Over long geological time periods, South America offered a large variety of waterbodies and habitats to aquatic organisms. Large river systems and streams of different water quality, different types of wetlands and high-altitude lakes favoured the development of a species-rich fish fauna with a large variety of adaptations. Therefore, South America is today a centre of megadiversity of fishes. In large areas, aquatic habitat health is still intact but environmental degradation is accelerating with negative impacts on the fish fauna. Inland fisheries are of great importance for protein supply of the local population, but stocks are already depleted in some areas. Fish culture is becoming increasingly important to supplement decreasing catches but cannot prevent loss of biodiversity. Lack of knowledge about fishes and their habitats, inefficient management of aquatic resources, large-scale uncoordinated modification of catchment areas, and the low priority given by the governments to fisheries and the protection of aquatic and wetland systems are reasons for decreasing fish stocks, deterioration of aquatic habitat health, and decreasing species numbers. A summary of the status of fish biodiversity, fishery, fish culture and habitat health in major South American rivers systems, the Altiplano and Patagonia is provided and the possibilities for sustainable management of aquatic resources and the protection of fish biodiversity are discussed.
The Neotropics are known as a centre of megadiversity of fishes. About 4,500 species have already been described (Reis et al., 2003), but the sampling of species is insufficient, even near large cities, as shown by the discovery of a new fish family that lives in a small stream in the vicinity of Manaus (Petry, 2003). According to different estimates, the total species number may rise to 6,000 or even 8,000 species (Schäfer, 1998; Reis et al., 2003). Little is known about the distribution of most species and their habitats. Information on the biology and ecology is often based on observations of ornamental fishes in aquariums by hobbyists. In recent years, increasing attempts have been made to elucidate the phylogeny of South American fishes, but many major questions are still unresolved (Vari and Malabarba, 1998).
The increasing efforts of scientists to improve our knowledge of fishes and their habitats coincide with an acceleration of the human occupation of even remote areas, which leads to dramatic large-scale changes in vegetation cover of the catchment areas and an increased use of waterbodies and aquatic resources, often with heavy negative side effects on fishes and their habitats. The governments of all South American countries face serious problems in coordinating and directing this development. As a result, all countries show deficiencies in environmental legislation, implementation of laws, administration of aquatic resources, and a lack of a policy for integrated watershed management as a basis for the sustainable management of waterbodies and their resources. Dramatic deterioration of habitat quality, habitat loss, and loss of fish diversity is the consequence. In remote areas, this goes often unnoticed by scientists and the public.
Inland fisheries are of fundamental importance for supplying local populations with inexpensive, high-quality animal protein. The dense river networks allow easy access to fishes, even in remote areas in the centre of the sub-continent, far from the sea. However, inland fisheries are often little regulated, with negative consequences for the stocks. Furthermore, fisheries compete heavily with other stakeholders who claim their rights on inland waterbodies, often to the injustice of the fishermen.
In this issue, an attempt is made to compile information on the South American freshwater fishes, fisheries, their biodiversity, and habitat health. To attain this goal, the fish fauna of major river basins and sub-basins, and of the Andean Altiplano and Patagonia, are described by different groups of authors. A river-basin approach rather than an individual-country approach was preferred because river basins are hydrological units that are characterized by specific fish communities and are not constrained by political boundaries. The same argument holds true, to a certain extent, for the large landscape units of Andean Altiplano and Patagonia. The papers in this issue cover about 80% of South America and provide a representative picture of the subjects in discussion, despite the variation in the level of information given in the chapters according to the considerable differences in catchment area, species richness and knowledge. In this summary, I present a sub-continental-wide view based mostly on the information given in the individual papers of this issue by Rodríguez et al. for the Orinoco River, Galvis & Mojica for the Magdalena River, Junk et al. for the Amazon River, Quirós et al. for the Paraná/Paraguay River, Agostinho et al. for the Upper Paraná River, Vila et al. for the waterbodies of the Altiplano, and Pascual et al. for Patagonian waterbodies.
The South American subcontinent covers an area of 17.85 × 106 km2. It reaches from about 13° N to 55° S and from 36° W to 82° W (Fig. 1). Most rivers drain to the Atlantic Ocean in the east because the sub-continent is delimited from the Pacific Ocean by the Andean mountains, which extend along its entire western rim. This mountain range developed during the separation of the sub-continent from Africa about 110 million years ago. Three major mountain systems lie inside the continent. The Guiana Shield in the north and the Central Brazilian Shield are separated by the Amazon basin. They are of Pre-Cambrian origin and are some of the oldest formations on Earth. The Central Brazilian Shield is, in part, covered by ancient sediment layers of different age and origin that form rather flat and deeply weathered plateaus. These plateaus connect to mountain ranges of different age that border in the east along the Tocantins valley and in the south at about 12° S, and overlie in part the Coastal Brazilian Shield (Harrington, 1962). It is assumed that the Magdalena, Orinoco, and Amazon rivers developed with the uplift of the Andes. More than 10 million years ago, the Magdalena River basin separated from the Orinoco River basin by the uplift of the eastern Cordillera. The older São Francisco, Tocantins and Paraná rivers were probably associated with the separation of South America and Africa in the Jurassic and Cretaceous periods (Potter, 1994).
