Eutrophication of surface waters
Environmental chemistry
The point versus non-point source dilemma
Management of water quality impacts from fertilizers
Economics of control of fertilizer runoff
Aquaculture
Problems of restoration of eutrophic lakes
"Eutrophication" is the enrichment of surface waters with plant nutrients. While eutrophication occurs naturally, it is normally associated with anthropogenic sources of nutrients. The "trophic status" of lakes is the central concept in lake management. It describes the relationship between nutrient status of a lake and the growth of organic matter in the lake. Eutrophication is the process of change from one trophic state to a higher trophic state by the addition of nutrient. Agriculture is a major factor in eutrophication of surface waters.
The most complete global study of eutrophication was the Organization for Economic Cooperation and Development (OECD) Cooperative Programme on Eutrophication carried out in the 1970s in eighteen countries (Vollenweider et al., 1980). The sequence of trophic state, from oligotrophic (nutrient poor) to hypertrophic (= hypereutrophic [nutrient rich]) is shown in Table 12.
Although both nitrogen and phosphorus contribute to eutrophication, classification of trophic status usually focuses on that nutrient which is limiting. In the majority of cases, phosphorus is the limiting nutrient. While the effects of eutrophication such as algal blooms are readily visible, the process of eutrophication is complex and its measurement difficult. This is not the place for a major discussion on the science of eutrophication, however the factors noted in Table 13 indicate the types of variables that must be taken into account.
Because of the complex interaction amongst the many variables that play a part in eutrophication, Janus and Vollenweider (1981) concluded that it is impossible to develop strict boundaries between trophic classes. They calculated, for example, the probability (as %) of classifying a lake with total phosphorus and chlorophyll-a concentrations of 10 and 2.5 mg/m3 respectively, as:
|
Phosphorus |
Chlorophyll |
Ultra-oligotrophic |
10% |
6% |
Oligotrophic |
63% |
49% |
Mesotrophic |
26% |
42% |
Eutrophic |
1% |
3% |
Hypertrophic |
0% |
0% |
The symptoms and impacts of eutrophication are:
· Increase in production and biomass of phytoplankton, attached algae, and macrophytes.· Shift in habitat characteristics due to change in assemblage of aquatic plants.
· Replacement of desirable fish (e.g. salmonids in western countries) by less desirable species.
· Production of toxins by certain algae.
· Increasing operating expenses of public water supplies, including taste and odour problems, especially during periods of algal blooms.
· Deoxygenation of water, especially after collapse of algal blooms, usually resulting in fish kills.
· Infilling and clogging of irrigation canals with aquatic weeds (water hyacinth is a problem of introduction, not necessarily of eutrophication).
· Loss of recreational use of water due to slime, weed infestation, and noxious odour from decaying algae.
· Impediments to navigation due to dense weed growth.
· Economic loss due to change in fish species, fish kills, etc.
TABLE 12: Relationship between trophic levels and lake characteristics (Adapted from Janus and Vollenweider, 1981)
Trophic status |
Organic matter mg/m3 |
Mean total phosphorus1 mg/m3 |
Chlorophyll maximum1 |
Secchi depth1 |
Oligotrophic |
low |
8.0 |
4.2 |
9.9 |
¯ |
|
|
|
|
Mesotrophic |
medium |
26.7 |
16.1 |
4.2 |
¯ |
|
|
|
|
Eutrophic |
high |
84.4 |
42.6 |
2.45 |
¯ |
|
|
|
|
Hypertrophic |
very high |
750-1200 |
|
0.4-0.5 |
1 Values are the preliminary OECD classification and are defined as the geometric mean. Secchi depth is a measure of turbidity of the water column in a lake.
TABLE 13: Parameters for measuring and monitoring eutrophication (Source: Janus and Vollenweider, 1981)
Resultant variables |
Causal variables |
|
Short-term variability: high |
Short-term variability: moderate-low |
|
Phytoplankton biomass |
Zooplankton standing crop |
Nutrient loadings Nutrient concentrations |
Major algal groups and dominant species |
Bottom fauna standing crop |
|
Chlorophyll-a & other phytopigments |
Epilimnetic D P, D N, D Si (D is difference between winter and summer concentrations) |
|
|
|
|
Particulate organic carbon and N |
Hypolimnetic O2 and D O2 |
|
Daily primary production rates |
Annual primary production |
|
Secchi disc visibility |
|
In their summary of water quality impacts of fertilizers, FAO/ECE (1991) cited the following problems:
· Fertilization of surface waters (eutrophication) results in, for example, explosive growth of algae which causes disruptive changes to the biological equilibrium [including fish kills]. This is true both for inland waters (ditches, river, lakes) and coastal waters.· Groundwater is being polluted mainly by nitrates. In all countries groundwater is an important source of drinking water. In several areas the groundwater is polluted to an extent that it is no longer fit to be used as drinking water according to present standards.
