known since the mid 1800's that PNG had a quite different freshwater fish fauna to other regions - particularly a lack of “true” freshwater fishes. A moment's caution with yield estimates, or, even better, some better estimates actually from the Sepik, might have prevented possible over-optimism at the planning stage. This indicates the danger of undertaking planning and investment based on potentially inaccurate yield/production estimates;
there were fundamental flaws in the economic logic of the project. In particular:
- the very high infrastructure costs involved with the development of a previously undeveloped fishery in a remote area;
- the economic analyses at the level of the fisherman in terms of the effort expected to be expended on solpis production compared with their returns. Takendu and Sadler (1980), for example, remark that fishermen obtained approximately four times the price for smoked fish, and three times for fresh fish, than they did via the extra effort producing solpis;
the project appears to have been funded more from the point of view of the amount of money available rather than the amount of fish being produced. In this respect it is worth noting that “warning bells” were sounding in year one of the project, yet investment and expenditure continued as originally planned.

The problems with the solpis project arose mainly for economic reasons and from poor planning, over-optimism, over funding and a lack of appreciation of what was actually happening in the real situation.
The technical aspects of the project produced few problems. A fine quality produce was actually produced and the demand for solpis was high (provided the project could ship it to areas where it was required at a competitive price - which it could not). The people involved at the technical level are to be commended on their enthusiasm. The employment of intermediate technology (e.g. Richards 1979) in the project was regarded by Bardach (1977) as one of the best he had seen in developing countries. The ability to transport fish from this region to other areas was a major achievement, unfortunately, belittled by over-riding economic and social failure.
Despite the constraints outlined above, the project was still being expounded in 1984 as a real alternative to imported tinned fish (Stuart 1984).

2.4 Present freshwater fisheries developments - stocking practices

Coates (1985), on the basis of fishery surveys, noted the low yields achieved from the Sepik River compared with other regions and suggested that a major constraint to development might be the inappropriateness of the resident fish stocks. Coates also noted the success of previously “accidentally” introduced species, for example, Sepik tilapia presently account for half of the catch from floodplains. One logical way of improving the fishery might be to introduce further species of fish. However, there may be environmental considerations here, depending whether or not introduced species affected existing fishes. This whole problem was discussed at length by Coates (1987) who suggested that stocking be seriously and rationally evaluated with due regard to further research and adherence to codes of practice relating to species transfers. Stemming from this work the Sepik River Fish Stock Enhancement Project was formulated in 1986 and initiated in March 1987. The project, primarily funded by UNDP, is implemented jointly by FAO and the PNG Department of Fisheries and Marine Resources (DFMR). An up to date outline of the adopted code of practice concerning the transfer of freshwater fishes is provided by Turner (1988).

2.5 Present estimates of yield

The following is a brief summary of estimates of fisheries yield from PNG freshwaters. Due to a lack of sufficient data for many regions these should be viewed with caution. They are, however, the best estimates presently available and give a rough approximation of the situation.

2.5.1 The Sepik and Ramu Rivers

The only estimate of freshwater fisheries yield based on data from the Sepik is that of Coates (1985) who estimated present yield of the floodplain to be about 4,000 t per year. This is a rough figure since yields from floodplain rivers can vary greatly between years and there was some degree of variation in estimates between the various methods he used. This figure, however, is considered acceptable for planning purposes. About half of the present catch is tilapia, 25% native fork-tailed catfishes, and 25% two species of native gudgeons (Coates 1985, 1987). A yield of 4,000 t per year is a considerable amount of fish by PNG standards but this is caught by a large number of people (about 68,000). Hence, yields per fisherman and the yield of the floodplain in total are considered to be low.
This estimate applies to the floodplain region which is 0only about 10% of the catchment. Yields from tributary streams and rivers have not yet been estimated accurately. These regions, however, are known to have much reduced fish stocks (Van Zwieten 1989). Based on the assumption that yields might average about 10% outside the floodplain, approximately the same amount of fish might be being produced in these areas. This gives an estimate of total yield for the Sepik of about 8,000 t per year, perhaps less if stream yields are on average lower than assumed.
A similar calculation for the Ramu River might give an approximate figure of 1,500 t per year (although the area of the Ramu is quite large it is less heavily populated in the main fishing areas than the Sepik).
Sepik and Ramu fish yield is, therefore, estimated at about 9,500 t per year. Obviously these figures need refining considerably, especially in stream areas. This is presently being undertaken. These estimates are the best presently available.
There are some interesting findings arising from this analysis:

about 43 t of fresh fish were landed and sold in the small market at Angoram on the Sepik River in 1982 (Coates 1985). This small-scale organised local marketing is entirely self supporting, involving limited, if any, government funding support. Interestingly, this is over twice the maximum production through a government funded fish plant at Wewak (WAMA Marine) which might cost the government in the region of about K100,000 per year to run (excluding extensive investment) depending on how one analyses costs. A similar situation exists with comparisons amongst government run coastal fisheries stations. Such observations have important implications for the logic of government supported fish processing and marketing plants;
organised fish sales occur only at Angoram and Pagwi, the latter being much smaller. A limited amount of fish is also sold at village level. In this respect, organised marketing, which infers more “commercial fishing”, is negligible compared with the amount of fish directly consumed or used for barter. Consequently, inland fisheries are considered predominantly subsistence in nature, “commercial” activities are negligible at present;
the impact of the “solpis” project with its maximum yield of 40.3 t, usually and presently much less, is negligible compared with overall production. The investment of K770,000 in this project produced, on average, an improvement in catch of less than 1% which is considered negligible compared with natural fluctuations. There are, of course, other ways of looking at this activity since it was designed to generate cash earning opportunities and some more positive aspects of the project were outlined above;
9,500 t per year is a significant fishery. At an approximate value of K 1.00 per kilo, the Sepik and Ramu fishery is worth about K 9.5 million per year (about US$ 11 million). It does not, of course, generate any foreign exchange or other government revenue, but does reduce government costs through the consumption of home produced protein. Notably, the majority of the fishery involves no direct costs to the government. Excluding foreign exchange earnings, the Sepik and Ramu fishery might rival or even surpass all other PNG fisheries depending on the aspects of “importance” considered;
the fishery is government-independent, being a private initiative without any government inputs; such fisheries might, however, be used as a gauge by which to assess governmental fisheries developments in terms of increases in total fish catches and utilisation;
the resource limits yield from the Sepik and Ramu. This differs markedly from coastal and marine fisheries in PNG, which are generally socially/economically limited not resource limited. However, a most significant factor in yield estimates is the number of people fishing, despite the resource. Regardless of the fact that yields per person or unit area are low, inland fisheries take on enormous regional importance because of the high percentage of people involved with them;
because of the large number of people directly involved with the fishery, only modest improvements might have a significant national impact. This is a most important point. For example, if each person in the Sepik/Ramu catchment increases his catch by only an average of a modest 20 kg per year (= average of 50 grams per day) - the resulting increase in production would be in the region of 10,000 t yr^-1. This, incidentally, is in excess of present total coastal fisheries production in PNG and a doubling of present Sepik/Ramu catches - i.e. an improvement of 100%. An increase of only 5 grams per day would be an improvement of 10% - still a significant amount in fisheries development terms. It is, however, essential that improvements in this fashion apply to the whole, or most, of the population in question; which, for example, was a failing in the “solpis” project. Whilst increased catches from inland fisheries can be expected to vary from these hypothetical figures, the point of the argument is valid. It makes sense, if possible, to direct certain fisheries developments towards modest improvements for the majority rather than significant improvements to a limited number of people. Experience shows that, up to now, although limited on scale, improvements have only been achieved for a minority of the population and through high government subsidy.

