Previous Page Table of Contents Next Page


Cambodia - some relevant facts
Cambodia and inland fisheries
Objectives of the study

Cambodia is the smallest country of continental Southeast Asia. It has two dominant geographical features: the Mekong and Bassac rivers system of the central lowland plains, and the mountain ranges near its borders. The Cardamon and Damrei mountains are situated in the southwest, the Dangrek range constitutes a large part of the northwestern border with Thailand, and the Darlac Plateau makes up part of the eastern border with Vietnam. Cambodia has a mainland coastline of 557 km on the Gulf of Thailand.

The Mekong river and the Tonle Sap lake are the main hydrological structures of the Cambodian lowland plains. The junction of the Mekong with the Tonle Sap channel constitutes the onset of the lower Mekong delta. During the rainy season, the large volumes of water from the upper and middle reaches of the Mekong find their way in two opposite directions, partly flowing northwest in the Tonle Sap channel, forming an inland delta towards the Tonle Sap lake, but mainly flowing southeast, forming the Mekong delta and finally discharging into the South China Sea.

The renewable natural resources of Cambodia are closely linked to the country’s geography. The Tonle Sap lake derives exceptional productivity from pulsed flooding through its position as part of the lower Mekong system where it functions as a natural overflow basin. The combination of cyclic flooding, soil conditions, natural vegetation of the flood area and high levels of biodiversity has enabled the development of this exceptional ecosystem. However, the dynamic processes that have contributed to its formation are at the same time the source of concern and problems for contemporary inhabitants and likely also the origin of its decline, in the long run succumbing to siltation.

King (1860) described the lowland area between the Tonle Sap lake and the surrounding mountain ranges as “diversified with forest and open prairie”. The borders of the lake are covered with seasonally inundated forest and other types of vegetation. The lowland flats have always provided the most suitable conditions for human settlements. The open grasslands here offer opportunities for agriculture, while the vicinity of forests, both in the mountainous areas and in the lowland plains, provides a source of wood and other forest products. Central in the flats lies the Tonle Sap lake, which is one of the most productive freshwater ecosystems in the world (Rainboth 1996). In 1860, King wrote, “during the months of January, February, and March, when the water is drained off the surrounding country, the lake appears alive with fish, and the inhabitants collect large quantities of them.”

Fig. 1 shows the present-day location of villages in Cambodia. The concentration of people along the Tonle Sap lake and in the central area around the capital Phnom Penh is remarkable. The site where the ancient Khmer civilization reached its highest peaks of power and domination during the Angkor period (eighth-thirteenth century), is precisely situated there where intensive rice culture was possible, and where forest products and lake resources were available in an optimal blend. Sculptures on the walls of constructions at Angkor testify to this. Other factors, such as protection against invaders using the Mekong to penetrate deep into the continent, played a determining role for the selection of the site near the present-day town of Siem Reap among several options around the lake offering similar food production opportunities (Hoskin and Hopkins, 1991).

Figure 1. Location of present-day villages in Cambodia. Each dot represents one village.

Intensive rice culture in the Angkorian era was possible thanks to the construction of large water reservoirs and an extensive irrigation system, allowing for three harvests per year, even though most of the soils of the lowland are not very suited for rice culture. The irrigation system collapsed with the decline of the Angkor civilization, and has been restored only to a small extent in recent times. This loss of technology and the consequent low rice production capacity has increased the dependency of the population on other food sources, in particular aquatic products of the lake and of other water bodies like small streams and rice fields.

Cambodia - some relevant facts

The Kingdom of Cambodia is administratively organized in 24 provinces, including four municipalities. The provinces are divided into 183 districts, comprising 1 609 communes. The communes are made up of 13 406 villages.

