THE STATUS OF FRESHWATER BIODIVERSITY

Awareness has been growing during the past decade of the unique nature of much freshwater biodiversity, of the array of factors that in the past and at present have an impact upon freshwater species, and the extent to which real damage has already been done (recent reviews: Abramovitz, 1996, McAllister et al. 1997). The evidence is uneven in geographic scope, but decline in habitat quality and species populations is typical in the countries and regions where good field information is available, and real concern for the status of freshwater species worldwide is justified. While much general interest has been stimulated in the more spectacular of the terrestrial habitats and species, knowledge of the diversity and importance of the species hidden beneath the surface of freshwaters has remained very largely within the academic scientific community.

Many human activities tend to promote fragmentation of natural and often species-rich habitats (eg. floodplain rivers) and the spread of highly-managed species-poor habitats (eg. channelised rivers and reservoirs). Small isolated populations tend to be more sensitive than larger connected ones to demographic factors (eg. random events affecting the survival and reproduction of individuals) or environmental factors (eg. spread of disease, changes in food supply). The risks of this kind of distribution pattern may be compounded by other external threats, such as excess exploitation, introduced predators or pollution events.

Notes: derived from qualitative trend information on freshwater and wetland species (19 mammals, 92 birds, 72 reptiles, 44 fishes).

Notes: derived from quantitative trend information on freshwater and wetland species (3 mammals, 49 birds, 8 reptile, 10 fishes).

The distribution and systematics of fishes are inadequately known, although they are certainly the best-known species-rich and cosmopolitan group in freshwaters; their conservation status may exemplify the situation in other groups of organisms. Recent experience is that wherever fish faunas are studied, more species than suspected turn out to be threatened, or cannot be re-recorded at all (example reviews: Moyle and Leidy, 1992; Stiassny, 1996; Reinthal and Stiassny, 1991, Kirchhofer and Hefti, eds., 1996, and see IUCN, 1996).

Notes: These are the 20 countries whose fish faunas have been evaluated completely, or nearly so, and which have the greatest number of globally-threatened freshwater fish species.

Source: The threatened species data in this table were collated for The 1996 IUCN Red List of Threatened Animals. The estimates of total fish species present are all approximations.

Table 9 shows a selection of the very few countries where the status of the native fish fauna has been fully evaluated using the new IUCN threat category system. The countries are those with the twenty highest counts of globally-threatened species (ie. the entire species is at risk of extinction). In several countries 20-30% of the fish species present are threatened at this level; the mean for all 20 countries listed is around 17% (see Figure 3). This is certainly an underestimate because it covers only the species that meet the criteria for Critically Endangered, Endangered or Vulnerable listing under the new category system; it does not include those that are declining in parts of the range but are not yet threatened as a species, nor those that lack the information needed to make an assessment.

Among other freshwater groups, four of the five river dolphins and two of the three manatees are threatened, as are several smaller aquatic mammals, also around 40 freshwater turtles, more than 400 inland water crustaceans, and hundreds of bivalve and gastropod molluscs.

Note: these are the 20 countries with the highest numbers of globally threatened fish species, selected from the few countries where the fish fauna has been comprehensively assessed.

Source: based on data compiled by WCMC, in part for The 1996 IUCN Red List of Threatened Animals.

A species is extinct when the last individuals have died without leaving offspring (or, in a different sense, may be termed extinct when over evolutionary time a given lineage has branched into two or more lineages). The fossil record suggests that extinct species greatly outnumber living ones, perhaps by one thousand to one; this, and evolutionary theory, suggests that extinction is probably the ultimate fate of all species.

Most extinctions indicated in the fossil record have taken place during about five geologically very short periods; a corollary of this is that extinction rates have been relatively low over geological time in general. The average lifespan of a species in the fossil record is 5-10 million years, and if 12-13 million species now exist, the general background extinction rate may be between one and three species per year. For mammals, the average lifespan of species in the fossil record is one million years, which suggests one natural extinction every 200 years among the contemporary fauna.

Recent extinctions are likely to be recorded with significant accuracy either where circumstances favour preservation of hard remains in good number or where naturalists of the past century recorded the fauna or flora with sufficient care that they set a firm baseline against which the composition of the modern biota may be assessed. It is exceptional to observe the actual process of extinction. Typically, many years elapse before sightings of a species become sparse enough to generate concern, and many more years are likely to pass before negative evidence (ie. failure to find the species) accumulates to the point where extinction is the most probable explanation.