2.5.2 Other freshwater regions

The Sepik and Ramu Rivers cover most of inland areas in northern mainland PNG. The major rivers in the south of the main island are the Fly and Purari systems. Despite some considerable attention to these rivers, primarily through mining or damming impact assessments, there is limited information on yields in the southern rivers. Estimates are further hindered by the fact that all southern rivers have a somewhat different fish fauna to the Sepik/Ramu. For example, the Fly River has approximately twice the number of fish species as the Sepik and several commercially and artisanally exploited Fly families of fish are absent from the Sepik (Allen and Coates 1989). The impact of tilapia in southern rivers (which is 50% of the Sepik/Ramu floodplain catch) is not known, and there are no records of this species occurring in the Fly, Purari or Kikori Rivers.
The rivers in southern PNG also have quite different morphological characteristics to the Sepik and Ramu rivers. In particular, deltas and estuarine zones are very extensive in southern PNG rivers but such areas are limited or non-existent in northern rivers. The Purari and Kikori Rivers have limited ox-bow lakes and floodplain areas, consisting, above the delta, of mainly river channels flowing through forest. Sepik/Ramu type floodplains, with extensive associated ox-bow lake systems, do occur in the Fly River above its extensive delta. Apart from the above mentioned differences between northern and southern rivers in terms of their fish faunas, southern rivers are also dominated by estuarine/mangrove fisheries in their lower sections and large floodplain fisheries tend to be absent except in the middle Fly.
Haines and Stevens (1983) provide some information on subsistence catches from the Purari/Kikori deltas. They estimated the annual subsistence catch of 20,000 people to be a minimum of 600 t of fish and 300 t of crabs. Amongst the fish, barramundi, sharks, threadfin salmon and catfishes are the major components of the catch. All of these crabs and fish (except catfishes) are not caught to any extent in the Sepik, if at all. Haines and Stevens' useful data clearly apply to a different type of fishery to the Sepik. For this reason such estimates from the Purari/Kikori areas are difficult to compare with Sepik/Ramu data. Haines and Stevens, however, note that Sepik catches are twice as high as their catch data. This was based on ADB hypothetical figures (they quote 8,000 t by 85,000 people in the Sepik-this is actually the tilapia production estimate). Based on the Coates (1985) figure of 4,000 t for Sepik floodplains and the loose comparison by Haines and Stevens, yields in the two areas might be similar. Note, however, that half of the Sepik yield is tilapia which does not feature in the southern fishery. The data of Haines and Stevens are biased towards the “estuarine” fishery; floodplain regions, the most productive area of the Sepik/Ramu, are absent from the Purari and Kikori Rivers.
Catches from the middle Fly river, the floodplain region, have been estimated to be between 2,000 and 5,000 t per year (Hortle 1989). Hortle also noted that Fly River floodplain and ox-bow lake gillnet catches were about twice as high as those from the Sepik recorded by Coates (1985). Although such direct comparisons of data between different river systems, at different periods and with different methods of assessment should be treated with caution, both Hortle's and Coates's estimates seem logical. Fly floodplains are less populated by people than those of the Sepik. Consequently, total Fly floodplain catch is in the region of total Sepik catch but Fly catches are perhaps twice as high on the basis of catch per net or per unit of effort. Higher catches from the Fly might be explained by this river having about twice the number of species, including the major commercially important groups - barramundi, clupeids and engraulids. Sepik/Ramu catches are also dominated by tilapia suggesting that comparisons of catches for native species alone would be higher for the Fly than the figure suggested by Hortle.
Most of the estimates of freshwater fish catches are rough approximations, sometimes simply guesses. Yield estimates, therefore, might relate more to numbers of people. Lowlands of the Fly and Purari are less heavily populated than those of the Sepik and Ramu, although highlands populations may be similar. On this basis, Coates guesses Fly/Purari freshwater fisheries yields (excluding estuaries/deltas) to be between 4,000 and 8,000 t per year, including highlands regions based on the same estimates as undertaken for Sepik highlands regions.
Freshwater yields from other southern regions might be about 500 t per year. The PNG Islands have more limited freshwater resources as major rivers are absent, lakes are poorly stocked and there are low populations of people. Although there is a significant extent of smaller rivers and streams, the islands people probably preferentially fish coastal resources. A guess of 500 t per year from islands freshwaters is made.
The above gives a very rough estimate in the region of 14,500 to 18,500 t per year from PNG freshwaters.
It is interesting to note that this figure (14,500 to 18,500 t) is in the region of the present coastal commercial, artisanal and subsistence fisheries production combined (excluding the industrial fisheries for tuna and prawns). However, about 87% of people in PNG live inland (Frielink 1983) as opposed to along the coast and there are, therefore, about 6.7 times more people inland. The above estimates and logic support the view that PNG inland fisheries may have low yields, especially for the Sepik/Ramu. Conclusions based from coastal fisheries, therefore, support the Coates (1985) estimate that Sepik/Ramu catches are much lower than that expected by comparison with other zoogeographic zones. Such broad estimates obviously need much refining but the logic suggests the figures presented here are valid approximations based on existing knowledge.

2.6 Potential catches

Estimates of potential catches achievable by comparison with data on, for example, African river systems, suggest a theoretical figure in the region of 100,000 t per year for all PNG freshwater capture fisheries (extrapolated from Coates 1985). This assumes that PNG freshwaters have both equivalent fish stocks and the potential for equivalent fishing effort to that found in other regions. PNG freshwater fisheries might suffer socio-economic constraints on their development to their full potential in the short-term but, more importantly, PNG freshwater fisheries are likely to be resource limited. Even if Fly/Purari lowland fish stocks could support catches achieved elsewhere, these areas have a low density of human population. The higher human population densities in Sepik/Ramu lowlands and all PNG highlands regions are all in areas where fish stocks are known to be particularly impoverished. It is concluded, therefore, that irrespective of the extent of resource limitation on Fly/Purari freshwater fisheries, overall, potential freshwater fisheries production in PNG is presently resource limited. This will be the major constraint to achieving the estimated potential production.
The Sepik River Fish Stock Enhancement Project is aimed at addressing freshwater fishery development problems in the Sepik and Ramu Rivers by examining the potential for reducing resource limitation through fish stock enhancement.

2.7 Summary of estimates of capture fishery yield

Table 2 summarises these estimates and guesses of freshwater fishery catches in PNG.
Table 2. Summary of estimates of fish catches from PNG freshwater (see text for explanations).
Previous catches

effectively the same as present (except increases arising through introduced species)

Present catches

Sepik floodplain 4,000 t per year (good estimate)
Sepik and Ramu total catchment 9,500 t per year (reasonable estimate)
Other areas 5,000 to 9,000 t per year (guess based on Sepik and Fly data)
Total freshwater production about 14,500 to 18,500 t per year (rough estimate).