The 1998 general population census (NIS, 1999) indicates that the total population in that year was about 11.4 million, growing at 2.49 percent per year, and half of them are considered economically active. Cambodia has a young population with over half of it younger than 18, and only 3.5 percent is over 65. In 1995, life expectancy at birth for both sexes was 52.9 years (FAO, 1999a). There are on average 5.1 persons in each of the 2.16 million households, a quarter of which is headed by women. The overall literacy rate among people of age seven and above is 71 percent for men and 55.4 percent for women. Fewer than one third of the households have access to safe drinking water, and for more than 95 percent of the households firewood or charcoal is the main type of fuel used for cooking.

At the 1996-1998 average annual exponential population growth rate of 2.49 percent (NIS, 1999), the present population of Cambodia is projected to double by the year 2028 (Fig. 2).

Figure 2. Cambodia population prospects 1998-2030

The gross domestic product (GDP) of the country is estimated at US$300 per person per year, ranking Cambodia among the poorest countries in the world. Being a predominantly agrarian society, with a rural population that is over five times bigger than the urban population, agriculture is the main occupation and is believed to contribute about half of GDP.

Food distribution remains a problem for a large part of the population, resulting in structural food insecurity at the national level, and acute or chronic food insecurity particularly in rural areas. This problem of availability, together with problems of access to and utilization of food, as well as the lack of diversity in the diet of most people, results in high levels of child and adult malnutrition, even though national food supplies are adequate. Five to fifteen percent of the households can be described as food insecure; another 15 to 20 percent manage to maintain minimum consumption levels albeit at a high cost-indebtedness and by recourse to risky coping strategies.

Several surveys (FAO, 1999a) confirm that the food consumption of the typical Cambodian consists primarily of rice with some fish or fish products. Over the last 30 years, the share of total energy from carbohydrates (rice) has declined (from 83 percent to 77 percent) while the contribution of protein to the overall energy supply has not changed. The main source of protein in rural areas is reported to be prahoc, which is fermented fish paste. A UNICEF/WFP survey (Kenefick, 1998) shows that fish is part of the daily diet of 74-80 percent of all children of over 11 months old, while nearly all (95 percent) reported to have eaten rice the previous day. Insufficient animal protein consumption and associated forms of malnutrition are linked to the degree of poverty. Apart from rice, fish was found to be by far the most common food type consumed in all regions of the country (FAO, 1999a).

In large areas, 82 percent of all children aged between 6 and 59 months are reported to be suffering from anaemia. Malnutrition continues to be a major problem in this age group for the whole country, with over 56 percent affected by some form of malnutrition-related growth disturbance. Among women of between 15 and 49 years of age, almost one out of three is malnourished. The immediate causes for malnutrition in children are low food intake and chronic illnesses, which are mutually reinforcing.

Cambodia and inland fisheries

i. Fish production and consumption
ii. Contribution of inland fisheries to the national economy
iii. Fish production and food security

Fish is one of the principal natural resources of Cambodia, together with its forests, the richness of the soil (gems, oil/gas) and agriculture production. The fish resources of the country are plentiful and diverse.

i. Fish production and consumption

Accurate assessments or even estimates of consumption of fish in Cambodia are difficult to make. The total annual production originates from a wide variety of sources and is incompletely reported. The quantity of fish that is exported is not precisely known, nor is the portion of the catch that is not used for human consumption but fed to husbanded animals like caged fish, crocodiles, pigs and other. Spoilage figures are not documented but could be as low as 10 percent (Csavas et al. 1994). Fish consumption tends to vary geographically (Nam and Thuok, 1999, Ahmed et al. 1998; Demuynck, 1995).

These uncertainties, combined with inaccurate population figures, have resulted in the past in considerable variation in annual per caput consumption estimates. However, the general population census of 1998 and the improvement in recent years of the quality and scope of the data on captured fish allow now a more realistic estimation.

Figure 3. Officially reported annual fish production and exports in Cambodia, 1961-1999. (Source: FAO, except for 1999 figures from the Department of Fisheries)

The officially reported Cambodian production and export of fish over the past 40 years is shown in Fig. 3. The dramatic increase in production in 1999 is due to the inclusion for the first time of data on fish catches from inland family fisheries and rice field fishing. A total of 231 000 tonnes from inland capture fisheries and 15 000 tonnes in aquaculture production is reported (Table 1).