Most of the subcontinent is well supplied with water. Dry regions with a mean precipitation of < 250 mm yr−1 exist only at the western slopes of the Andes, in some areas of the Argentine Chaco, and the centre of the dry north-eastern part of Brazil. There is a pronounced seasonality in rainfall over most parts of South America, leading to heavy drought stress during the dry season in some areas and large, rather predictable monomodal flood pulses in all large river systems and wetlands.
The lowlands near the equator are covered with tropical rain forest. In the west at the foothills of the Andes, the forest changes from 700 to 900 m upwards to different types of mountainous and cloud forests and later to high Andean Puna grasslands. North and south of the forest belt, different types of Cerrado and Savanna vegetation occur. The southern Cerrado belt in Brazil (about 2 × 106 km2) has been under strong human pressure for the past approximately 30 years. About 50% of the area has already been converted into pasture and cropland, mostly soybean. Agriculture is slowly advancing into the Amazon rain forest, as shown by the frequent fires observed during the dry season along a 2,000-km long “fire arch” that extends along its southern border. To the south, the Cerrado vegetation changes to the Argentine Chaco and further south to the southern Nothofagus forest. Along the Atlantic coast down to Argentina, a forest belt (Atlantic forest) has been mostly converted to crop land and pasture.
The subcontinent is characterized by the following large river systems (from north to south): Magdalena River; Orinoco River; Amazon River, including Tocantins River and Araguaia River; São Francisco River and, the Paraná/Paraguay rivers. The size of the catchment area, discharge and water-level fluctuations are given in Table 1. Large glacial lakes are a conspicuous feature of Patagonia and along the Andean mountain range. Lake Titicaca is the largest and deepest lake on the sub-continent, lying in an endorheic basin of the Andes at about 4000 m altitude and having a specific endemic fish fauna. Lakes in the lowlands are shallow, often embedded in floodplains, and are of young age. Most of them are subject to strong oscillations in size and depth and may dry out temporarily.
The total human population in South America is about 365 × 106 (20.5 people km−2). About 60% live in urban areas. In the countryside, the population concentrates along the large rivers that provide water, fishes, transport facilities, and often fertile alluvial soils for agriculture and animal farming. Even so, population density in large parts of the river catchments is low, except for the Magdalena River basin, parts of the Atlantic coastline, and the areas at the mouths of the Amazon and La Plata rivers (Table 1). The altiplanic population of Bolivia, Peru and Chile is estimated in 5.4 million and that of Patagonia in about 2 million.
Limnological conditions and habitats
Because of predictable pronounced dry and wet seasons in their catchments, most large South American rivers are characterized by a predictable monomodal flood pulse. Only the Magdalena River shows a bimodal flood curve. At high water level, extended areas along the middle and lower courses of the large rivers are inundated, forming large floodplains of different shape with many floodplain lakes. In addition to these riverine floodplains, there are large interfluvial wetlands that are flooded during the rainy season mostly by rainwater, and that are periodically or permanently connected to the large rivers. The different types of wetlands are indicated in Fig. 2.
A large delta is formed by the Orinoco River. The huge amounts of sediment from the Amazon River are distributed by the South Equatorial Current along the northern coastline up to the Guianas.
All large South American rivers have a low to moderate electrolyte content that varies in nutrient-poor and acidic black water and clear water tributaries of the Amazon, Orinoco and Paraguay rivers between 6 and 20 μ S cm−1, and reaches up to 300 μ S cm−1 in others. Brackish and saline waters are found in the endorheic Titicaca basin (Table 2). At high water, floodplain lakes often show pronounced hypoxia or anoxia near the bottom. Therefore, the benthos is often impoverished, and fish kills can occur when holomixis occurs during heavy rainstorms and temperature declines, despite many adaptations of the fish fauna to low oxygen concentrations.
Vegetation cover of the large floodplains corresponds to a certain degree with the vegetation in the surrounding uplands. Floodplains in the rain forest belt are covered by highly flood-adapted forests that tolerate mean annual inundation for up to 8 months. Lower-lying areas are covered by flood-tolerant perennial herbaceous plants. The lowest-lying areas are colonized during the low-water period by annual grasses, sedges, and herbs. In nutrient-rich whitewater rivers (e.g. Amazon, Purus, Madeira, Orinoko, Paraná/Paraguay rivers), abundant emergent and free-floating macrophyte communities develop during high water levels. Submerged vegetation is rare or absent because of unsuitable light conditions. In nutrient-poor blackwater rivers (e.g. Negro River), aquatic macrophytes are rare because of the low nutrient status of the water and/or the low pH value. Floodplains in the savanna belts have flood-tolerant forests along the river courses and drought-tolerant forests or shrublands in little-flooded areas, where drought and fire stress are heavier than flood stress. Large parts are covered during the terrestrial phase by grasses, sedges, and herbaceous plants and, at high water, by abundant submerged, emergent and free-floating aquatic macrophytes. Floodplain lakes are also, in part, colonized by submerged, emergent and free-floating aquatic macrophyte communities.
