This paper draws on approaches in ecology, biology and policy analysis to examine the tensions between dams and fisheries in the Lower Mekong Basin. We review the exceptional importance of Mekong fisheries in terms of total catch, economic value and their role in rural livelihoods. The ecological conditions necessary to sustain the fish production are also analysed. The paper then considers the implications of dam development in the Mekong Basin, drawing on recent research to review predicted changes in hydrology and the resulting impacts on fishery resources. We then consider why, given the importance of fisheries, these potential impacts are not being addressed in regional policy processes.
The Mekong Basin's inland capture fisheries are of exceptional importance. Their tremendous size is matched only by their economic value, key role in rural livelihoods, and central place in regional food security. Unfortunately, this importance is not always communicated to policy makers or reflected in regional policy decisions.
According to commonly agreed upon figures, the Mekong Basin produces around 2.5 million tonnes of fish annually (Figure 1; 2.64 million tonnes based on catch estimates according to Van Zalinge et al. (2004); 2.06 million tonnes based on consumption studies according to Hortle (2007). This translates into more than four times the production per km2 of the North Sea (Jensen, 2001a), making it the most productive inland fishery in the world.
The economic value of this massive resource at the level of fishers has been estimated at $1.4 to $1.9 billion US (Van Zalinge et al., 2004; MRC, 2005). More than two thirds of the value of inland fish production comes from capture fisheries, compared to only 10% from reservoir fisheries and the remainder from aquaculture (Baran et al., 2007a). The Mekong's fisheries are also crucial to the livelihoods of the people living within the basin. The bulk of the catch is harvested by part-time and subsistence fishers who are poor and generally use fishing as part of a diversified livelihood strategy (Dixon et al., 2003). This diversification helps to ensure food security, generates money with little capital investment, provides cash in times of need (e.g. to buy rice seeds at the end of the dry season) and “insures” against the risk of crop failures (McKenney and Tola, 2002).
The nutritional needs of the population also depend on the basin's fisheries. The fish catch in Cambodia, for example, represents 3 times the pig production, 20 times the chicken production (Kurien et al., 2006), and up to 75% of animal protein requirements (Ahmed et al., 1998). When averaged, the annual fish consumption in the basin represents 29 kg per person (Hortle, 2007), or 80 g per person per day for each of the 60 million people of the basin (Table 1). Fish also provides necessary calcium as well as essential amino acids to the diet of the region's people (Jensen, 2001b; Guttman and Funge-Smith, 2000).
However, while fish is central to food security and to livelihoods in the region, the omnipresence of fish in life and meals gives it a low value as a food item paradoxically (Levy-Ward, 1993) and no specific emotional or symbolical status (Ivanoff, 2003). This translates into long-lasting, limited interest at the policy level in the conservation of this natural resource.
Ecological conditions necessary for sustainability
The Mekong Basin is largely a wetlands ecosystem whose productivity is governed by three main factors: hydrology, the quality of the floodplain environment, and fish migration patterns (Baran et al., 2001a, 2001b; Kurien et al., 2006; Figure 2). The Mekong River features the greatest hydrological variability in the world (Welcomme, 1985). This variability is largely related to the river's annual flood pulse, which occurs during the monsoon season (May to November) and inundates some 84,000 km2 of floodplains (Scott, 1989). These floodplains, equivalent to the surface area of Ireland, constitute huge feeding and spawning grounds for fish. Lastly, migrations are a major feature of Mekong fish species and have been detailed in a number of studies (MRC, 2001; Poulsen et al., 2002a).
Three hydrological factors play a major role in the sustainability of Mekong fisheries (Baran et al., 2001a). The first of these is the level of the flood: in years with higher flood levels, there is greater surface area available for fish to feed, leading to larger yields. The second is the duration of the flood: when floods are longer, fish have more time to grow, resulting in larger fish being caught. Finally, the timing of the flood is important: when the flood comes earlier, the fish larvae and juveniles have a higher survival rate.
The quality of the floodplain environment created by the Mekong's annual flood pulse is the second major factor in the sustainability of the basin's fisheries (Lamberts, 2006). Three main factors contribute to this quality: accessibility of floodplains, nature of the vegetation flooded and presence of refuges in the dry season. The accessibility of floodplains is put at risk by infrastructure development that fragments the habitat; in the Tonle Sap Basin for instance, 14,000 structures were identified in 2006 on the floodplain (Baran et al., 2007b). The nature of the vegetation flooded is seen as an important factor driving fish production, in particular the flooded forest (Van Zalinge et al., 2001), but studies detailing the relationship between flooded vegetation and fish productivity or biodiversity are still lacking. The availability of refuges in the dry season, such as floodplain ponds for the guild of “black fish” and deep pools in the Mekong mainstream for migrant “white fish” (Baran et al., 2001b; Poulsen et al., 2002b), is the last condition for the replenishment of stocks from year to year.