Table 1. Total fisheries production for 1999 (Source: Department of Fisheries)


Production (tonne)

Inland Fisheries:

231 000

· Large and medium scale commercial fisheries1

71 000

· Family fisheries

115 000

· Rice field fisheries

45 000

Marine Fisheries

38 000

Total Capture Fisheries

269 000


15 000

Total Fisheries Production

284 000

1 Data for September 1998-May 1999 fishing season. The first four months of the 1999-2000 season are reported to have yielded 35 615 tonnes, which is almost three times the volume reported for the same period of the 1998-.

2 Including 50 tonnes of shrimp.

The most recent estimates (Jensen 2000c; van Zalinge and Thuok, 1999) put the annual freshwater fish catches at between 290 000 and 430 000 tonnes. Together with the marine catch and aquaculture production, this results in an estimated total annual fish production for the country of between 343 000 and 483 000 tonnes. The export of fish products in 1999 amounted to 48 742 tonnes (Department of Fisheries). The reported import of fish in 1997 was negligible (4 tonnes) and no change was reported for 1999.

Based on the 1999 estimated fish availability and the calculated total population of 11 722 454, the maximum consumption of fish in Cambodia would have been between 25.2 and 37.1 kg.year-1.caput-1. The real figures are probably lower, given post-harvest losses and uses other than human consumption. In contrast to this, a household survey of fishing communities in the vicinity of main fishing waters carried out by Mekong River Commission (MRC) (Ahmed et al. 1998) puts the annual equivalent consumption of fresh fish as high as 75.6 kg.year-1.caput-1(consumption of processed products was reported as their equivalents in fresh fish. This seems very high, the more as it would leave the rest of the Cambodian population with fewer than 16 kg.year-1.caput-1 equivalent of fresh fish in the current most optimistic estimate of fish availability. These data support the indication of considerable regional differences in fish consumption. Past estimations of annual per person consumption of fish in Cambodia based on official data were as low as 13.3 kg.year-1.caput-1 (Csavas et al. 1994).

With these new population data and best-effort estimations of the annual fish production, it is now possible to update past predictions of fish consumption and requirements (e.g. Csavas et al. 1994; Nam and Thuok, 1999). Requirements based on the assumption that consumption patterns remain unchanged are presented in Table 2. However, these requirements will depend on a number of factors such as the level of urbanization development and availability of alternative food sources.

Table 2. Current consumption and projected requirements of fresh fish at different levels of demand in Cambodia, 2000-2030. (The shaded areas indicate where the highest estimation of the present production does not meet projected demand.)


Projected population (millions)

Future annual fish requirements (1 000 tonnes)

consumption levels (kg.year-1.caput-1)















































1 Based on most recent official statistics.

2 Mean between the high (37.1) and low (25.2) estimations, corresponding to an annual freshwater fish production of 360 000 tonnes.

ii. Contribution of inland fisheries to the national economy

There is little doubt about the considerable contribution of inland fisheries to the national economy of Cambodia. Traditional Khmer villagers pattern their livelihood around rice growing and fishing (Ahmed et al., 1998). The inland capture fisheries contribute between 80 and 90 percent of the total fisheries production, and about two thirds of that is produced and caught in the Tonle Sap ecosystem - 40 percent of the total in the lake and 26 percent in the channel (World Bank, 1995; Csavas et al., 1994).

The contribution to the national economy lies in the domestic generation of food, tradable goods and employment. Fish is an essential part of the Cambodian diet and constitutes a part of the food security for most individuals, households and communities (FAO, 1999a; Kenefick, 1998; MAFF, 1996). In 1999, reported exports of fish and fish products to neighbouring countries amounted to almost 50 000 tonnes but are in reality probably higher since not all exports are declared. Exported fishery products are usually of high value and represent a considerable source of foreign currency. More than three quarters of Cambodia’s active population is employed in the primary sector, i.e. agriculture, hunting and fishing (NIS, 1999). However, it is difficult to separate these people along occupational lines. Over one million households are believed to be involved in fishing at least some time of the year (World Bank, 1995).