Notes; Data refer to globally extinct species, not national or geographic populations. 91 fish species were listed as extinct in the wild in 1996; this table includes 50 Lake Victoria cichlids all treated here as becoming extinct during the 1980s, and 31 other species for which estimated extinction times are available. A further 10 species could not be assigned to a decade.

Source: based on data compiled by WCMC, in part for The 1996 IUCN Red List of Threatened Animals.

Evidence for extinction of aquatic species is even less likely to be available than in terrestrial environments. However, some 81 fish species are recorded to have become extinct during the past century, and a further 11 are extinct in the wild but remain as captive populations (see Table 10 and Figure 4). This is far higher than estimated background animal extinctions. A major proportion of known extinctions have resulted from the ecological effects of the apparently deliberate introduction of the Nile Perch Lates niloticus into Lake Victoria in the mid-20^th century. The state of knowledge of freshwater fish faunas is so incomplete that other species could well have been lost before being discovered by scientists and formally described.

Notes: Data refer to globally extinct species, not national or geographic populations. 91 fish species were listed as extinct in the wild in 1996; this table includes 50 Lake Victoria cichlids all treated here as becoming extinct during the 1980s, and 31 other species for which estimated dates are available. A further 10 species could not be assigned to a decade.

Source: based on data compiled by WCMC, in part for The 1996 IUCN Red List of Threatened Animals.

Changes to the structure and quality of the freshwater environment are brought about by many different human activities. Many such changes are sufficiently large-scale and radical as to be obvious to a human observer (eg. dam construction); others occur on a much smaller scale and without visible effect (eg. release of sublethal pollutants). It is now clear that multiple habitat changes can have a cumulative impact on freshwater species, and there is evidence of a widespread and often severe decline in freshwater biodiversity.

External factors affecting populations of freshwater species include:

Although humans have always made use of freshwater systems and species, the last 200 years (the Industrial Revolution, the growth of cities, the spread of high-input agriculture) have brought about transformations on an unprecedented scale. The rate of water withdrawal rose steeply at the start of the present century, and further after mid-century. Over the same period the volume of river water polluted to some degree by waste water has similarly risen.

Major changes in the distribution of water on the continents has resulted mainly from withdrawals for irrigation, and secondarily from domestic and industrial use (L'Vovich, et al., 1990). Other factors are impoundment, wetland drainage and flood control. These physical changes have consequences for aquatic species: many large reservoirs have been created, river systems have been heavily disturbed, wetlands have been drained and the load of inorganic and organic pollutants in flowing waters has increased. From a water quality viewpoint, the major challenge is to address the increasing volume of polluted waste water from industrial and agricultural processes.

The various anthropogenic factors that impact upon freshwater systems can usefully be classified according to spatial scale and the location of effects (Table 11). As a general rule, wherever impacts have been investigated and changes in biological diversity demonstrated, multiple factors are involved.

Acid deposition through precipitation has been recognised as a regional transboundary phenomenon since the 1960s. Industrial emissions of sulphur and nitrogen oxides (SO₂, NO_x), mainly a result of fossil fuel combustion, are the principal source of acid rain. Most evidence of acid rain and its effects relates to North America and Europe, but emission rates are rising steeply in rapidly industrialising countries elsewhere. Acid rain in one country may be a consequence of compounds released into the atmosphere by industry in another country hundreds of kilometres distant. The geology, soil and vegetation of drainage basins will strongly influence the acidification process: coniferous forests (with acidic leaf litter) over granitic rocks will tend to promote acidification, whereas calcareous soils over limestone will exert a strong buffering effect on percolating water. Acid rain has been shown to decrease species diversity in lakes and streams. It has not been implicated in any recorded species extinction nor any major species decline. It has not yet been shown to be a significant issue in tropical freshwaters, where global freshwater diversity is concentrated.

Removal or extension of forest cover, or any anthropogenic interference with soils and land cover (eg. agriculture, urbanisation, road construction, mining), will modify the rate of runoff from catchment slopes and also the density of particles carried in the drainage system. All moving waters will carry some mass of suspended material, and there is considerable natural variation in this in space and time, but logging can increase sediment load by up to 100% for a short period, and 20-50% over the longer term. Sediment reaching lakes will be deposited and in effect enter long-term storage; depending on water velocity, sediment in rivers will settle out on floodplains or other parts of the course, or be carried into the coastal marine environment.