Note: Almost all is subsistence production. Commercial production is negligible.
Potential catches
Based on a theoretical comparison with other regions (e.g. Africa), this should be in the order of 100,000 t per year. This assumes equivalent fishing effort and fish stocks. A major constraint to achieving this potential is the impoverished fish stocks in the Sepik and Ramu lowlands and all highlands regions in PNG.

3. AQUACULTURE

Except clam culture in Manus Island (Maclean 1978) and modest fish culture on Bougainville Island, traditional aquaculture is non-existent in PNG. There have been numerous attempts at aquaculture in both marine and freshwater environments. None of these have been successful either economically or socially.

3.1 Freshwater aquaculture

Interest in freshwater aquaculture started in the early 1950s. The indigenous freshwater fish fauna of PNG was assumed to be poor, particularly in the highlands. The solution to this was thought to be to bring in various species of non-native species for culture in ponds (Schuster 1951, Van Pel 1956). Rapson (1957) and Filewood (1967) provide briefs on the approaches at the time. Numerous species were tried, 29 species introductions are listed by West and Glucksman (1976), although not all for aquaculture, many were not successful either for environmental reasons or because of the lack of fish farming skills (Table 1).
The only significant impacts of these introductions have been the inadvertent production of valuable wild stocks. Tilapia escaped from ponds, where they were never a success, into the Sepik River system where they now account for over half of the yield of the fishery (Coates 1985). Tilapia also formed the basis of the only commercial inland fisheries venture, salting and drying fillets (the “solpis” project). Common carp were brought to PNG to replace tilapia in pond trials and again these escaped into the rivers. In the Sepik River they are increasing in abundance and providing a significant part of subsistence catches (Coates 1984, Ulaiwi 1989).
Present freshwater aquaculture activities (including rearing and stocking) are as follows:

3.1.1 Crocodiles (Crocodilus porosus and C. novaeguinae)

There is significant crocodile farming in PNG. For the present purposes this is not regarded as aquaculture and the activity does not come under the jurisdiction of DFMR. Freshwater fish are used to feed the crocodiles in some areas. It is a potential method of utilisation of waste fish.

3.1.2 Rainbow trout (Salmo gairdneri)

Trout were introduced in 1949 to stock highland streams, primarily for the colonial angling fraternity. In 1971, a trout hatchery was constructed at Mendi to support this fishery. After independence (1975) trout continued to be introduced but with the emphasis on stocking streams for local people to fish. Recently two small trout hatcheries operated on this basis at various times, at Mendi and at Keglsugl (Mount Wilhelm). Eggs are imported from Australia, hatched and fingerlings released. Both of these hatcheries are funded from the government purse. Stocking has occured, and still occurs, on a smaller scale at other locations on a less “official” basis. The economics of this activity have never been studied and in particular the following questions never addressed:

are trout the correct fish to use ? They were originally brought to PNG for an elite sport-fishery. The objectives of stocking have since changed. Trout, since they are predators eating other species of fish, are also well known to have detrimental effects on native fish species in other countries;
what happens to the trout after they have been stocked ? There is no information available on the utilisation of trout in highland streams by local people. Without this it is impossible to do any costversus benefit analysis of this government funded activity;
do trout self-reproduce in highland streams ? This is not known. If they do - why continually stock them ? If they don't - why are trout being used ?

There have been attempts at commercial trout farming. In 1970, a “private” trout farm was established at Goroka (Kotuni Trout Farm) which in effect was funded from the government purse. It operated until about 1984 but never showed a profit and generally a substantial loss. It produced government subsidised fish to a minority market. There are regular suggestions to start commercial trout farming again (either at Goroka or numerous other highlands locations).
Kotuni trout farm was re-opened in early 1989 and renamed “Gana Trout Farm”. This is supposedly run under the auspices of a private business group but in fact is being funded as before, from government funds. Presently (September 1989) the farm estimates that they have about 80,000 fingerlings about 25 g in size. The business group involved have now exhausted their funds and cash-flow problems suggest this resurrection of the activity is also doomed to failure.
Some pertinent factors relevant to commercial trout farming in PNG are:

there appear to be no technical problems. The water temperature, quality, flow etc. appear to be satisfactory;
economic constraints are considerable and include:

high cost of feeds - trout need a high protein diet and fish feeds need to be imported. There are presently no sources of locally produced high protein wastes in PNG;
high costs of transportation and marketing;
limited market outlets;
lack of suitably trained and motivated manpower;

It is worth mentioning that attempts have often been made to disguise these factors. In particular to look at the “export markets” for trout which are non-existent based on production costs in PNG, unless the government wishes to subsidise this activity.

