Data Requirements for Fisheries Management in the Tonle Sap

"No data is better than misleading data."

Nicolaas VAN ZALINGE
MRC Fisheries Program
Phnom Penh, Cambodia

Cambodia has very few income generating possibilities beyond its natural resources and is economically almost fully dependent on agriculture, forestry and fisheries. The country's inland fisheries support a thriving industry of great economic and social importance and have a potentially bright future.

Cambodia's freshwater capture fisheries production of over 400 000 tonnes per year is large, even by world standards. It may be among the world's ten largest producers.

The most recent estimates of the National Institute of Statistics in Cambodia indicate fisheries contribute 16 percent to the national GDP.

The Tonle Sap contributes over half of the fish produced in the country. More than 1.2 million people in the Tonle Sap area alone depend on fishing for their livelihoods.

An MRC/DoF socio-economic survey of 4.2 million people living in central Cambodia estimated that average fish consumption was 67 kg/person per year (in fresh weight equivalents, 1995/96 data (Ahmed et al. 1998). There is no other food supply as readily available or as inexpensive that can replace fish in the diet of the Cambodian people.

Management of these important resources is a crucial matter. It requires an understanding of the issues at stake and reasonably accurate statistics are critical to a proper perspective. The work done by MRC/DoF in the Tonle Sap area since 1994 provides a basis to discuss what we can hope fisheries management could realistically achieve and how the needed information could be collected.

The Tonle Sap fishery

Since the Great Lake in Cambodia was formed some 5 000 to 6 000 years ago (Carbonnel 1963 in Rainboth 1996) it must have abounded with fish. The rise of the Khmer Angkor Empire may, to a large extent, have been possible due to the availability of a rich fishery resource and well-developed rice irrigation schemes. The abundance of fish pictured on the reliefs of the Bayon and Angkor Wat temples and the proximity of the temple complex to the Great Lake in Siem Reap province are testimony to this.

Fishing Lots in 2001 (Lot areas: red; Flood plains: light blue)

The combination of rice and fish is still the staple food for the great majority of Cambodians. Recognizing the value of the fisheries, the French colonizers modeled their taxation system on the traditional royal fund-raising practice of issuing leases for fishing lots, introducing the first fishery laws of the country (Petillot, 1911 in Van Zalinge et al. 2000).

Petillot also reported that in 1910 about 50 000 tonnes were exported in the form of dried, salted and live fish, as well as fish oil and paste. In the 1920s and 1930s exporting dried fish to Java was a big business. Chevey and Le Poulain (1940) reported that an average of 25 000 tons was shipped annually from Cambodia mainly via Singapore by Chinese traders. Given a fresh-to-dried fish ratio of 3 to 1 (Chhouk, 1996 in Van Zalinge et al. 2000) this corresponds to 75 000 tonnes of fresh fish. Chevey and Le Poulain (1940) estimated the total fish yield of the Tonle Sap at 125 000 tonnes per year. At present, this trade no longer exists, although similar quantities are being exported to Thailand and Viet Nam, mostly as fresh, dried and smoked fish or as paste and sauce.

The main reason for the enormous wealth of fish in Cambodia is the monsoon, which each year swells the rivers and creates a flood of water that inundates the highly productive floodplains. The temporary access to enormous quantities of food from a wide range of natural habitats drives the huge production of fish. White fish have evolved to synchronize their time of spawning with the onset of the monsoon so that fry and juveniles are ready to enter the plains as they flood. Black fish spawn and feed in the inundated floodplains. Without the floods and the floodplains, the fish catch would be only a small fraction of what it is now.

The Tonle Sap floodplain at maximum inundation varies considerably in size from year-to-year (roughly between 10 000-15 000 km2). Thus, in a dry year (e.g. 1998-1999) fish production is much less than in a wet year (e.g. 2000-2001). This is illustrated for the dai fishery in Figure 1 below. The relationship between the maximum flood level of the season and the fish catch shows that a permanent lowering of the average peak flood levels (e.g. due to flood controls) would result in a more than proportionally lower fish catch. Among other variables, fish productivity is related to the extent of floodplain inundation. Thus, flood controls like dams, irrigation canals, river canalization and diversions have a negative effect as they lower peak flood levels.