Increased sedimentation can have several effects on aquatic biodiversity: deposition can radically change the physical environment of species restricted to particular conditions of depth, light penetration and velocity; it is a major carrier of heavy metals, organic pollutants, pathogens and nutrient; it can interfere mechanically with respiration in gill-breathing organisms; and it can damage coral reef systems in the coastal environment.

Floodplain areas of large rivers tend to be regarded as wasteland suitable for draining and agricultural development; this destroys highly productive floodplain fisheries and modifies flow in the main course.

Dam construction and channelisation also strongly disrupt natural production cycles, including migration of fishes that ascend rivers from downstream areas or the sea in order to spawn. A review of hydrological change in the northern hemisphere as a result of dams and flow regulation is provided by Dynesius and Nilsson (1994). Fish production can be maintained or increased in some circumstances, in reservoirs or floodplain canals, although natural aquatic biodiversity is expected to decrease. For example, dam construction has so severely disrupted flow in the Colorado River (USA) that all native fishes in the lower reaches are in decline or extirpated (Moyle and Leidy, 1992).

Dam construction is the prime cause of extinction in the gastropod fauna of the Mobile Bay drainage in USA. Historically, the freshwater snail fauna of Mobile Bay basin was probably the most diverse in the world, followed by the Mekong River. Nine families and about 118 species were known at the turn of the century to occur in the Mobile Bay drainage. Several genera and many species were endemic, particularly in the Pleuroceridae. Recent surveys suggest at least 38 species are extinct (32%); decline in species richness ranges between 33% and 84% in the main river systems. The richest fauna was in the Coosa River and this system has undergone the greatest decline (from 82 to 30 species). Almost all the snail species presumed extinct were members of the Pleuroceridae and grazed on plants growing on rocks in shallow oxygen-rich riffle and shoal zones. The system has 33 major hydroelectric dams and many smaller impoundments, as well as locks and flood control structures. A combination of siltation behind dams, and submergence of shallow water shoals has removed the snails' former habitat. Where habitat remains it has diminished in area and become fragmented.

Globally, pollution and habitat modification are the most widespread and pervasive factors known to cause decline in fisheries. Water quality maintenance has generally been given much lower priority than industrial growth, and many river systems in developed countries are degraded as a result. Some countries have devoted resources to habitat restoration, with recent evidence of success. For example, the Rhine was a wild salmon-rich river two centuries ago, but by the 1970s, heavy pollution (combined with dam construction, channelisation, floodplain modification, and introduction of non-native fishes) led to marked decrease in populations of many fish species and collapse of fisheries. Since the end of the 1970s, water quality has improved and the decline in populations has slowed or reversed (Lelek, 1989, 1996).

Many smaller lakes have been affected, particularly by domestic and industrial wastes, and fisheries have declined or disappeared. Low levels of nutrient enrichment may stimulate production. There is growing concern for larger lakes, including the Rift Valley system in eastern Africa, where increased urbanisation and agricultural development are affecting catchment areas, with eg. increased sediment loads entering lake waters locally. The impact of these developments on biodiversity and fisheries is not known in detail.

There are major regional differences in the present and expected future impact on biodiversity and fish production of habitat degradation. Some of the most developed countries appear to have passed the peak of freshwater habitat modification, and are investing in water quality controls and rehabilitation measures. Some of the countries that are now undergoing rapid industrial development are a considerable distance from this state, and their freshwater habitats are likely to come under increasing pressure in coming years. The heaviest impacts are likely to be felt in eastern Europe, South and South-east Asia, with increasing industrial effluent and hydroelectric development, and in Africa, where water extraction and agricultural development for increasing human populations may be the principal impacts.

Unplanned or poorly planned introduction of non-native species and genetic stocks is a major threat to freshwater biodiversity (eg. Moyle, 1996). Such introductions can have negative or positive effects on fishery production. Table 12 and Figure 5 show the scale of introductions in recent decades; it is a reasonable assumption that all successful introductions will have an impact on existing population levels and community structure, and many changes are likely to be undesirable.