Eighty-seven percent of Mekong Basin species whose migrations status is known are migratory species; these 165 species belong to several families, with a strong dominance of catfishes (Pangasiidae, Siluridae, Bagridae, etc.) and of Cyprinidae (Henicorhynchus spp., Paralaubuca typus, Cyclocheilichthys enoplos, etc.) (Baran, 2006). These species undertake cyclical seasonal journeys between breeding grounds (often in upstream tributaries) and feeding grounds (on downstream floodplains). In fact, migratory fishes include long distance migrants (“white fish” as detailed above), but also short distance migrants (“grey fishes”) that move between local rivers and floodplains. The major factors triggering these migrations are hydrological factors, as 90% of Mekong fish species with known migration cues respond to variations in water level or discharge (Baran, 2006). Species with a large role in aquaculture and the highest commercial value such as Pangasiid catfishes are especially sensitive to these hydrological triggers. Thus, changes to these hydrological factors or to the connectivity of natural environments pose a clear threat to the sustainability of the Mekong's harvest.
Impacts of dams
In 2001, there were 11 completed hydropower projects larger than 10 MW in the Lower Mekong Basin (Nam Ngum, Xeset, Theun Hinboun, Houay Ho and Nam Leuk in Lao PDR; Sirindhorn, Chulabhorn, Ubolratana and Pak Mun in Thailand; and Nam Dray Ling and Yaly in Vietnam; MRC 2001). A recent review (King et al., 2007) has identified 30 dams in operation in the Mekong Basin, and this number is set to increase dramatically with another 92 large dams either planned or already under construction. Therefore, it is vital to consider the implications of dam building on fish production in the Mekong Basin.
A recent modelling approach (Koponen et al., 2007), based on data from the Mekong River Commission, considered three scenarios to predict the possible impacts of dam development in the Mekong Basin: The baseline scenario describes the situation existing in 2000, with one dam in the Upper Mekong Basin and five in the Lower Mekong Basin (but without integrating the small Thai dams created between 1965 and 2000). The intensive development scenario sees seven new dams built in the basin by 2025, with a water storage capacity of 55 km3 (China and Lao PDR would then account for about 83% of the total Mekong capacity). The extreme development scenario adds 7 dams to the previous scenario, assuming 20 dams in the Mekong Basin and 85 km3 of additional water storage capacity. The effects of these dams on the basin's fish resource, detailed in Baran et al. (2007b), are summarised below.
The most obvious of these is simply the loss of connectivity between natural environments: the physical presence of dams on the Mekong and its tributaries would block the migration routes used by fish and therefore prevent them from completing their natural lifecycle. In an environment where fish need to migrate each year between upstream and downstream habitats, dams located downstream near floodplain habitats would have the most significant impacts. Dams located upstream and on tributaries would obstruct a smaller proportion of the longitudinal fish migration network and would therefore have a smaller ecological impact on fisheries.
A large number of studies have shown that the existence of a seasonal flood is crucial in sustaining fish productivity, and this has been at the origin of the flood pulse concept (Junk et al., 1989). The flood increases the size of aquatic environment available to fish and brings in nutrients which stimulate rapid growth of the whole food chain, but it also allows decaying terrestrial matter to be promptly recycled and used (Lowe-McConnell, 1987; Arthington et al., 2004). Thus, in years of lower flooding the fish production is often much lower, as illustrated by the close correlation between maximum flood water level and the fish production of a local fishery during 12 years in Cambodia (Halls et al., 2007). Koponen et al. (2007) have shown that dams would reduce inflows into the lake between 4 and 10% in a wet year and between 10 and 25% in a dry year depending upon the scenario. The corresponding loss of suitable habitat would equate to a loss of productivity.
Dams also imply shorter periods of flooding and would correspond, for the scenarios mentioned above, to a shortening of 1 to 2 weeks of the flood season in the upper areas of the Tonle Sap floodplain. This would also translate into a smaller volume of fish production. These dams would then release water in the dry season, which is seen as beneficial to agricultural production, but would increase the minimal water level in the dry season resulting in the inundation of the flooded forest, which is most unlikely to survive permanent flooding. An increase in dry season flows would also impact the effectiveness of the artisanal gears designed to harvest fish migrating at low water levels (Baran et al., 2005).
Flood timing is another factor key to the sustainability of the fisheries. Most development scenarios predict that dams will store water in the monsoon season and release it in the dry season. Thus, the hydrological cues necessary to trigger migrations at the end of the dry season are likely to become blurred or disappear, which could prevent fish species sensitive to these natural signals from undertaking their seasonal breeding migrations (Baran, 2006). Additionally, variations in the flow regime in a system where the upstream movement of adults compensates for the downstream drift of larvae are likely to result in a very different distribution of fry in downstream areas (Welcomme and Halls, 2003).
A final impact of damming is related to the trapping of sediment, which is associated with a loss of productivity for fisheries as well as for agriculture (Kummu et al., 2005). Approximately half the sediment reaching the Mekong Delta derives from the Upper Mekong in China (Plinston and He Daming, 2000). Sarkkula et al. (2003) were able to show that flow reductions and sediment trapping by Chinese dams on the Mekong would have a dramatic impact on the overall sedimentation and productivity of the Tonle Sap Lake.