3.1.3 Common carp (Cyprinus carpio)

Carp introductions began in 1958/9 with a view to replacing tilapia (Oreochromis mossambicus) and gouramy (Osphronemus gouramy) that had been disappointing in highland fish ponds. A carp breeding and distribution centre was established at Aiyura (Eastern Highlands Province). By the mid-1970's the carp project was considered a failure (further details in Haines 1978, Uwate 1984). However, in 1983, the carp project was re-initiated at Aiyura following reports by Paton (1982) and Sagom and Paton (1982).
The modest carp breeding station at Aiyura is intended to produce carp fingerlings to be distributed to villagers for use in fish ponds. It is essentially a subsistence aquaculture project. In addition to fingerlings production, research is being undertaken on the utilisation of potential waste feeds (e.g. coffee pulp) that might be converted into fish protein and on integrated fish farming (e.g. using carp, ducks and pigs).
In 1986, technical assistance for the carp project was obtained through FAO (FI:TCP/PNG/4503). Pitt (1986) made proposals for the improvement of facilities at the Highlands Aquaculture Development Centre, Aiyura. Tacon (1986) made technical recommendations concerning the development of carp feeds, particularly utilising local wastes. Kovari (1986) reported on the investigation of the re-establishment of carp fishing in the highlands (although “fishing” is perhaps a misnomer since the project actually aims at aquaculture not capture fisheries). There are presently suggestions to implement these findings through up-grading of the facilities with further fundings.
Despite these previous technical inputs, several fundamental problems with this aquaculture project have not been addressed. These relate to the need for development and application of technical skills in the aquaculture field; regular feed supply; attention to market outlets and economics etc. Such conditions are not presently available and, therefore, requirements for successful aquaculture are not presently being met. These are essentially the same problems that prompted Kearney (1975) to conclude that the prospects for aquaculture in PNG had been regarded over-optimistically and in fact the prospects were not bright. This view was shared by others (e.g. Haines and Kelleher 1979). Despite over 20 years of trials, most aquaculture activities were abandoned by the late 1970's, yet remnants continue with the Aiyura station and Gana Trout Farm.
The establishment of common carp culture needs to clarify the following before it can proceed with full, and hopefully more successful implementation of the project:

how to teach people to culture fish when they have no tradition of fish culture and limited knowledge of animal husbandry of any kind. Perhaps more importantly, how to promote the attitudes and approaches necessary for successful fish culture;
how many people in highlands regions are willing to undertake aquaculture and especially to quantify the amount of area suitable for aquaculture?;
what has happened in the past with the carp project and particularly how much biomass of carp was produced and actually eaten - in other words, what happens to the carp once they leave Aiyura and what impact do they have on nutrition and the economy of the highlands ?

Without the above information it is not possible to determine the costs versus benefits of carp culture. It is also worthy of note that no attention has ever been given to the economics of aquaculture versusalternative activities, particularly agriculture or increases in yields via stocking rivers and other natural water bodies.
The re-establishment and up-grading of the carp project is apparently in response to a demand in the highlands for carp fingerlings (Kovari 1986) which has not been quantified nor viewed on an economic basis. This perceived demand requires clarification also. It is probably a demand for fish, not necessarily carp produced in ponds.
Recent attention has been directed to solving technical problems with carp production whilst the fundamental problems with aquaculture in general are left unanswered. It also appears, in this instance, that the various consultants in question were not asked to address these problems but to give only technical advice to DFMR on how to culture carp at Aiyura.

3.2 Marine aquaculture

There have been various attempts at oyster culture which failed by the mid-1950s (Uwate 1984). Further culture trials were done with some limited technical success. Two pearl projects were initiated in the mid-1960s. One was a village scheme at Milne Bay, and the other was an Australian-Japanese joint-venture at Fairfax Harbour (Port Moresby). Initial results were promising, but by the mid-1970s, the Fairfax Harbour operation had closed for reasons of economics and pollution (Uwate 1984), but neither problem was quantified. The Milne-Bay operation developed into a multi-village activity. Financial problems resulted in the drastic decline of the Milne Bay industry by the 1980s. A few villages were still culturing pearl oysters in 1982 (Lock 1982) - remnants of the previously much larger fishery. Lock (1982) should be consulted for further information on this activity.
Essentially, the problems encountered with marine aquaculture have been the same as those already listed for freshwater aquaculture.
The only known presently existing marine aquaculture activity is research being undertaken on the culture and stocking in the wild of giant clams. The PNG activities of this project are mainly undertaken by The University of PNG under a regional Australian Council for International Agricultural Research project. Research has been undertaken mainly on the technology of clam culture (e.g. Gwyther and Munro 1981). From a practical PNG fisheries point of view, it appears that the need for clam re-stocking in PNG, the present levels of exploitation and the economics of this activity have not been studied. This activity relates more to stocking than aquaculture which is involved only in the early stages of clam production.