Figure 1: Productivity and Water Level

Natural floodplain habitats like the flooded forests have the highest productivity and species diversity. Hence, flooded forest destruction or conversion to rice fields also has a negative effect.

Species composition of the catch

Long-distance migratory species or 'white' fish (about 60% of total catch)

Migrations take place annually between the spawning areas in southern Lao PDR and northeastern Cambodia and the floodplains around the Tonle Sap, south of Phnom Penh and the Vietnamese portion of the Mekong Delta and back. Larger species tend to spawn later in life. Many large species have dramatically declined in number, some nearly to extinction, such as the famous Mekong Giant Catfish Pangasianodon gigas and the Giant Barb Catlocarpio siamensis. The Catfish is reported to spawn for the first time at a weight of 150-250 kg by which time it may be six or more years old (Pholprasith and Tavarutmaneegul, 1997). Very few individuals survive the heavy fishing pressure long enough to reach sexual maturity. In the year 2000, eleven Giant Catfish were caught in the dai fishery and only seven in 2001. Thus, the later a species matures the more vulnerable it is to overfishing.

In the dry season, illegal fishing with explosives takes place in the deep pools and channels of the Mekong River in the northeast of Cambodia. This further reduces the spawning populations of some of the bigger species. Smaller species are usually early spawners. Most have not declined and dominate present catches. A good example is the Cyprinid, Trey Riel, Henicorhynchus siamensis. It spawns for the first time when about one year old. As most of the larger species are predators, a decrease in their number leads to an increased survival rate of their smaller prey. Smaller species are not overfished and could even be fished with greater intensity.

Short-distance migratory species or black fish (nearly 40 per cent of total catch)

Movements are much more limited, usually from flooded forests to lakes and rivers and back. Stocks have probably not declined as these fish do not run the same gauntlet of fishing gear as the long-distance migratory species. Snakeheads (Channa spp.) are the most important species group. They spawn in the flooded forests and are the most valuable species in the catch of the Great Lake fishing lots.

Fishery management system

Fishery Laws: The 1987 Fiat Law is still in force. Many of the regulations are largely based on colonial legislation. The fisheries of Cambodia can be divided in two broad categories: limited and open-access fisheries.

Limited access fisheries: the fishing lot system

The most productive part of the Cambodian fisheries has been privatized for more than a century through a system of government leases on fishing lots. The rest is open-access.

Resource rent: In the recent past, the lot system provided over US$ 2 million annually in tax revenues and more in an informal way. The open-access fisheries, however, do not contribute to public taxes.

Fishing lots vary from a simple anchoring position (dai) in the Tonle Sap to a large area of floodplain. The value of the lot depends on the expected fish production. Many lots occupy relatively large areas of floodplain (the largest Great Lake lot is 500 km2). The Tonle Sap Great Lake lots contain mostly natural habitats, but there are also rice fields and sometimes villages within their boundaries. The natural habitats comprise flooded forests, shrub forest and grasslands, which are essential for the feeding and breeding of many fish species. In the past, there were close to 300 lots but now there are 164 covering approximately 30% of the floodplain area that was occupied in 1919.

In the recent past the open-access areas witnessed a rapid expansion of fishing effort in waters outside the lots. Catch rates were falling and this has caused an increase in conflicts over access to the fish resources. Many conflicts between lot operators and local villagers ensued (Degen and Loeung, 2000). The government intervened on behalf of the family fishers and further reduced the size of the lots. Community fisheries management is encouraged in the freed-up areas. It is still too early to judge what impact community management might have.

Open-access fisheries

Open-access fisheries have expanded dramatically in the past two decades and have contributed to the recent increase in fishing pressure. Close to 200 different types of fishing gear and methods are used in these fisheries (Degen et al. 2002). The majority of the fishers in the Tonle Sap Great Lake area (the 1995- 1996 estimate was 1.2 million) are engaged in these open-access fisheries. Most fishers are living at the edges of the floodplain, but quite a number have adjusted their lifestyle to 'living with the floods' by creating floating villages or houses on tall stilts.

Middle-scale fisheries: A number of gear types specified by the fishery law require a license, such as gillnets, seines and arrow-shaped traps.

Family or small-scale fisheries: The remaining gear types are free for anybody to use, although not everywhere or at any time. These gear types include small castnets, small dipnets, small gillnets and certain types of traps. Rice field fisheries fall into this category.