The total value of the annual fish production is subject to the same uncertainties as those described for total production. Furthermore, there is the aspect of the price of fish and fish products. Not only are there large variations in the value depending on the species, size and freshness of the fish, there are also large overall seasonal variations that are related to the seasonal fishing activity and fluctuating supply of fish to the markets. The domestic prices for marine species are generally lower than for freshwater fish, resulting from a preference for the latter among the Khmer people (Csavas et al., 1994).

Fish in urban areas is usually more expensive than in rural markets, and increasing distance to the fishing grounds - and therefore in most cases an increasing number of intermediaries - tends to push retail prices up. Actual prices paid for fish are available from a number of sources (Demuynck, 1996; NIS, 2000; Mekong Secretariat, 1992). There is large difference between the prices paid at the landing sites or by intermediaries of between 200 and 3 000 Riel/kg according to the species (Ho, 1999) and the retail prices of 5 000 to 6 000 Riel/kg quoted in the Consumer Price Index for fresh fish (NIS, 2000).

The lack of a more precise valuation of the fisheries makes them vulnerable to underestimation. This is particularly risky at policy level where decisions are to be made that may have an important and possibly irreversible impact on the inland fisheries, and for which the value is an issue of consideration. An example of this is the possible construction of dams on the main stream of the Mekong, for which the associated economical benefits will be considered against environmental and social implications such as impact on the inland fisheries. With the past underestimation of total production and value, and the current uncertainty in this matter, there is a real possibility that policymakers will not have all the arguments required to reach responsible conclusions. Furthermore, undervaluation of the production makes it more susceptible to misappropriation.

The issuance of licences for the large and middle-scale fisheries generates a revenue of about US$2 million annually for the government (van Zalinge et al., 1999). The Department of Fisheries reported for 1999 an official revenue from the fisheries sector of just over US$3 million.

Table 3. Contribution of fishery to GDP according to different recent sources (US$1 = 3 800 Riel)

Total annual fish production (tonne)

Average fish price (US$/kg)

Total value (US$ million)

Percentage of GDP (1998)





about 5

World Bank 1995

> 1.2 million


900-1 000


Jensen 2000b/Combined data for all four Mekong (MRC) countries; value at retail market

400 000




Jensen 2000b, c

295-420 000



van Zalinge et al. 1999/Value based on landing site prices

284 000




Department of Fisheries, Nam and Thuok 1999





S. Nuov, personal communication, 2000/Percentages as cited





Mekong Secretariat 1992/Commercial and subsistence fisheries; GDP of 1992 applies

The total value of the fish production would increase if product quality management and marketing efficiency could be improved (van Zalinge et al., 1999). The actual contribution to the GDP is considerably higher than most of the estimates so far if retail prices were used instead of prices obtained by the fishers.

iii. Fish production and food security

Food security for Cambodians means in the first place rice security. The verb for “to eat” in Khmer translates literally as ‘to eat rice’, and this is where most of the attention and efforts are focusing on. The contribution of fish to the national food security appears to be only recognized as such by those involved in the fisheries sector.

Food security in Cambodia has three components of concern (MAFF, 1996; FAO, 1999a):

The precise role of fish and fish products in the food security of Cambodia is difficult to assess. As with all other issues for which proper quantification of the volume is required, the lack of accurate production and availability figures makes it impossible to quantify with some degree of confidence the contribution of capture fisheries and aquaculture to food security. In spite of this, numerous statements have been made regarding the importance of fish for food security in Cambodia at all levels, mostly without proper substantiation. Some “data” are widely quoted but have an obscure origin and, leading a life of their own, they have become self-endorsing. Results from surveys tend to be biased by local consumption patterns and usually are not representative for the whole of the country (e.g. Ahmed et al., 1998; FAO, 1999a).