Currents and waterfalls are specific habitats that harbour rheophilic plant and animal species and many are found in the foothills of the Andes, the Guiana and Central Brazilian shields, and the eastern South American mountain ranges. Furthermore, they act as biogeographic barriers for the aquatic fauna unable to cross the obstacles upriver. Eighty percent of the fish species of the Iguaçú River, a tributary of the Paraná River, are endemic because of the presence of several cascades along the river course and the huge Iguaçú waterfalls in the lower reach, which act as natural barriers to dispersion (Garavello et al., 1997).
Today, the Checklist of the Freshwater Fishes of South and Central America (Reis et al., 2003) recognizes 4,475 fish species. Approximately 300 of these species are restricted to Central America, but many areas in South America have been insufficiently sampled. Reis et al. (2003) estimate another 1,550 not described species, bringing the estimated existing fish species in neotropical freshwaters to 6,025. This is much less than the extrapolation of Schäfer (1998) of more than 8,000 species. In any case, the Neotropics contribute 20 to 25% of the total fish fauna on Earth.
Most diversification of the modern neotropical fish fauna occurred during the roughly 70 million years from the late Cretaceous through Miocene periods. Late Miocene through Holocene history events on Earth had very little impact on the taxonomic diversity. The great antiquity and static history of almost all modern lower taxa can be shown by fossil traces. For instance, over the course of the last 13.5 million years or longer, fishes such as the Amazon tambaqui (Colossoma macropomum) persisted, apparently without changing its diet of fruits and seeds (Lundberg, 1998; Lundberg et al., 1998). There are only few examples of allopatric speciation, for instance, in Pleistocene Lago Valencia, which contains four endemic species (Mago-Leccia, 1970). Reproductive isolation by assortative mating is used to explain rapid intra-lacustrine speciation in the large African lakes, but seems to be of less importance in the hydrologically variable South American river systems. There are many cases of late and post-Miocene local extirpation of modern groups, for instance, in Argentina, Chile, the Magdalena Basin, the north coast of Venezuela, and the high Cuenca Basin of Ecuador (Lundberg et al., 1998).
Modern neotropical freshwater fishes have been assigned to approximately 60 families. Characiformes and Siluriformes are the dominant orders. In Amazonia, 43% of the species are Characiformes, 39% Siluriformes (Lowe-McConnell, 1987). In the upper Paraná River basin and in the Paraguay/La Plata River basin, each of these groups make up about 40% of the species (Agostinho et al., this issue; Quirós et al., this issue). These orders are followed by Cichlidae, Gymnotiformes and Cyprinodontiformes. The diversity of external morphology and internal anatomy is comparable to the high species number, as shown for Characiformes (Vari, 1998) and Siluriformes (de Pinna, 1993). The endemic Gymnotiformes are characterized by unique morphological, neurological, and electrosensory adaptations.
The morphological and anatomical diversity allowed the species to make use of the many diverse habitats and food resources. Colonization of large, periodically anoxic wetlands stimulated the development of a large variety of anatomical, morphological, physiological, and ethological adaptations to low oxygen concentrations, but also provided access to new habitats and terrestrial food items normally absent in lakes and rivers, such as fruit from floodplain forests and large numbers of terrestrial invertebrates (Goulding, 1980; Junk et al., 1997). The Orinoco River fish fauna seem to rely largely on algae or on herbivorous invertebrates. Food webs show remarkable food chain compression. Fish production derives mostly from 1st and 2nd trophic levels, sustaining high fish production on a relatively small amount of algal carbon (Rodríguez et al., this issue). Fish species that colonize ultra-oligotrophic lakes of Patagonia feed mostly on benthos and in the littoral zone because plankton is scarce (Pascual et al., this issue).
Changing water levels in rivers and floodplains required mobility and resulted in life history traits that include small- and medium-scale dislocations, but also small- to large-scale migrations in all large South American river systems, covering up to 4,000 km between the estuary of the Atlantic Ocean and the headwaters of the Amazon River at the Andean foothills (Junk et al., this issue). The reproductive strategies are comparatively diverse and include external and internal fecundation and different forms of complex parental care, as shown for the fish fauna of the upper Paraná River system (Agostinho et al., this issue).
The centre of species richness is the Amazon basin (Fig. 3), but this large basin shows considerable variation in species composition. A recent preliminary approach, led by The Nature Conservancy and World Wildlife Fund subdivide the basin into 13 eco-regions according to catchments of major tributaries or parts of catchments (Petry, 2003). With an increasing number of inventories, further subdivision of the eco-regions can be expected. The second largest diversity is found in the Orinoco River basin, which has a considerable number of species in common with the Amazon River basin. Differences in fish assemblages at the watershed scale are strongly associated with landscape characteristics and water-quality features. Transparency and electrical conductivity are particularly strong predictors of ichthyofaunal assemblages, as shown for the Orinoco River basin (Rodríguez et al., this issue). The geographic distribution pattern of the species is very complex. Some species, such as the catfish Pseudoplatystoma tigrinum and the characid Hoplias malabaricus, occur in most large lowland tropical and subtropical river basins. Others have a restricted distribution, such as the 190 described species of the catfish genus Corydoras, which occur mostly in small tributaries and lowland headwaters of the large tropical and sub-tropical rivers.