Illegal fisheries: A number of types of gear and methods have been declared illegal, such as brush parks, explosives, poison and electrified gear.

Over-exploitation

The high intensity of fishing operations in the Tonle Sap area supports a general impression that the system is overfished. This impression is strengthened by anecdotal evidence from many fishers who claim that their catches have been decreasing over time. However, overall catches are probably higher now than in the past, although individual catch rates have declined because the increase in population and number of fishers outstripped the increase in catch (Table 1). At species level the situation is more complicated.

Table 1 Tonle Sap Great Lake Region

Due to the reduction of larger fish species in the catch and the shift to smaller sizes, the average value per kilogram has decreased. Thus, not only has the catch rate per fisher dropped, the value of his catch has decreased as well. Nevertheless, the overall tonnage of fish caught is still increasing. A number of larger species are overfished, but smaller species are not overfished at all. Multi-species river systems cannot be intensively exploited without loss of the larger species in the fish population, which are less abundant and reproduce more slowly (Welcomme, 2001).

During the fishing-down process, fishing effort on the large and medium-sized fish species increases to levels above what allows for the maximum sustainable yield of these species. Meanwhile, the yields of smaller species may be still growing and thereby increase total fisheries output. Species usually do not become extinct and the potential for fish productivity does not diminish as long as the natural habitats remain intact and the average level of flooding remains stable. The fishing-down process is illustrated in Figure 2. In 1940, the Tonle Sap Great Lake region catch of 125 000 tonnes consisted mainly of large and medium-sized fish, while the 1995-1996 catch of 235 000 tonnes contained hardly any large fish and was dominated by small fish.

Figure 2: Catch versus fishing effort

Fisheries management problems in the Tonle Sap stem from lack of governance and public sector reform, which hinges on two main issues: 1) the legal framework is still inadequate, especially with regard to land tenure and community fisheries legislation is not yet in force and 2) government staff (DoF, military, police, commune heads) are not being paid adequate wages, hence they have to use the power of their authority to make ends meet.

History of recent data collection in the Tonle Sap

From 1980 to 1998, government-produced statistics were generated for internal reporting on progress and for planning. They did not accurately reflect catch levels and could not be used for fisheries management purposes. Between 1995 and 1997, more accurate estimates of catch levels (280 000 to 445 000 tonnes) were produced by the MRC/DoF/DANIDA Project for Management of the Freshwater Capture Fisheries of Cambodia and these figures were subsequently used by the government. At present, only the dai fishery in the Tonle Sap River is annually monitored. There is a great need for cost effective and comprehensive data collection on catches and exports.

Need for accurate fishery information

Accurate information and statistics is necessary for balanced government planning and decision making in national or regional water resources management because:

1) Fishery resources are vulnerable to upstream river interventions

2) The fishery is vitally important for food security and livelihoods of the rural poor

3) Governments need to know the true contribution of fisheries to the national economy (presently estimated at 16% of GDP in Cambodia), to employment creation and to foreign exchange earnings through exports

4) Lack of knowledge of the true value of the wild fisheries could result in permitting or even stimulating agricultural expansion in prime fish producing floodplain habitats to alleviate perceived food shortages

Attempts to improve fisheries data collection systems

Catch assessment system

In the catch assessment system, catches were estimated on the basis of CPUE by gear type and frame survey data. Initially, the large-scale fisheries (fishing lots, barrages and dais) and middle-scale fisheries (licensed mobile gear and traps) were targeted. The plan was to expand data collection to include family fisheries. It was hoped that this would provide the basis for an improved fisheries statistics gathering system in Cambodia. However, this system was not expanded to cover the family fisheries and was actually stopped. Main reasons for this decision were:

Failure to collect accurate data on the lot fisheries. Neither lot operators nor officials were sufficiently prepared to cooperate. The only exception is the collection of data on the dai fishery in the Tonle Sap River.

Collecting more accurate data on the middle-scale and family fisheries with the personnel available proved to be unsuccessful. As the use of most fishing gear is seasonal, regular counts of gear in operation (frame surveys) need to be carried out. The great variety, dispersed nature and often-patchy distribution of the gear make catch assessment a prohibitively cumbersome and expensive exercise.