On the other hand, the anecdotal evidence of the availability and consumption of fish and fish products in most places and the high level of accessibility of fish for food (low price, common property resource, low investment in fishing gear, low opportunity cost), indicate that this is an important contributor to food security, especially at individual and household level, keeping in mind the constraints to food security mentioned above (availability, accessibility and diet diversification). Access to fish appears possible for almost all, even the poorest and least apt.

The weakness of the data, and the historic underreporting of the availability of fish for food in the country contribute to the present lack of acknowledgement of the importance of fish for national food security. Although the importance of fisheries is acknowledged, this is usually not reflected in subsequent policy statements and actions proposed to maintain and improve food security.

The resolutions of the National Seminar on Food Security and Nutrition in Cambodia (20-21 April 1999) do not directly address the issue of fisheries and aquaculture as a (possibly important) contributing factor to food security. At the national level, they mention the strengthening of extension and intensification services to introduce improved production practices for all agriculture sectors and fisheries. It is not clear how this will apply to capture fisheries. At the community level, some of the adopted resolutions may have an impact on the preservation and durable exploitation of the capture fisheries, albeit mostly indirectly (environmental awareness promotion, improved transport infrastructure and education on fishery laws and regulations). The resolution whereby the Royal Government commits itself to zoning the land for agriculture, forest production and forest conservation has large potential for the conservation of capture fisheries but it does not mention the sector at all. Attention is almost exclusively oriented at improving aquaculture productivity though aquaculture makes up only a fraction of the production of the capture fisheries.

Objectives of the study

Gillnets are the most popular of all gear used for subsistence fishing, and their use is mainly limited to this sector. Several sources report the majority of fish caught for subsistence purposes is obtained by gillnets.

Until very recently, official fisheries statistics provided data on the commercial and middle-scale fishery sectors only and did not include catch figures from non-licensed subsistence fishing activities.

However, the subsistence sector is an important part of the inland fisheries, not only of the Tonle Sap but also in most other water bodies. Its annual production is estimated to be at least equal and probably even higher than that of the licensed sectors of the inland fisheries. The very nature of subsistence fishing and the widespread scale at which it is practised make it play a major role in constituting food security at household and community levels. In many other cases, non-licensed fishing provides dietary variation and supplementation.

Little information is available on the practice of gillnet fishing in the Tonle Sap lake or in the inland waters of Cambodia. In 1996-1997, a study was undertaken in Siem Reap province to assess the capacity of gillnets as an important fishing gear in different habitats of the Tonle Sap lake, and to evaluate the environmental links of gillnet fisheries. The results are presented in this case study, which has two objectives. The first is to describe the different habitats as identified by type of macrophyte vegetation in the Tonle Sap lake and the floodplain in terms of basic limnological parameters; the second, to assess gillnet catches in different habitats in the Tonle Sap lake and the floodplain in Siem Reap.


i. Study area and characterization of the habitats
ii. Identification of habitat types by macrophyte vegetation and flood regime
iii. Selection and location of the sampling areas and sites
iv. Sampling procedure

i. Study area and characterization of the habitats

The study was carried out in two communes, Chong Kneas and Kampong Kleang, bordering the Tonle Sap lake in the northwestern province of Siem Reap.

The area in which gillnet fishing was studied consists of the lake and of what usually is described as the floodplain of the Tonle Sap lake. The area in Siem Reap province that gets flooded annually is about 1 780 km2.

Figure 4 Map of the lake-bordering districts in Siem Reap province. The sampling areas are shaded

For the gillnet fishery activities, both the permanent habitats in the core area of the lake and the temporary aquatic habitats in the floodplain are relevant. Due to cyclic flooding, most habitats in the Tonle Sap lake and the floodplain are determined by physical and biological factors, many of which show extreme variations over time.