Generally, species numbers decrease considerably with increasing altitude and towards the cold southern end of the sub-continent. At the foothills of the Andes, fish fauna changes at an altitude of 500 to 700 m, which also represents the border of the tropical lowland rainforest. The endorheic freshwater systems of the Altiplano, which extend between 3,600 and 4,500 m altitude, harbour only three native fish genera: the killifish Orestias (43 species) and the catfishes Astroblepus (1 species) and Trichomycterus (13 species). Patagonia covers about 1.4 million square kilometers and harbours 17 described native species (Fig. 3).
Fish species can roughly be classified according to their occurrence into “true” lake species, headwater species, and species occurring in large rivers and their floodplains. The occurrence of “true” lake species is restricted mostly to Lake Titicaca and other Andean lakes. Many of the headwater species have a rather small distribution area. Species living in large rivers and their floodplains often have a wide distribution range. Welcomme (1985) classifies this group into “white fish”, which live preferentially in the river channel and migrate in large numbers during spawning; “black fish”, which prefer floodplain habitats and are highly adapted to low oxygen concentrations; and “grey fish”, which have an intermediate behaviour. This classification is also used by Quirós et al. (this issue). The classification of the fishes of the upper Paraná River basin according to reproductive strategies, given by Agostinho et al. (this issue), separates long-distance migrants from non-migrants and short-distance migrants. Long-distance migrants would fit into the “white fish” category of Welcomme (1985), and short-distance migrants into the “grey fish” and “black fish” categories.
The differentiation of fishes according to their occurrence provides hints to their vulnerability. True lake species and many headwater species are very vulnerable to human impact because of their restricted distribution areas, as shown for Lake Titicaca and for Patagonian waters (Vila et al., this issue; Pascual et al., this issue). Probably half of the species described for Amazonia are restricted to low-order headwaters that are actually seriously threatened by large-scale habitat modifications. Species occurring in high-order rivers and their floodplains have a higher resilience against human activities because of the larger range. However, studies along the Paraguay, Paraná and Rio de la Plata river axis show considerable differences in species composition between major river reaches. No fish migrations have been reported from the lower to the upper extreme of the riverine axis, and it is highly probable that fish stocks of the two extremes differ (Quirós et al., this issue). Reasons for these differences are not clear, but the river axis extends from the tropics far into the subtropics with considerable changes in temperature between summer and winter. To what extent species composition changes along the main stems of other large South American rivers remains an open question, but most of these rivers are situated in the tropics and subtropics where annual temperature changes are of minor importance. Species exchange along the river axis of these rivers seems to be more pronounced, as shown by the large number of long-distance migrants in the Amazon basin (Junk et al., this issue).
The role of inland fisheries and fish culture
Inland fisheries are of great importance for the protein supply of the local population throughout South America. Stocks in the Magdalena River of Colombia are heavily over-fished (Galvis and Mojica, this issue), and stocks of São Francisco River and Paraná River in Brazil are depleted (Sato and Godinho, 2004; Agostinho et al., this issue). The fishery potential of the Amazon River, Orinoco River, and Paraná/Paraguay rivers is not fully exploited (Junk et al., this issue; Rodríguez et al., this issue; Quirós et al., this issue). For instance, Bayley and Petrere (1989) estimate an inland fishery potential of about 9 × 105 t y−1 for the Amazon River basin, but actually only about half of it is used, and this use concentrates on only a few species. Some of these species already show signs of over-fishing, such as Pirarucu (Arapaima gigas), Tambaqui (Colossoma macropomum), and the large migrating catfishes (Pseudoplatystoma, Brachyplatystoma). Over-harvesting of a few large species is also reported from the Orinoco and the Paraguay river basins. It is assumed that selective fisheries have drastically modified the relative abundance, population structure and distribution of the fish stocks (Rodríguez et al., this issue). In Amazonia, the former abundant stocks of other aquatic species, such as manatee (Trichechus inunguis) and the large river turtles (Podocnemis spp.) were already strongly depleted at the beginning of the 20th century and are now in danger of extinction. The economic importance of inland fisheries of native species is low in Patagonia and the Altiplano, except in Lake Titicaca (Pascual et al., this issue; Vila et al., this issue).
According to all authors of this issue, fishery legislation is not adequate, and its implementation is insufficient in all large river basins, the Altiplano, and Patagonia, except for the newly enacted fishery law in Venezuela. Another problem that seriously hinders the development of efficient management strategies in all large South American rivers is the lack of long-term reliable fishery statistics. Without such statistics, the impact of natural variables and different kinds of human impacts on the stocks cannot be analysed adequately.
Recreational fisheries and ecotourism are becoming locally important in all catchments and may become an alternative source of income in traditionally managed wetlands, such as the Pantanal, the Llanos of Venezuela and high Andean lakes and rivers. However, there is a need for adequate regulation of sport fishing activities to overcome negative side effects, such as pollution and environmental destruction, to allow participation of the local population in the benefits of fishing tourism, and to avoid conflicts of interest with local fishermen.