3.3 General

Delmendo (1979) was consulted to address the prospects and potential for aquaculture development in Papua New Guinea. This was at a time when, after over twenty years of effort, the Fisheries Department had effectively abandoned all attempts in this field (Kearney 1975). Delmendo's report relates more to the technical feasibility of aquaculture. Although she recognised the limitations already known at the time (and as outlined above) the problem of how to account for these factors was not addressed. Kearney (1976) had already warned of the dangers of over-optimism with the prospects for aquaculture in PNG.
Kan (1981, 1982) gives further brief information relating to aquaculture in PNG.
The most comprehensive studies of the prospects for aquaculture in PNG are those of Uwate et al. (1984), Uwate and Kunzatuba (1984) and Uwate (1984). These result from an extensive survey of aquaculture in the Pacific Islands and should be consulted by all those interested in aquaculture development (there are extensive references to PNG). It is useful to list the conclusions of Uwate (1984):

aquaculture development projects have had little success in the Pacific Islands region, even when outside consultants were used;
most Pacific Island countries have had consultants visit and have had feasibility studies and resource assessments completed;
the potential for aquaculture as espoused by consultants often only relates to technical feasibility, not to economic development;
business and management skills have been limited among both expatriates and locals in aquaculture projects;
in some cases, advanced phases of a project have been initiated even though the objectives of the initial phases were never met; and
projects with social value have been run at a financial loss that has been subsidised.

Critical issues that surfaced included:

other possible motivational factors for large public expenditures in the area of aquaculture;
the question of who really benefits from aid projects;
missing components (such as training, economic analysis, and commercial development) in aquaculture projects;
unrealistic estimates used in proposals and projects; and
the possibility that the Southeast Asian experience may not fit the Pacific Islands region.

Uwate (1984) mentions that it would be easy to conclude that aquaculture is not viable for the entire Pacific Islands region but qualifies this by saying a blanket “no” for all aquaculture may not be justified. Specific cases may work. He further suggests a case by case basis for examining aquaculture projects.

3.4 Summary of aquaculture production

The following is a summary of aquaculture production in PNG:
Previous production: limited
Present production: negligible
Potential for production increase: technically feasible but at present considerable economic and sociological constraints exist for full aquaculture development.

4. DISCUSSION

There has been much interest in the field of inland fisheries development over the past thirty years. Much of the activity has been based on limited data and information on PNG inland fisheries. Previous activities do, however, reflect the interest within the country in inland fisheries development. Such interest stems from an appreciation of the importance of such fisheries in view of about 87% of the population living inland. Previous activities have also reflected a demand for fish in inland areas.
Successful previous inland fisheries developments are hard to find, if not non-existent. The only significant success has been the inadvertent entry of tilapia into the Sepik and Ramu River systems. Tilapia formed the foundation for the only commercial fishery ever to develop based on a freshwater fish. Recently, the inadvertent entry of common carp into the same river systems has also produced a substantial spin-off via the establishment of wild stocks in natural water bodies.
In PNG, research has suggested that inland fisheries may be resource limited, especially in the Sepik/Ramu. By comparison, analyses of marine fisheries suggest these are primarily socio-economically limited: major limitations here include lack of transportation facilities, market outlets, high handling and processing costs, generally widely dispersed consumer populations and other economic factors. An appreciation of the general fisheries situation in PNG leads to two important conclusions in respect to the Sepik River Fish Stock Enhancement Project:

it is becoming increasingly more evident that the greatest practical benefits in terms of increased domestic fish utilisation will arise from fisheries projects with minimal capital and infrastructure inputs aimed at modest improvements but applying to a large number of people;
although some inland fisheries are presently resource limited this problem may be addressed by a suitable stocking strategy.

The resultant higher levels of fish stocks available through stocking could then begin to be exploited more effectively. Inland fisheries development might encounter problems of exploitation, processing, handling, transportation and marketing etc. inherent in most fisheries activities within the region.
In the short term, the priorities under the Sepik project are to improve the subsistence fishery and, hence, protein intake and availability in under-privileged areas. Commercial yields can only be predicted when enhanced stocks become available. The development of a successful and important Sepik commercial fishery based on sufficient stocks might be expected to be a lengthy process. It is, however, obvious that commercial fisheries are unlikely to develop without a reasonable resource there in the first place. To attempt significant developments without a sound resource may damage the existing stocks perhaps irreparably. Previous attempts at commercial developments in the Sepik have also shown disappointing yields and economic returns (ref. solpis project).
Aquaculture should be regarded as a parallel activity to stocking natural waters in order to improve fish protein availability on a wide base. Aquaculture development may involve the importation of further exotic species more suited for this purpose and these should be regarded as having the potential to establish populations in natural waters.
PNG has had thirty years experience with aquaculture projects, yet the only significant benefits in terms of fish production have arisen through the establishment of wild stocks of fish. Further attention to socioeconomic considerations regarding aquaculture in PNG might alleviate some of the problems previously experienced but this is likely to be a difficult process.
Attention to the previous history of aquaculture and freshwater fisheries in PNG can leave little doubt that the greatest immediate potential benefits to improving fish availability in the Sepik and Ramu Rivers will arise through fish stocking. The Sepik River Fish Stock Enhancement Project recommends that any future fish introductions be undertaken under a carefully considered scientific rationale and on an environmentally sound basis. Guidelines on established procedures and a code of practice for the transfer and introduction of freshwater fishes in this fashion have been produced by the European Inland Fisheries Advisory Committee and the International Council for the Exploration of the Sea (see Turner 1988) and are being followed by the project.