Two factors are particularly relevant for the characterization of the habitats in this area: the water level and the macrophyte vegetation. Apart from the mere existence of the aquatic habitats, these factors largely determine the spatial structure of the habitats as well as their accessibility for gillnet fishing. Furthermore, they are the source of much of the variation of the ecological parameters which characterize these habitats. The habitats selected for this study are those types which are based on water level and macrophyte vegetation are distinctive, and that are present in the study area. Eight different types of habitats which make up most of the study area were identified. Other parameters of relevance (e.g. for migratory fish species) such as distance to the permanent water body, degree of vegetation patchiness and human activities are not considered as such.

Nowadays, human activity plays a determining role in the evolution of the vegetation of the floodplain. For many centuries, the pristine primary forest in the flood area has been a major source of wood and non-wood forest products and continues to be so. The collection of wood for use as fuel, construction material, fishing gear, etc. has led in large areas to a clearing of nearly all tree trunks, resulting in secondary vegetation types of scrubland consisting of shrub morphs and lianas. Only where functional in situ (e.g. as wave and wind breakers or as mooring places) or in areas of low population density that of larger trees can still be found. With the changing morphology and composition of the vegetation, the exploitation of the flooded forest has shifted accordingly. The constant pruning of the larger sprouts maintains the shrub morphs. Alternative sources of timber and wood products for the riparian population in Siem Reap province are distant in the northern hills.

In many places, the impact of flooded forest exploitation has gone much further than just degradation to secondary vegetation types. Often the vegetation has been and still is being cleared with the use of slash-and-burn techniques, resulting in barren soils on which young-grasses vegetation offer good grazing opportunities for cattle. In some cases where the flood regime permits, these soils are taken into culture for seasonal crops, mainly mung beans, water melon and flood recession rice.

The determining influence of human activity on most of the habitats of the floodplain implies important conservation and management issues. The ecosystem is insufficiently understood to assess with a high degree of accuracy the impact of ongoing activities and trends, and therefore the cautionary principle should guide the future exploitation of these natural resources. The risk of irreversible damage and subsequent degradation of the highly productive ecosystem is real.

The high productivity of the ecosystem is likely due to its unique combination of pulsed flooding with loaded waters and a large and rich floodplain, and it is becoming more apparent what the impact will be of further transformations of an important component of this system - the flooded forest - and how this may affect the productivity. Most human interventions so far have led to a loss of diversity, stability and biomass, and presumably also to a reduction of primary and secondary productivity of the flooded vegetation.

ii. Identification of habitat types by macrophyte vegetation and flood regime

The description of the habitats coincides in part with that of the vegetation types as identified by McDonald and Veasna (1996) for the Preach Sramoach area in Kampong Kleang district.

The description of the habitat types does not take into consideration factors like density or height of the vegetation, or soil conditions, making allowance for some variation in these factors. Some habitats are relatively stable sub-climax stages while others are pioneer or transitional or artificial entities that display little stability and require sustained disturbance. This approach to habitat description does not explicitly take into account other components such as plankton or aquatic birds, even though these often form a specific and important part of the ecosystem and some of them are also of considerable human interest.

A list of the macrophyte species that are dominant in each vegetation type is given in Annex 1. The following habitat types were identified:

· Scrubland

The secondary scrublands of inundated forest can be characterized as a usually dense and almost impenetrable tangle of sprawling shrubs and lianas that reach 3-5 m in height. The canopy is closed, and the understorey is devoid of herbaceous vegetation. The soil is covered by a very thin (1-2 cm) humus layer during the dry season. Only seedlings of the dominant shrubs appear in the shaded understorey, but these are sparse, owing to the absence of sunlight. The vegetation is dominated by euphorbs and legumes. The density and the height of the vegetation vary but the overall characteristics do not change. This vegetation type probably accounts for the majority of the plant cover surrounding Tonle Sap lake. It is present throughout the flood zone, except in areas that are permanently flooded or near the permanent water line. This vegetation presents large, elaborate and varied organic spatial structures, both hard and soft, with large surface areas. Depending on the extent and the duration of submersion, many of the shrubs appear to lose most of their leaves during the flooding. It is not clear whether this is abscission or a passive process as a consequence of the submersion.