While these problems were primarily attributed to mineral fertilizers by FAO/ECE (1991), in some areas the problem is particularly associated with extensive and intensive application of organic fertilizers (manure).
The precise role of agriculture in eutrophication of surface water and contamination of groundwater is difficult to quantify. Where it is warranted, the use of environmental isotopes can aid in the diagnosis of pollutant pathways to and within groundwater (IAEA, pers. comm. 1996). RIVM (1992), citing Isermann (1990), calculated that European agriculture is responsible for 60% of the total riverine flux of nitrogen to the North Sea, and 25% of the total phosphorus loading. Agriculture also makes a substantial contribution to the total atmospheric nitrogen loading to the North and the Baltic Seas. This amounts to 65% and 55% respectively. Czechoslovakia reported that agriculture contributes 48% of the pollution of surface water; Norway and Finland reported locally significant eutrophication of surface waters arising from agriculture; high levels of usage of N and P are considered to be responsible for proliferation of algae in the Adriatic; similar observations are made in Danish coastal waters; substantial contamination of groundwater by nitrate in the Netherlands was also reported (FAO/ECE, 1991). Appelgren (FAO, 1994b) reported that 50% of shallow groundwater wells that supply more than one million rural residents in Lithuania are unfit for human consumption because of a wide range of pollutants which include pesticides and nitrogen species. In the 1960s Lake Erie (one of the North American Great Lakes) was declared "dead" by the press due to the high levels of nutrients accompanied by excessive growth of algae, fish kills, and anaerobic bottom sediments.
Although the ECE (1992) regarded livestock wastes as a point source and excluded it from calculations of the contribution of agriculture to eutrophication in Europe, their statistics indicated that livestock wastes accounted "on average" for 30% of the total phosphorus load to European inland waters, with the rest of agriculture accounting for a further 17%. The situation for nitrogen, as for phosphorus, was quite variable from country to country. Danish statistics indicated that manure contributes at least 50% of the leaching of inorganic N (Joly, 1993). Nitrogen from agricultural non-point sources in the Netherlands amounted to 71% of the total N load generated from within the Netherlands (ECE, 1992).
A study by Ryding (1986) in Sweden demonstrated how lakes which were unaffected by industrial or municipal point sources, underwent long-term change in nutrient status as a result of agricultural activities in the watershed. Over the period 1973-1981 the nutrient status of Lake Oren increased from 780 to 1000 mg/m3 for Total-N, and from 10 to 45 mg/m3 for Total-P. Lake transparency declined from 6.2 to 2.6 m and suffered periodic (heavy) algal blooms.
As noted in Chapter 1, the US-EPA regards agriculture as the leading source of impairment of that nation's rivers and lakes with nutrients ranking second only to siltation as the pollutant most affecting rivers and lakes.
The values cited in Tables 14 and 15 indicate the wide range of nutrient losses that are measured at the plot, field and sub-basin scales. Heavily fertilized crops such as maize tend to have large losses relative to non-intensive uses such as pasture. Agricultural uses associated with poor land management practices that lead to erosion also produce significant nutrient losses. Wastes, manures and sludges, through biological concentration processes, can supply soils with 100 times more hazardous products than fertilizers for the equivalent plant nutrient content (Joly, 1993). This is considered a major environmental (and water quality) problem in periurban areas of many developing countries. Numerous authors report that a high degree of variability at individual sites is expected as a consequence of changes in hydrological regime from year to year. The implication is that estimation techniques using "typical" values of nutrient yield can expect to have a high degree of uncertainty and could be very much in error if estimated from data collected over a single year.