Declining fisheries related to hydropower development has sometimes been said to be offset by the creation of alternative fisheries in reservoirs. This is largely a myth. Out of 160 fish families living in freshwater, only 17 are fully able to live in lakes at one stage of their lifecycle, as most species have to return to free-flowing rivers to breed (Fernando and Holcik, 1982). Only nine species in the Mekong Basin are known to breed in reservoirs such as the ones that could be created behind dams (Baran, 2006). At the moment, reservoir fisheries contribute about 10% of the fish production basinwide, but it is unlikely that this reservoir production can ever replace the river capture fisheries. In terms of mitigation, effective measures to counteract dams' negative impacts have yet to be found. This is partly related to the scale of proposed developments. While “run-of-the-river” dams are said to be less detrimental to fisheries than retention dams, their average size amounts to 54 meters, making them serious obstacles to fish migrations (Hill and Hill, 1994). The effectiveness of other measures, such as fish ladders, has also been questioned in the context of the Mekong (Warren and Mattson, 2000; Jensen, 2001c), and the inefficiency of existing projects such as the Pak Mun dam fish ladder has already been demonstrated (Roberts, 2001).
Dams will give a major boost to the economies of the region, but they will also negatively impact other sectors relying on river flows such as fisheries. Making informed development decisions requires detailed information about proposed hydropower projects as well as a clear understanding of the full range of possible impacts. Unfortunately, in the rush to bring new hydropower generation online, many of the details of proposed projects have not been made public. Without this information, it is impossible to fully assess what effects continued damming will have on sectors such as fisheries or the viability of proposed mitigation measures.
What is known now is that a vast number of dams are either planned or already under construction. Updating the dam database detailed by King et al. (2007) using information in Cambodge Nouveau (2004, 2006), Watershed (2005), Lieng Vuthy (2006), Nippon Koei (2007), JICA (2007), Middleton and Sam Chanthy (2008) and complementary Internet searches has revealed previously unidentified plans for dams in Thailand and Lao PRD (mainstream dams in Luang Prabang, Sayabouri, Pa Mong, Don Sahong, Thakhek, and Ban Koum) and in Cambodia (25 dams in the Mekong Basin and 17 additional schemes in small coastal basins). An overview of the dam construction plans in the Mekong Basin is proposed in Table 2.
This information, combined with the development scenarios reviewed in the previous section, gives a reasonably clear picture of both what is at stake and who has the most to lose from damming in the basin. The Table 2 shows that Cambodia is now the country where the largest number of dams is being considered, even before Lao PDR (although the average installed capacity of 135 MW per dam in Cambodia is much smaller than in Lao PDR where it reaches 249 MW). Given that fisheries contribute 12% of the country's GDP, Cambodia is clearly the country with the most to lose from extreme hydropower development, followed by Vietnam, whose economy is more developed and diversified and thus less vulnerable to a loss of capture fisheries. Based on this information, Table 3 details, in a qualitative way, the risk vis-à-vis fishery resources incurred by each country in relation to the size of the local fishery and the number of dams planned.
This situation is complicated by issues related to national development strategies in the region. First, governments in the region have development priorities that are often at odds with each other (Campbell, 2005). For instance, China and Lao PDR see hydropower development as a cornerstone of national development plans. Yet, these plans will likely have a negative impact on downstream countries such as Cambodia, where fisheries are key to both the economy and rural livelihoods, and Vietnam, where sedimentation, saline intrusion into the delta, and water for irrigation are the main concerns.
Second, the development policies within countries often involve trade-offs. The case of Cambodia is illustrative in this regard. While Cambodia has recently started showing strong interest in regional hydropower development, inland capture fisheries remain a vital resource. Understanding the relative value of these two resources, and the trade-offs involved with regard to the economy, rural livelihoods, and food security in developing one at the expense of the other, depends on a clear picture of the situation. Unfortunately, details related to planned hydropower developments are not the only source of missing or conflicting information. The monitoring of fisheries at the regional level is often limited or inaccurate (Coates, 2002), and various interest groups, from the private sector to conservation groups, vie for attention without presenting a comprehensive assessment of the situation.
Before effective development strategies can be implemented, impartial assessments based on accurate quantitative data must be undertaken. The most comprehensive assessment of large dams to date has pointed out that the advantages of hydropower development are often exaggerated and its disadvantages underestimated (WCD, 2000). The Mekong fisheries are the most important inland fisheries in the world and provide food and income for millions of people. The loss of even a small percentage of this 2.64 million tonne resource represents the loss of tens of thousands of tonnes of fish and millions of dollars as well as uncalculated costs to the livelihoods and food security of the people living in the basin. Only with more rigorous quantitative assessments that take into account all possible impacts of dam construction will well-informed decisions be possible in the context of national and regional development plans.