Figure 5. Habitat types. From left to right, top to bottom: scrubland, grassland, floodplain pool, lotus field, rice field, forest, lake coastal and lake pelagic









· Grassland

The ephemeral grasslands of burned inundated forest are formed when secondary scrublands are managed by the use of fire. This is usually done to open new grazing areas for cattle or for use as agriculture land. Almost no native arborescent tree species survive this disturbance and after the burning a pioneer vegetation emerges which is dominated by weeds and grasses. This situation is transitional, after a couple of years the more robust shrubs shade them out. This type of vegetation can be found wherever there is scrubland.

· Floodplain pool

Unmanaged reservoirs of different size occur in the floodplain after the recession of the lake water. A small number of species (five to ten) of aquatic vegetation is found in these pools. Sedges and grasses predominate on the edges and in the canals, while the common floating, submerged and emergent plants include species such as Lemna minor, Pistia stratiotes, Azolla imbricata, Utricularia aurea and Eichhornia crassipes. Locally, the floating aquatic plants form thick, dense mats of vegetation. The depth of these reservoirs varies but all are shallow, with maximum depths of at least 1 metre.

· Lotus field

Large areas of permanent or nearly permanent water away from the core area of the lake are used for the cultivation of dense stands of lotus (Nelumbo nucifera). In these areas, vestiges of a once rich aquatic flora (c. 23 species) can be found, but overall there is little vegetation other than the lotus. Young lotus plants are collected from the wild and transplanted into these stretches of remaining water in the floodplain in February or March. The harvest of leaves and flowers starts in April and the seeds are collected in August before the water rises. The stems of the plants remain after the fruit is harvested but apart from these, very few spatial structures are present in these water bodies, as most other vegetation is shadowed out by the lotus leaves.

· Rice field

The rice fields in the floodplain are used for flood-recession dry-season rice culture. Small temporary dykes are built in many places to retain the floodwater longer. The rice is harvested once the floodwaters receded long enough at a particular place to make rice transplanting possible. The rice is harvested before the paddy-fields become flooded again. The fields are not ploughed before the flooding, and after the harvest, the stubble, about one half metre tall, remain on the land. Some of the harvested rice fields are briefly used for grazing ducks or cattle.

· Forest

On the border of the permanent part of the lake, in some places, scrubland with some taller trees can be found. The vegetation species composition is the same as that of the scrubland. Their wave and wind-breaking role is probably the reason why the larger forms are not cut or pruned to the size of the other scrub. They offer the only shelter opportunities for fishers out on the lake guarding their traps or nets. Recent surveys in the floodplain around the Tonle Sap lake have shown that there still are extensive stretches of well-preserved primary forest (A. McDonald, personal communication, 1997). The tall trees in it have large trunks but in most cases the crown has been constantly pruned. The habitat described here as forest refers to the small fringe of large trees on the edge of the lake in the selected area that, at least from a spatial point of view, resembles this primary forest. There is no primary forest remaining in the floodplain of Siem Reap and Sotr Nikum districts.

The above-mentioned habitats are all situated in the floodplain and the area is annually flooded by water from the Tonle Sap lake.

· Tonle Sap lake, coastal and pelagic habitats

The aquatic vegetation of the Tonle Sap lake is not well known. At times, thick dense mats of floating vegetation are formed. Most of the year, the lake is characterized by strong winds and large waves, which make the use of gillnets often difficult or impossible. A distinction is being made between the sampling in the coastal zone (within 2 km from the low-water shoreline) and the sampling farther out in the pelagic zone.

iii. Selection and location of the sampling areas and sites

Two sampling areas were chosen, one east of the seven villages of Chong Kneas commune, Siem Reap district, and the other near Preach Sramoach, Kampong Kleang commune, Sotr Nikum district (Fig. 4).