Lake Victoria, the largest tropical lake in the world, provides a classic example of the potential negative impacts of species introductions. Until some 30 years ago, when the large top predator, the Nile Perch Lates niloticus, was introduced, the lake supported an exceptional 'species flock' of around 300 species of haplochromine cichlid fishes as well as smaller numbers from other families. Not all the species have yet been formally described; many of these are known among aquarists and others only by informal common names. At least half and up to two-thirds of the native species are believed to be extinct or so severely depleted that too few individuals exist for the species to be harvested or recorded by scientists. The evolutionary processes behind this adaptive radiation, involving an immense variety in teeth and jaw morphology and feeding niches, have been the subject of considerable scientific research which has contributed to development of modern theories of evolutionary diversification. Additional factors in decline of the Victoria cichlids are excess fishing pressure, already evident before introduction of Nile Perch, and possible competition from tilapiine cichlids that were also introduced. The lake itself has now become depleted of oxygen, and a shrimp tolerant of oxygen-poor waters provides a major food source for the Nile Perch. In recent years the Nile Perch, and one of the introduced tilapiines form the basis of a high-yielding fishery, and an important national and export trade. It is unlikely that such high yields will be maintained.

Although it is of interest to distinguish the general factors that adversely affect freshwaters and their biodiversity, it is more useful from a management viewpoint to distinguish the various specific types of impact and the specific human activities that generates those impacts. Richter et al. (1997) termed these 'stressors' and 'sources' respectively, and analysed expert opinion on their identity in relation to more than 100 threatened aquatic species in the USA. Some 40 individual stressors were identified, grouped into six principal classes.

Source: compiled from data in Welcomme (1988).

Source: compiled from data in Welcomme (1988).

The threat classes identified were reported to arise from a relatively small number of primary sources, chiefly from different kinds of land use. Analysis of reported stressors and their sources for all 135 threatened species assessed indicates that three fundamental threat sources are most important (1-3 in Table 13). However, there was significant variation in results with respect to the groups of organism assessed, their geographic origin within the USA, and to historic versus current conditions. An important conclusion from this analysis is that it is not possible to derive a single global ranking of threats and their effects to guide conservation action; it will always be essential to evaluate local history, local conditions and individual species ecology in order to focus management efforts.

Note: information relates to analysis of species in USA.

Source: summarised from text in Richter et al., (1997).

Although there is much evidence for widespread decline in the health of many freshwater habitats, this is very variable in scope and quality. We have attempted to derive a semi-quantitative global assessment of change over recent decades in the condition of freshwater lakes. The study is based on Project Aqua, a project initiated by the Societas Internationalis Limnologiae in 1959, with the aim of documenting information on more than 600 water bodies judged worthy of conservation. A provisional list was issued in 1969 by the International Biological Programme, and a revised enlarged version was published with the additional support of IUCN in 1971 (Luther and Rzóska, 1971). So far as possible, data on each system were collated by national or regional specialists and presented in a standard numbered format; the information relates essentially to the 1960s.

A substantial number of systems treated in Project Aqua are also treated in later information sources, and in these cases it is often possible, taking the 1960s data as a baseline, to make an assessment of condition at a later time point and determine the direction of change. Of later sources, we have relied mainly on volumes such as The Directory of Asian Wetlands (Scott, 1989) and companion works dealing with other continents. The directories in many cases contain information relating to the 1980s and 1990s. We have also extracted information, often from the 1990s, from the Lakes Database (see references for web address) maintained by the International Lake Environment Committee Foundation. We have compared available entries (at least two, sometimes three) for 93 systems and scored each according to whether its condition appears to have deteriorated (or impacts have increased), improved, or no change is reported (this can mean 'no new information').

Note: sample of 93 lakes (and a small number of other wetland types); "worse" = condition deteriorated or impacts increased; "?" = no change reported; "better" = condition improved or remedial measures reported.

Source: data sources cited in text; prepared by WCMC for WWF Living Planet Report (Loh et al., 1998).

Note: sample of 93 lakes (and a small number of other wetland types) ); "worse" = condition deteriorated or impacts increased; "?" = no change reported; "better" = condition improoved or remedial measures reported.

Source: data sources cited in text; prepared by WCMC for WWF Living Planet Report (Loh et al., 1998).

The results are based on uneven sampling and non-standard reporting, but may be taken as valid indication of the general direction of change in recent decades (see Table 14). Most lakes in the sample have declined in quality, particularly those in Asia. However, not all remedial measures taken within the past ten years or so, or their beneficial effects, will have been reported in the information sources used. The data are graphed as percentages in Figure 6.