5. REFERENCES

(Note: all those marked unpublished are available through the DFMR fisheries bibliography in the Dept. Library)

Allen, G.R. and Coates, D. (1989) An ichthyological survey of the Sepik River, Papua New Guinea. Records of the Western Australian Museum. In press.

Anon. (1980) Freshwater Crayfish. U.S. Investment Mission Supplementary Information. Department of Primary Industry, Fisheries Division. Unpublished.

Anon. (1989) Fisheries Sector Survey Papua New Guinea. Report prepared for the Government of Papua New Guinea on behalf of the United Nations Development Programme. Project PNG/88/004/A/01/31. (ANZDEC Ltd. Aukland). Pages variable. Two volumes. Unpublished.

Bardach, J.E. (1977) The fisheries of Papua New Guinea: promises and problems. University of Hawaii, unpublished report.

Coates, D. (1984) The occurrence, spread and potential effects of common carp, Cyprinus carpio L., in the Sepik River. Department of Primary Industry, Fisheries Research and Surveys Branch, Report 1984–13.

Coates, D. (1985) Fish yield estimates for the Sepik River, Papua New Guinea, a large floodplain system east of “Wallace's Line”. J. Fish Biol. 27: 431–443.

Coates, D. (1986) Inland Fisheries in Papua New Guinea. In: Reports and papers presented at the Indo-Pacific Fisheries Commission Expert Consultation on Inland Fisheries of the larger Indo-Pacific Islands, Bangkok, Thailand, 4–9 August, 1986 (Ed. T. Petr). FAO Fisheries Report No. 371 (Suppl.): 119–129.

Coates, D. (1987) Consideration of fish introductions into the Sepik River, Papua New Guinea. Aquaculture and Fisheries Management 18: 231–241.

Delmendo, M.N. (1979) Aquaculture development in Papua New Guinea: Prospects and potentials. UNDP Technical Assistance Report. Unpublished.

FAO (1982) Developing the freshwater crayfish industry in the Fly River system. FAO FI/TCP/PNG/6702.

Filewood, L.W.C. (1967) Fishpond culture in the New Guinea Highlands. Department of Agriculture Stock and Fisheries Report. Unpublished.

Frielink, A.B. (1983) The rural coastal population of Papua New Guinea. Department of Primary Industry, Fisheries Research Report, 83–11.

Gwyther, J. and Munro, J.L. (1981) Spawning induction and rearing of larvae of Tridacnid clams (Bivalvia: Tridacnidae). Aquaculture 24: 197–217.

Haines, A.K. (1978) Highlands Aquaculture. Department of Primary Industry, Fisheries Report. Unpublished.

Haines, A.K. and Kelleher, M.K. (1979) Fisheries in the Highlands: prospects and proposals. Department of Primary Industry, Fisheries Internal Report. Unpublished.

Haines, A.K. and Stevens, R.N. (1983) Subsistence and commercial fisheries. In: The Purari - tropical environment of a high rainfall river basin (ed. T. Petr). Mongraphiae Biologicae 51. Dr. W. Junk, The Hague, pp 385–408.

Hortle, K. (1989) Potential introductions of Fly River fishes to the Sepik River system. Presented at the Workshop on Introduced and Translocated Fishes and their Ecological Effects. Australian Society for Fish Biology, Magnetic Island, Townsville 24–25 August, 1989. In press.

Kan, T.T. (1981) Some aspects of the state of aquaculture in Papua New Guinea. Conference on Economic Development and Growth of Papua New Guinea. University of Technology, Lae. Unpublished.

Kan, T.T. (1983) The state of aquaculture in Papua New Guinea. In: Report of an International Conference on Development and Management of Tropical Living Aquatic Resources. Malaysia. Unpublished.

Kearney, R.E. (1975) The prospects for fisheries development in Papua New Guinea. Paper for the Waigani Seminar, Port Moresby, May, 1975. University of Papua New Guinea. Unpublished.

Kearney, R. E. (1976) The expansion of fisheries in Papua New Guinea: prospects and problems. Report for Waigani Seminar, May, 1976. University of Papua New Guinea. Unpublished.

Kovari, J. (1986) Investigations into the re-establishment of carp fishing in the Highlands. FAO: FI: TCP/4503. Field document 2. 15p.

Lock, J. M. (1982) Bivalve culture in Papua New Guinea. In: Bivalve Culture in Asia and the Pacific. Proceedings of the IDRC Workshop, Singapore, February, 1982.

Lock, J. M. and Waites, D. C. (1985) Papua New Guinea Fisheries Bibliography. Department of Primary Industry, Technical Report 85/3. 102p.