The villages of Chong Kneas are closely linked to the water of the Tonle Sap lake. There are no permanent buildings, and some of the villages consist entirely of boathouses. These floating villages are moving according to the water level, and so do the villages that are built along the dyke supporting the road that heads southeast from Phnom Krom towards the lake. The area around the villages is characterized by different stages of regenerating primary forest and agricultural land. Chong Kneas is the port area for the town of Siem Reap. Pollution levels from port activities like ship traffic and associated small businesses are high. High population density and high levels of small engine-powered boat traffic make the shallow stagnant water during the dry season highly turbid and chemically and organically polluted. The pollution seems to be mostly limited to the immediate surroundings of the villages.

In contrast to the villages of Chong Kneas, most constructions in Kampong Kleang are permanent. Most buildings are either on high grounds or on long stilts. Most of the sampling sites in Kampong Kleang were situated in an area protected for this purpose, with restricted access for local users. This protected area covered a surface of approximately 2 600 ha, and was situated west of the road between Damdek, the district centre of Sotr Nikum, and the commune centre of Kampong Kleang. During the rainy season, it got entirely flooded by the rising lake waters. There is no permanent human settlement inside the area but it is used by inhabitants of neighbouring villages for fishing and forest products collection. To make the area accessible, about 20 km of 3-metre wide strips were cleared throughout the area at 500 m intervals along north-south transects. These strips permitted to move in the area at almost any time of the year, by boat or on foot. They were used in this study to place gillnets as they often offered the only possibilities to submerge the nets in the scrubland without loosing all prospects of successfully retrieving them.

Figure 6. Distribution map of the habitats in the sampling area in Kampong Kleang, based on the vegetation map of the protected area by Malleux (1997)


The habitats and the fish populations in Kampong Kleang are less disturbed by nearby human population and port activities and are therefore more representative for the whole flooded area than those found in the area of Chong Kneas. Some habitats are present in both areas; others are specific to one place. In principle, of habitats present in both areas, those situated in the protected area in Kampong Kleang were sampled. Only when this was not possible, for a variety of reasons ranging from security matters to adverse weather conditions, similar sites in Chong Kneas were used.

Sampling sites were chosen based on habitat characteristics, to coincide with the habitat types described. In patchy patterns of vegetation, the larger areas were chosen, wherever possible, to be sampled.

Another factor that was taken into consideration was the accessibility of the sampling sites. The areas are large and little disclosed by roads, and the vegetation and at times the condition of the lake limit access both over land and water. Especially during the transitional seasons (beginning of the flooding, and particularly at the end of the flooding when the soil is saturated with water), large areas become very difficult or impossible to access. Wind action often results in considerable wave formation on the open waters of the lake, making it impossible to use gillnets as the waves cause the panels to roll and collapse. Despite taking this into consideration for the selection of the sites, at times it was not possible to reach the selected sites.

As there are almost no permanent reference points in the flood area, the location of the sampling sites was marked using a Global Positioning System (GPS) device. Attempts to mark the sampling sites with buoys proved ineffective. The same sites were used for sampling throughout the study.

Table 4. Location of the sampling sites. The lake sites in Chong Kneas were only used when the sites in Kampong Kleang were inaccessible.


Kampong Kleang area

Chong Kneas area





Floodplain pool


Lotus field


Rice field




Tonle Sap lake pelagic



Tonle Sap lake coastal



Figure 7. Schematic transect of the floodplain perpendicular to the shoreline. The scrubland holds patches of grassland, rice and lotus fields and floodplain pools. (Not to scale)

iv. Sampling procedure

A typical sampling trip took a full day per sampling site, i.e. per habitat. The day before the sampling, the team, comprising the data collector, one or two assistants and a boat operator, would prepare all the equipment and travel to the village nearest to the sampling site and spend the night there. On the sampling day, the team would go by boat to the sampling site, in time to set the gillnets at dawn. Then, the limnological samples and data were collected, and the gillnets were retrieved after the appropriate soaking time had lapsed. The catch would then be collected and either analysed on board or transported to a nearby facility for analysis and data collection. Later that day, preparations would be made for the sampling of the next day in another habitat.

Previous Page Top of Page Next Page