In the Amazon basin, the capture of ornamental fishes is of regional importance in Colombia, Brazil and Peru. Every year, 16 to 35 million specimens of ornamental fishes belonging to about 50 species are exported from the middle Negro River in Brazil and 12.4 million from the Peruvian Amazon, mainly to the United States, Germany and Japan (Junk et al., this issue). Environmentalists criticize this fishery and point to a negative impact on the stocks, but available data do not indicate stock depletion. Scientists assume that a switch of the human population actually engaged in the export of ornamental fishes to slash-and-burn agriculture, cattle ranching and timber exploitation would certainly have a much stronger and irreversible negative impact on stocks and habitats than the actual fishery.
The social aspect of inland fisheries is also important in other regions. Fishermen are the weakest stakeholders in conflicts over aquatic resource management, and are heavily affected by environmental degradation, over-fishing, and land-use change. For instance, historically, there were larger fish stocks in the Magdalena River basin and fewer fishermen that used the floodplain also for subsistence farming. Today, there are 35,000 fishermen in the basin because many people without land try to survive as fishermen. They capture 17,600 t yr−1, corresponding to 1.4 kg d−1 per fisherman. They no longer have access to floodplain soils, which are now used by the politically stronger cattle ranchers. Conflicts of interest also arise in the Altiplano and even in some parts of Amazonia despite a much lower human population density and larger fish stocks. The multiple reasons for these conflicts require specific solutions that often surpass the problems directly related to fishery legislation. In the Magdalena River basin, the reform of land tenure and the reduction of the influence of drug lords are required. In Amazonia, different models of decentralized, community-based management of aquatic resources are being tested to solve the conflicts (Galvis and Mojica, this issue; Junk et al., this issue).
Fish culture in the Amazon, Orinoco and Paraguay river basins is only at its beginnings because of the still-abundant natural stocks. But in other areas, fish culture is advancing rapidly. For the past approximately 70 years, the African tilapias have been the backbone of the fish culture in water reservoirs of the dry north-eastern part of Brazil (Pereira et al., 2000). Stocking with exotic species, such as Oreochromis niloticus and Tilapia rendalli, and cage culture are also common practices in large reservoirs of the upper Paraná River basin, where the species now dominate the landings (Agostinho et al., this issue), and in the São Francisco River reservoirs (Sato and Godinho, 2004).
Since the beginning of the 20th century, rivers and lakes in Patagonia have been stocked with trout (rainbow trout, Oncorhynchus mykiss; brown trout, Salmo trutta; brook trout, Salvelinus fontinalis) and salmon (Atlantic salmon, Salmo salar; chinook salmon, Oncorhynchus tshawytscha; coho salmon, O. kisutch and cherry salmon, O. masou). This process started about 50 years later on the Altiplano, and included also stocking with the pampean silverside (Odontesthes bonariensis). Today, exotic species contribute considerably to fish production and recreational fisheries, but also replace native species by predation or food competition. In Lake Titicaca, the introduction of rainbow trout led to a decrease in the capture of native fishes, which were the principal fish sustenance of the Altiplanic people. Furthermore, it led to reshaping of the Patagonian fish fauna, the extinction of endemic Altiplanic species, such as Orestias cuvieri, and the decline of Orestias. pentlandii and Trichomycterus sp. On the other hand, about 6,000 people culture trout in net cages around the lake. Marine production of salmon in Chile grew dramatically from 53 t in 1981 to 300,000 t in 2000 (Pascual et al., this issue; Vila et al., this issue).
With about 42,000 t yr−1, fish farming in Colombia produces more fishes than all river fisheries together. Galvis and Mojica (this issue) state that the collapse of wild fisheries worldwide will be followed by a growth in fish farming. This is certainly true, but a productive fishery does not exclude a flourishing fish culture. In modern societies, a productive fishery is the expression of healthy aquatic ecosystems and the sustainable management of aquatic resources including biodiversity. Therefore, efforts should be undertaken to achieve both.
Habitat health, threats, and protection
The conservation assessment of Freshwater biodiversity of Latin America and the Caribbean (Olson et al., 1998) differentiates in South America 27 freshwater ecoregion complexes subdivided in 72 ecoregions. Ecoregions were defined as “relatively large areas of water and land that share a large majority of species, dynamics, and environmental conditions”. The status of six ecoregions was considered as critical (Magdalena Basin, Maracaibo Basin, Atacama/ Sechura Deserts, Lake Poopo, North Mediterranean Chile, Parano-Platense Basin); seven of them were considered as globally outstanding in biological distinctiveness and received highest priority for conservation being considered as vulnerable (Lake Titicaca, Llanos, Southern Orinoco, Amazon Main Channel, Rio Negro, Upper Amazon Piedmont and Pantanal) (Olson et al., 1998).
At first glance, this situation seems to be not so bad, but the accelerating environmental destruction in all South American countries is alarming and raises the question of long-term strategies for the sustainable use and protection of lakes, rivers and wetlands and their resources.