Maclean, J. L. (1978). The clam gardens of Manus. Harvest 4 (3): 160–163.

Mitchell, M. (1972) Sepik fisheries - the big potential. Harvest 2 (3): 102–106.

Mitchell, D. S., Petr, T. and Viner, A. B. (1980) The water fern Salvinia molesta in the Sepik River, Papua New Guinea. Environmental Conservation 7: 115–22.

Osborne, P. L. (ed) (1987) A draft inventory of wetlands in Papua New Guinea. Department of Environment and Conservation, Papua New Guinea. 163 p.

Paton, J. (1982) Report on a visit to Aiyura fish farm. Internal Department of Primary Industry, Fisheries Report. Unpublished.

Petr, T. (1983) Limnology of the Purari basin. Part 1. The catchment above the delta. In: The Purari - tropical environment of a high rainfall river basin (Ed. T. Petr). Dr. W. Junk, The Hague. Monographiae Biologicae 51: 141 – 178.

Pitt, R. M. (1986) Carp cultivation and the Highlands Aquaculture Development Centre, Aiyura. FAO: TCP/PNG/4503 Field Document 1. 108p.

Rapson, A. M. (1957) Coastal and Highland freshwater fishery problems of Papua New Guinea. Indo-Pacific Fisheries Council, 7th Session. C57/Tech. 64.

Richards, A. H. (1979) A polythene tent fish dryer for use in Papua New Guinea's Sepik River salt fish industry. Department of Primary Industry, Fisheries Department internal report. Unpublished.

Roberts, T. R. (1978) An ichthyological study of the Fly River in Papua New Guinea with descriptions of new species. Smithsonian Contributions to Zoology 281. 72p.

Rodwell, L. J. (1982) Economic assessment of the East Sepik fisheries project. Department of Primary Industry, Fisheries report. Unpublished.

Room, P. M. and Thomas, P. A. (1985) Nitrogen and establishment of a beetle for biological control of the floating weed Salvinia in Papua New Guinea. Journal of Applied Ecology 22: 139–56.

Sagom, P. and Paton, J. (1982) State of carp production in the Eastern Highlands. Internal Department of Primary Industry, Fisheries Division Report. Unpublished.

Schuster, W. H. (1951) A survey of inland fisheries of the territory of New Guinea and Papua. Australian Journal of Marine and Freshwater Research 2 (2): 226–238.

Stuart, P. (1984) Sepik Project helps dam the canned fish tide. Pacific Islands Monthly, September 1984: 37.

Takendu, D. R. and Sadler, G. (1980) Internal review of the East Sepik Rural Development Project. Department of Primary Industry, unpublished.

Tacon, A. G. J. (1986) Development of Carp Feeds. FAO: FI: TCP/PNG/4503. Field Document 3. 42p.

Thorpe, J. P. (1977) East Sepik rural development project, inland fisheries situation report. July-December, 1979. Department of Agriculture Stock and Fisheries Report. Unpublished.

Thorpe, J. P. (1980) Review of the inland fisheries project. Department of Primary Industry, Fisheries Report. Unpublished.

Turner, G. E. (1988) Codes of practice and manual of procedures for consideration of introductions and transfers of marine and freshwater organisms. EIFAC/CECPI Occasional paper No. 23. 43 p. (FAO).

Ulaiwi, K. W. (1989) The occurrence and spread of common carp, Cyprinus carpio (L.) in the Sepik River, Papua New Guinea. Proceedings of the Second Asian Fisheries Forum, Tokyo, Japan, April 1989 (Asian Fisheries Society). In Press.

Uwate, R. K. (1984) Aquaculture assessment project - final report. Pacific Islands Development Programme, East-West Centre, Hawaii. 76p.

Uwate, R. K. and Kunzatuba, P. (1984) Aquaculture development in the Pacific Islands Region. Pacific Islands Development Programme. East-West Centre, Hawaii. 87p.

Uwate, R. K., Kunzatuba, P, Raobati, B. and Tenakanai, C. (1984) A review of aquaculture activities in the Pacific Islands Region. Pacific Islands Development Programme. East-West Centre, Hawaii. 112p.

Van Pel, H. (1956) Notes and suggestions on the development of fisheries in the territory of Papua New Guinea. South Pacific Commission, Noumea, 18p.

Van Zwieten, P. A. M. (1989) Preliminary analysis of biomass, density and distribution of fish in tributaries and hillstreams of the Sepik-Ramu river system (Papua New Guinea). Proceedings of the Second Asian Fisheries Forum, Tokyo, Japan, April, 1989. In press.

West, G. J. and Glucksman, J. (1976) Introduction and distribution of exotic fish in Papua New Guinea. Papua New Guinea Agriculture Journal 27: 19–48.

Wilson, P. T. (undated). A report on the fisheries situation to the Government of Papua New Guinea with recommendations for development. Unpublished.

Sigabaduru