The cost of running a catch assessment system is currently beyond the means of the DoF.

Only the 'dai' fishery has been successfully monitored since 1995. Regular monitoring has revealed the variability of these stocks in relation to the extent of the annual inundation making it possible to detect long-term trends in migratory fish populations.

Household survey

The household survey was limited in area coverage but successfully estimated fish consumption (67 kg/capita/year), as well as fish catches made by the middle-scale and family fisheries, including gear use and species composition information. In fact, the present estimate of some 300 000-400 000 tonnes produced annually in Cambodia is based on the results of this survey.

Recent small test surveys carried out in a few communes over a period of several months produced average fish consumption rates that were very close to the ones found in the household survey. In the latter survey, randomly chosen households were interviewed only once.

When the survey was set up (1994 -1995), it was limited to districts and communes, which were close to water bodies in only eight of Cambodia's 17 inland water provinces. Fisher communities in the major fishery provinces with a total population of 2.4 million people were directly targeted. Consequently, the data cannot be extrapolated to the rest of the country. During the survey, more than 5 000 interviews were conducted over a relatively short period of time (four months) with teams of four to seven data collectors per province. However, data entry and analysis took a relatively long time.

The fish yield per unit of habitat approach

Another rough way of gauging the overall catch level in the country would be to estimate fish yield per unit of habitat type (flooded forest, secondary flooded forest, grasslands, marshes, rice fields, etc.) and to use land cover information to determine the extent of these habitats. The open nature of the terrain makes the estimates rather imprecise. Using one ha enclosures may give a low estimate of the standing stock in a certain habitat and misses out on the productivity over time. Adding up all fish catches made in the Tonle Sap floodplains, lakes and rivers and dividing this by the surface area gives an overall yield of 139-190 kg/ha (Lieng Sopha and Van Zalinge, 2001). This figure is the average yield of all the habitats found here.

Conclusions and recommendations

Data collection is costly and no data is better than misleading data. The question then arises: "What could be a realistic focus for fisheries management in Cambodia under the present conditions and what data are the minimum data needed to support this?"

Fisheries management focus

Given the continuing increase in population, it can be expected that fishery resources will experience proportionally increasing pressure. There will be a trend towards greater use of the floodplains for agriculture, as well as an increase in fishing effort as a result of the need for employment. Limiting access to fisheries would be one choice. However, the political reality is that government favours opening up access to the fishing grounds and decreasing the use of fishing lots.

The best option for the protection of the resources and to limit species extinctions is to focus on maintaining the essential floodplain habitats and breeding grounds and their connectivity as best as possible.

First, further degradation of the floodplain habitats and their conversion to agricultural use should be stopped. Upstream breeding grounds (deep channels) should be protected and connectivity should be maintained (no mainstream dams). Secondly, the present flooding regime of the plains needs to be maintained if present fishing levels are to be sustained. Damming may change this.

To achieve this, a twofold government action is required. First, the government should try to engage more of the many NGOs in Cambodia to support a fisheries management approach and work with floodplain and upstream communities. Community involvement in the management of fishery resources is still in its infancy. Even though major efforts to facilitate co-management are still to come, success in protecting fisheries resources is not guaranteed.

Second, the government needs to strongly defend the nation's fish wealth in regional fora dealing with the future use of the Mekong's water resources.

Data requirements and collection methodologies

The following are considered to be the essential items of information needed to support this approach to fisheries management in the Tonle Sap. Appropriate collection methodologies are suggested.

To be collected at regular 5-year intervals:

Economic value of the fisheries resources by fishing method and species.

To determine the economic value of the fisheries resources a different approach needs to be taken to fishing concessions and open-access fisheries. Fishing concessions are notoriously difficult to monitor. Staff should be very carefully selected.

Floodplain lots: Stratification may be possible using habitat distribution in drainage areas. Monitoring during the main fishing period is essential. This period is often limited to about 4 months around the Great Lake (March to June).

Barrage lots: Targeting migratory fish in the early stage of operations (October to March), thereafter sweeping channels (April to June).

Dai lots: A stratified sampling scheme is worked out and operational. Stratification needs updating.

Open-access: In our experience, assessment through a stratified multi-staged random household fish consumption survey will be sufficiently accurate and cost effective. The survey should be carried out once during the best season for fishing (November to February) and again during the low season (May to June).