TABLE 14: Selected values for nutrient losses
Location |
Land use |
Phosphorus (kg/ha/yr) |
Nitrogen (kg/ha/yr) |
Comment (source) |
Southern Ontario |
Cropland |
0.415 |
|
Sediment-associated P, mean of 14 catchments (Spires and Miller, 1978) |
Unimproved |
0.08 |
|
|
|
Maize, potatoes |
|
26.0 |
From 11 catchments in S. Ontario (Nielsen et al. 1978) |
|
Cereals, beans. veg. & tobacco |
|
3.6 |
As above |
|
Hay, unimproved pasture |
|
0.1 |
As above |
|
Unimproved |
|
0.0 |
As above |
|
Great Lakes Basin (N. America) |
Cropland |
|
0.2-37.1 |
Range reported for total N for 15 studies of stream-discharged N (not tile drains) Nielsen et al. 1978) |
Hungary |
Cropland |
1.142 |
|
1984 results for 73 km2 watershed tributary in Lake Balaton (Jolankai, 1986) |
Denmark & Netherlands |
Livestock/crop systems |
|
316 |
Where 680 kg/N/ha are added (Joly, 1993) |
USA |
Cropland |
|
64 |
Where livestock is not intense (Joly, 1993) |
Côte d'Ivoire |
Agriculture |
29.0 |
98 |
Lower Côte d'Ivoire (FAO, 1994a) |
TABLE 15 Relative leaching losses of nitrogen and phosphorus (% change between no fertilizer and with fertilizer) (Source: Bolton et al., 1970 as reported in Bangay, 1976)
Management |
Crop |
Percent change |
|
P |
N |
||
Rotation |
Maize |
+10 |
+65 |
Oats, alfalfa |
- 05 |
+33 |
|
Alfalfa 1st year |
+17 |
- 08 |
|
Alfalfa 2nd yr |
+59 |
+09 |
|
Continuous |
Maize |
+12 |
+102 |
Bluegrass sod |
+12 |
- 69 |
|
Average |
|
+17 |
+27 |
Note: Values of +/- 15% are likely within detection limits of the methods used.
The huge increases in fertilizer use worldwide over the past several decades are well documented. Figure 9 illustrates the historical trends and predicted future needs of fertilizer use. However, fertilizer use (either mineral or organic) is not, of itself, the primary factor in downstream water quality. More important are the land management practices that are used in crop production.
FIGURE 9: Fertilizer use development and crop yield evolution in Asian, European and Latin American countries and the United States (Source: Joly, 1993)
There is a danger, however, in assuming that all waters have natural levels that are low in nutrients. In some areas, such as lakes located in areas of rich agricultural soils, waters have historically been highly enriched by nutrients associated with natural erosion of fertile soils. In the prairie lakes of Canada, for example, early settlers reported that the lakes were green with algae. In other parts of the world, as in Asia, ancient civilizations so profoundly impacted water quality that there are no longer "natural" levels of nutrients. In such situations the existence of eutrophication, while undeniable, must be measured against arbitrary standards that reflect water quality criteria established on the basis of societal needs for beneficial use of the water.
The importance and, in some cases, the major problems associated with organic fertilizers, deserve special mention. Manure produced by cattle, pigs and poultry are used as organic fertilizer the world over. To this is added human excreta, especially in some Asian countries where animal and human excreta are traditionally used in fish culture as well as on soils. However, intensive livestock production has produced major problems of environmental degradation, a phenomenon which has been the subject of European and North American legislation and control. The problem is particularly acute in areas of intensive livestock production, such as in the Eastern and Southern parts of the Netherlands where the production of manure greatly exceeds the capacity of the land to assimilate these wastes.
In addition to problems associated with excessive application of manure on the land, is the problem of direct runoff from intensive cattle, pig and poultry farms. Although this is controlled in many western countries, it constitutes a serious problem for water quality in much of the rest of the world. For example, Appelgren (FAO, 1994b) reports that discharge of pig wastes from intensive pig raising in Lithuania is a major source of surface water pollution in that country. The FAO/ECE reports similar problems in the Po River of Italy. The Canadian Department of Agriculture calculated in 1978, on the basis of detailed study of several feedlot operations, that cattle feedlots and manure storage facilities contributed 0.5-13% of the total loading of total phosphorus at that time to the Canadian portion of the Lower (agricultural portion) Great Lakes (Coote and Hore, 1978).
To the typical pathways of degradation, that of surface runoff and infiltration into the groundwater, is added the volatilization of ammonia which adds to acidification of land and water. In a review of environmental impacts caused by animal husbandry in Europe, the FAO/ECE (1991) reported the following major categories of impacts:
· Fertilization of surface waters, both as a result of direct discharges of manure and as a consequence of nitrate, phosphate and potassium being leached from the soil.· Contamination of the groundwater as a result of leaching, especially by nitrate. Phosphates are less readily leached out, but in areas where the soil is saturated with phosphate this substance is found in the groundwater more and more often.
· Surface waters and the groundwater are being contaminated by heavy metals. High concentrations of these substances pose a threat to the health of man and animals. To a certain extent these heavy metals accumulate in the soil, from which they are taken up by crops. For example, pig manure contains significant quantities of copper.