Protection measures can be classified in a hierarchical order. Certainly a prerequisite at the top level is the elaboration and implementation of integrated catchment area management plans because rivers represent the physical, chemical and biological conditions of the entire catchment area by the run-off of precipitation and groundwater. Human activities in the catchment will sooner or later affect the connected aquatic systems. Therefore, any management plan has to consider the impact on the environmental health of the river and wetland systems and to try to avoid or minimize negative side-effects. At the next level, the elaboration and implementation of integrated management plans for the different rivers and associated wetlands is required. This includes the inventory and classification of waterbodies and wetlands according to climatic, geographic, hydrological, chemical and biological parameters. At the third level, the environmental impact of the different management practices have to be analysed to optimise their economic and social output and minimize negative side-effects on the environment. There are serious deficits at all levels in all South American countries.
All authors of this issue agree that there are serious deficits in our knowledge of aquatic habitats and their fishes. Inventories and classifications of waterbodies and wetlands are incomplete. Only about two-thirds of the fish species are known, and even less is known about their distribution. Under these circumstances, an efficient protection of fish biodiversity is impossible in areas of accelerated agro-industrial development, for instance in the Brazilian Cerrado, the fire belt along the southern border of the Amazon rain forest, and the Argentine Chaco.
Furthermore, most countries do not have adequate legislation that deals specifically with wetlands, despite these habitats having key functions in water storage, water purification, aquifer recharge and maintenance of biodiversity. When there is such legislation, for example for the protection of the riparian vegetation, the implementation is usually insufficient. In Central Amazonia, environmental impact analyses concentrated on the rain forest and ignore aquatic resources. Fishery legislation often does not correspond to the local conditions, and there are serious deficits in its implementation. The abundance of water in most parts of South America results in negligence by the respective governments and a split of responsibilities that leads to rivalries between ministries and related organizations, and hinders concrete actions for the establishment of integrated watershed management plans. All authors of this issue agree that, without such plans, the protection of aquatic habitats and its resources is impossible in the long term.
Human impact is largest in basins with a high human population, mainly in the Magdalena River and the upper Paraná River basin, but also at Lake Titicaca. The expansion of agriculture led to large-scale modification of the natural vegetation. Water pollution by domestic and industrial waste and mining industries is severe because wastewater treatment plants are rare and do not meet modern standards. Efforts to improve wastewater treatment are offset in urban centres by the rapidly increasing human population numbers. Destruction of riparian vegetation and smaller wetlands is common. In Lake Titicaca, the coverage of the “tatora” reed (Schoenoplectus californicus ssp.tatora) has decreased from 70% in the 1970s to 15% in the 1990s principally because of eutrophication, and this decline has led to a reduction of reproduction habitats for fishes and birds (Northcote, 1991).
The Orinoco River, Amazon River, lower Paraná River, and Paraguay River basins have a very low human population and are largely still in rather pristine conditions. However, human impact is steadily rising and increasingly impacting habitat health and biodiversity. As already pointed out, agro-industries are destroying, on a large scale, the natural cerrado vegetation around the southern headwaters of the Amazon River, which harbour endemic aquatic species, many of them still not described. The destruction of riparian vegetation facilitates sediment input because of increased soil erosion and reduces habitat diversity and diversity of aquatic organisms (Wantzen, 1998). Most floodplain forests along the lower Amazon River and its large tributaries are destroyed by clear-cutting for cattle ranching or degraded by heavy logging. Large-scale deforestation is also mentioned as a major threat to aquatic habitat health and fish stocks in the Orinoco and lower Paraná River basin. The fish fauna of the Paraná River shows increased levels of persistent organic pollutants (Quirós et al., this issue). Local sources of pollution in remote areas are mining industries, e.g. gold mining in the Amazon River basin, the Pantanal and some tributaries of the Orinoco. Gold mining delivers large amounts of sediments and mercury to the river systems. Mining industries pollute the environment also at Lake Titicaca and the Paraguay River catchment.
All large rivers are important navigation routes and have increasingly been used since colonization by Europeans. Today, ocean-going vessels travel along the Amazon River about 3,000 km up to Iquitos in Peru and 1,000 km along the Paraná River up to Asuncion in Paraguay. There are plans for improving navigation of several large southern tributaries of the Amazon, such as the Madeira, Tapajos, Xingu and Araguaia rivers. Plans to straighten and deepen the upper Paraguay River inside the Pantanal threaten one of the most beautiful wetlands of the world and would heavily affect downriver areas.