Among other factors, fish consumption, sources of supply, fishing involvement, gear use and catch composition should be assessed. A manual including a questionnaire is available (Van Acker, 2000).

Fish consumption and livelihoods

Household fish consumption as above and land use surveys, especially in the floodplains. The last Landsat and aerial photography survey dates from 1992 - 1993. An update is under preparation. More work needs to be done on developing yield-per-habitat indicators as a monitoring tool.

To be collected on an annual basis:

Exports and fish prices, preferably through continuous monitoring.

A strategy has yet to be developed.

Indicator fisheries

On migratory fish: The performance of migratory stocks over time can be followed by continuing the annual monitoring of the dai fishery in the Tonle Sap River as these stocks migrate out of the floodplains of the Great Lake to the Mekong to survive the dry season and for breeding.

On non-migratory fish: Snakehead (Channa spp.) is the most valuable species and makes up the main catch of the lake lots. The fisheries are presently not being monitored regularly.

Flood levels

Gauges in Kampong Luong and Kampong Chhnang are good indicators of the level of inundation of the floodplains and hence of fish production. Other gauges along the Mekong River, especially in Pakse, are indicators of long-term trends in water discharge during the wet months. The long-term trend over the period 1924-1998 is negative, hence floodplain inundation and fish productivity must have been larger in the past.

References

Ahmed, M., Hab Navy, Ly Vuthy and M. Tiengco, 1998. Socio-economic assessment of freshwater capture fisheries of Cambodia: A report on a household survey. Mekong River Commission Secretariat, Phnom Penh. 185 pp.

Chevey, P. and F. Le Poulain, 1940. La pÍche dans les eaux douces du Cambodge. Travaux de L'institut Océanographique de l'Indochine. 5e Mémoire, Gouvernement GÈnÈrale de l'Indochine, Saigon. 241 pp.

Degen, P., F. Van Acker, N. Van Zalinge, Nao Thuok and Deap Loeung, 2000. Taken for granted: Conflicts over Cambodia's freshwater fish resources. Presented at the IASCP Common Property Conference, Indiana, USA, 31 May - 4 June 2000.

Degen, P. and Deap Loeung (Compilers). 2002. Cambodian gear catalogue. In preperation.

Lieng Sopha and N. Van Zalinge, 2001. Fish yield estimation in the floodplains of the Tonle Sap Great Lake and River, Cambodia. Inland Fisheries Research and Development Institute of Cambodia, IFReDI Fisheries Technical Paper Series, Volume III, p.23-26.

NIS, 2000. Report on the Cambodian socio-economic survey 1999. National Institute of Statistics, Ministry of Planning, Phnom Penh, Cambodia. 156 pp.

Pholprasith, S. and P. Tavarutmaneegul, 1997. Biology and culture of the Mekong giant catfish, Pangasianodon gigas. Extension Paper 31. National Inland Fisheries Institute, Bangkok. 79 pp. [in Thai].

Rainboth, W. J. 1996. Fishes of the Cambodian Mekong. Mekong River Commission Secretariat and Food and Agriculture Organization of the United Nations, Rome. 265 pp.

Van Acker, F., 2000. Cambodia. Fish consumption at household level survey. (Manual). Unpublished report. Cambodian Capture Fisheries Project, Mekong River Commission Secretariat, Phnom Penh.

Van Zalinge, N., T. Nao, T. S. Touch and Deap Loeung, 2000. Where there is water, there is fish. Cambodian fisheries issues in a Lower Mekong Basin perspective, |p. 37-48 In: M. Ahmed and P. Hirsch (eds.). Common property in the Mekong: Issues of sustainability and subsistence. ICLARM Studies and Reviews 26, 67 pp.

Van Zalinge, N. P., Nao Thuok and Sam Nuov, 2001. Status of the Cambodian Inland Capture Fisheries Sector with special reference to the Tonle Sap Great Lake. Inland Fisheries Research and Development Institute of Cambodia, IFReDI Fisheries Technical Paper Series, Volume III, p. 10-17.

Welcomme R. L., 2001. Inland fisheries: ecology and management. Food and Agriculture Organization of the United Nations. Fishing News Books, 358 pp.