A major threat is the quickly rising number of hydroelectric power plants. Hydroelectric power generation is already of large importance and will further increase. Brazil generates > 93%, Paraguay nearly 100%, Peru 74%, Venezuela 73%, Ecuador 68%, Colombia 68% and Chile 57% of their electric energy from hydropower (World Commission on Dams, 2000). Approximately 150 large reservoirs, including the giant Itaipu Reservoir (14,000 MW), lie along the upper and middle Paraná River and its tributaries (Agostinho et al., this issue). In the São Francisco River catchment area, there are 33 reservoirs, including six very large ones on the main stem (Sato and Godinho, 2004). Most of them are used for hydroelectric power generation, but also for water supply for agriculture and urban areas. The rising energy demand will dramatically increase the number of reservoirs in the next decades. The hydroelectric potential of the Brazilian Amazon basin is estimated to be 100,000 MW. For the production of this amount of energy, 90 reservoirs would be necessary, covering a total area of about 100,000 km2 and affecting all major tributaries of the Amazon River (Junk and Nunes de Mello, 1987). The Tocantins River and its upper tributaries would become transformed into a cascade of 20 reservoirs. Some reservoirs have a very low energy output per unit area and are considered very harmful to the environment, such as the Balbina reservoir at the Uatuma River in the Brazilian Amazon region, which produces 0.05 MW km−2. The reservoir covers an area of about 2,360 km2 of former undisturbed rain forest and will be a long-term source of methane because of large amounts of inundated organic material and its large aquatic-terrestrial transition zone (Fearnside, 1989, 1995).
Reservoirs have multiple effects on the fish fauna. They change lotic habitats to lentic habitats, inundate former floodplains, interrupt longitudinal connectivity, retain nutrients and sediments, accumulate pollutants, and modify the hydrologic regime downriver. Exceptions are high altitude reservoirs in the Magdalena River basin that are above the upper limits of fish migrations (Galvis and Mojica, this issue). The construction of fish ladders in the upper Paraná River basin has been shown to be very costly and of limited benefit to the highly diverse fish fauna. Furthermore, their importance for fish conservation is contested because reservoirs in cascades may not provide relevant nursery areas (Agostinho et al., 2002). The few available data on the upper Paraná River and São Francisco River basins indicate that the fishery yield of the reservoirs is low because fishes are not adapted to the new environment. Reservoirs do not compensate for the losses of yields in the undisturbed river floodplain systems (Agostinho et al., this issue; Sato and Godinho, 2004).
Water is a limiting factor mostly along the dry Pacific coast and in northeastern Brazil. The Brazilian government plans to allocate water of the São Francisco River to adjacent river basins to improve water availability for agricultural irrigation and urban centres. Environmentalists, large parts of the local population, and many scientists strongly oppose these plans, pointing to heavy negative environmental and socio-economic impacts. Water allocation would require large engineering efforts that would benefit mostly the construction contractors and large farmers. In an open letter of September 9, 2005 to the Minister of National Integration, Dr. Ciro Gomes, the Brazilian Society of Limnology raised serious doubts with respect to the project and stated that efficient wastewater treatment and improvement of storage capacities and infrastructure for a better distribution and efficient use of the available water would bring about more benefit for the local population and protect the environment (http://www.sblimno.org.br). A few weeks later, Bishop Frei Luiz Flávio Cappio of Barra, Bahia, started a hunger strike against the project. These democratic actions by parts of the society against the central government in favour of the protection of a river represent a new and encouraging development for environmental protection in South America.
An ambiguous issue is the introduction of exotic fish species. In general terms, the introduction of exotic terrestrial or aquatic species has to be treated with great caution because there are many examples of their uncontrollable spread with dramatic negative economic and ecological side effects. For instance, the predatory peacock bass (Cichla ocellaris), introduced from the Amazon basin in the Pantanal some years ago, is spreading and foraging severely on native species. The Asian golden mussel (Limnoperna fortunei) was introduced with ballast water by ships and first observed in 1993 in the mouth area of the La Plata River system. In 2001, it had already reached the Pantanal, about 2000 km upstream. The mussel creates serious problems by densely covering hard substrates and clogging water intake tubes (Darrigran and Ezcurra de Drago, 2000; C.T. Callil, Federal University of Mato Grosso, Brazil, pers. com.).
On the other hand, fish culturists in all river basins claim the need to introduce exotic species for economic reasons. A fast growing industry is already producing large amounts of protein in the Magdalena River Basin, Lake Titicaca, and in Patagonian waterbodies. However, little is done to prevent the escape of species from fish culture stations to adjacent waterbodies and the release of exotic species by private individuals. For instance, Orsi and Agostinho (1999) estimate that 1.3 million fishes, belonging to ten exotic species, reached natural waters in one sub-basin of the Paranapanema River during the catastrophic flooding of January 1997. To what extent these species will be able to build viable populations in nature is an open question, but the permanent danger of the spread of exotic fish species and associated exotic parasites and diseases exists. Probably in the 1940s and 1950s, the epizootic protozoan parasite Ichthyophthirius multifiliis was introduced with exotic fishes to Lake Titicaca and caused the death of 18 million Orestias in December 1981 (Wurtsbaugh and Tapia, 1988). Adequate legislation and its rigid implementation to control the introduction and cultivation of exotic species is, therefore, urgently needed throughout South America.
According to several authors of this issue, a major problem for the sustainable management of aquatic resources is centralized administration. It hinders active participation of the local population in the decision-making process, neglects cultural traditions of ethnic minorities, and hinders the input of local knowledge. This reduces the acceptance of regulations and hinders its implementation. Several experiments with participative management are ongoing, e.g. in Brazilian Reserves for Sustainable Management (Junk et al., this issue). Such programs are strengthening community-based political and social organization, stimulate environmental education, and may offer viable alternatives to the inefficient centralized management models in the future.
Environmental legislation and its implementation are precarious in all South American countries. Low political priority, inefficient administration, and a high level of corruption facilitate the destruction of rivers, wetlands, and their resources. For the past three decades, national and international environmental organizations have played an increasing role in environmental protection, mainly by establishing protected areas, getting involved in environmental education of the local population, influencing governmental development policies, and realizing research projects for environmental protection. The increasing globalization of commerce requires also the globalization of efforts for environmental protection. South America will strongly benefit from international cooperation in research and the formation of human resources in environmental sciences. This is not only true for North-South cooperation but also for a strengthened South-South collaboration, considering the leadership of Australia and the advanced status of South Africa in the implementation of modern concepts for the sustainable management of rivers, wetlands and aquatic resources, such as the Environmental Flow Assessment (Tharme, 2003).
Sustainable management also includes the establishment of protected areas. During the last decades, priority has been given in South America to the protection of terrestrial ecosystems, assuming that they would cover aquatic ecosystems occurring therein. This is certainly true but only satisfies the requirements of protection of the diversity of aquatic habitats and fishes to a limited extent. For instance, in Argentine Patagonia, continental protected areas cover about 37% of temperate rain forests and headwaters in the Andean region, but less than 5% of large rivers and shallow lakes in the steppe. In Amazonia, several reserves have been established for the protection of wetland areas, lakes and parts of rivers. For example, the Mamirauá Reserve for Sustainable Development at the confluence of the Japurá and Amazon rivers near Tefé (11,240 km2) and the nearby Jaú National Park (22,720 km2), which covers the entire catchment of the Jaú River, a tributary of the Negro River in Brazil, are large enough to maintain habitat diversity related to river dynamics and certainly protect the local river and wetland flora and fauna. However, large rivers cross different landscapes over distances of several thousand kilometres and have catchment areas of hundreds of thousands square kilometres with different plant and animal communities. Some fish species migrate for spawning over hundreds and even thousands of kilometres. New concepts are required for the establishment of protected areas that represent entire river systems and their aquatic biota.
With few exceptions, South America is a wet continent, dominated by large rivers that are accompanied by large fringing floodplains. Some rivers are also connected to large interfluvial wetlands. Probably about 15% of the entire continent is periodically or permanently flooded. Most of the wetlands occur in the moist tropics and subtropics. Permanent deep lakes are restricted mostly to the Andes and Patagonia.
Habitat diversity in river-floodplain systems is large and results in high fish species diversity. About 4,500 fish species in the Neotropics have already been described, 4,200 of which are in South America, and a total of 6,000 to 8,000 species are thought to occur, corresponding to about 20 to 25% of the world's fish fauna. Species diversity dramatically decreases with increasing altitude in the Andes and with increasing latitude towards the southern tip of South America, but even these regions considered generally inhospitable for fishes, have very peculiar endemic species.
Most large lakes, river-floodplain systems and interfluvial wetlands are still in a rather pristine stage, except for the Magdalena River, the upper Paraná River, the São Francisco River and areas of high human population density in other regions. Human impact on Lake Titicaca is rising.
Inland fisheries play an important role in the supply of the local population with high-quality animal protein. Fish stocks in the Magdalena River, São Francisco River, the upper Paraná River and Lake Titicaca are already depleted; those of the Amazon River, Orinoco River and Paraguay River are still not fully exploited, but stocks of highly valuable species show signs of over-fishing.
Fish culture, mostly with exotic species, plays already an important role in the Magdalena River basin, upper Paraná River basin, the semi-arid Brazilian North-East, Patagonia and the Altiplano. In the basins of the Amazon, Orinoco and upper Paraguay rivers, fish culture is still in its infancy, probably because of still abundant fish stocks. There is rising concern about the release of exotic species and their impact on the native fish fauna.
Major threats to ecosystem health and fishes are water pollution in basins with high human density (Magdalena River, upper Paraná River and São Francisco River, and reaches of other rivers below large cities), reservoir construction, which interrupts longitudinal connectivity and leads to river-floodplain degradation, and large-scale agro-industrial projects in the catchments that lead to accelerated aquatic habitat degradation because of increased soil erosion and pesticide input.
Knowledge on the ecology of natural waterbodies and wetlands, their structure, functions and multiple services, including biodiversity, is insufficient. Abundant water results in the lack of political will to manage water resources sustainably. Inadequate environmental legislation, inefficient administration and control, and corruption lead to accelerating degradation of lakes, rivers, associated wetlands and their aquatic resources. To overcome governance deficiencies, different types of decentralized, community-based management systems are being tested that hand over part of the administration of natural resources, including fisheries, to local communities.
In a time of accelerated globalization of commerce, the international community should also make use of their strength in stimulating efforts for protection and sustainable use of aquatic resources in South America. Large national and international research projects are necessary to increase knowledge of South American waterbodies and to train human resources. The role of national and international non-governmental organizations should be strengthened because they play a major role in raising awareness of the local population and political leadership.