· Acidification as a result of ammonia emission (volatilization) from livestock accommodation, manure storage facilities, and manure being spread on the land. Ammonia constitutes a major contribution to the acidification of the environment, especially in areas with considerable intensive livestock farming.
FIGURE 10 The N cycle in soil (from Stevenson, 1965)
FIGURE 11 Schematic diagram of nitrogen and phosphorus losses. Arrows are proportional to loss
The key hydrological processes that link rainfall, runoff and leaching, and which give rise to erosion and transport of chemically enriched soil particles, are important components of the environmental chemistry, transport and fate of fertilizer products. These hydrological processes are described in Chapter 2 and are not repeated here.
The environmental dynamics of nitrogen and phosphorus are well known although the detailed transformations of nitrogen that occur in soil and water are difficult to study and document. The nitrogen cycle is depicted in Figure 10.
Nitrogen is comprised of the forms: soluble organic N, NH4-N (ammonium), NO3-N (nitrate), NO2-N (nitrite), and N associated with sediment as exchangeable NH4-N or organic-N. Nitrogen cycling is extremely dynamic and complex, especially the microbiological processes responsible for mineralization, fixation and denitrification of soil nitrogen. Generally, in soils that are not waterlogged, soil N (held as protein in plant matter) and fertilizer-N are microbiologically transformed to NH4 (ammonium) through the process of ammonification. The ammonium ion is oxidized by two groups of bacteria (Nitrosomonas and Nitrobacter) to NO3 with an unstable intermediate NO2 product in a process called nitrification. Urea is readily hydrolysed to ammonium. Denitrification occurs under anoxic conditions such as wetlands where NO3 is reduced to various gaseous forms. The N cycle is largely controlled by bacteria, hence the rate of N cycling is dependent upon factors such as soil moisture, temperature, pH, etc. NO3 is the end-product of aerobic N decomposition and is always dissolved and mobile.
From a water quality perspective, the ammonium ion (NH4) can be adsorbed to clay particles and moved with soil during erosion. More importantly, however, NH4 and NO3 are soluble and are mobilized through the soil profile to groundwater during periods of rain by the process of leaching. NO3 is also observed in surface runoff during rainfall events. Prevention of nitrogen pollution of surface and groundwater depends very much on the ability to maintain NO3 in soil only up to the level that can be taken up by the crop, and to reduce the amount of NO3 held in the soil after harvesting. The processes described above are depicted in Figure 11.
In contrast, the behaviour of phosphorus is quite simple. Phosphorus can exist in a variety of forms: as mineral (generally apatite) phosphorus (AP); non-apatite inorganic-P (NAIP); organic-P (OP - bound up with carbon and oxygen in plant matter); and as dissolved soluble reactive ortho-P (SRP). The phosphorus species AP, NAIP and OP are associated with the particulate phase. In studies of phosphorus movement from agricultural lands the largest amount is sorbed onto clay materials and transported as erosion products. SRP is readily available to aquatic plants to the point where measured SRP in surface water may only represent a residual amount after most of the SRP has been taken up by plant life. Consequently, in aquatic studies, the focus is often on the sediment-associated forms of P as these tend to dominate total phosphorus flux. The NAIP fraction is considered to be available to plant roots and is rapidly solubilized under conditions of anoxia in the bottom of lakes and reservoirs. It is for this reason that lake sediments can represent a very large internal (autotrophic) load of phosphorus which is recycled into the water column during periods of bottom anoxia. This load can be so large that, without attention to lake sediment remediation, phosphorus management programmes in tributaries can be quite meaningless.
The relative losses of N and P to groundwater are illustrated in Table 15 where it is seen that P losses are generally smaller relative to the much more soluble N. Indeed, insofar as maize is the most heavily fertilized crop, the leaching of nitrogen is especially noticeable.
The dilemma in many countries is to ascertain the role of agriculture relative to the impacts of (often untreated) municipal sewage. In a large number of countries the database required to make this distinction is lacking and frustrates the development of a rational pollution abatement programme and inhibits cost-effective investment in control measures. In developing countries it makes sense that the focus should initially be on point source control; however, it has been the experience in the developed countries that point source control for nutrients has not had the desired level of environmental benefit until agricultural control measures were seriously addressed. It is significant that the trend of fertilizer usage worldwide has been one of huge increases in the past 40 years suggesting that, in the absence of major changes in land use to control fertilizer runoff in large parts of the world, one may expect that agriculture will be responsible for an ever-increasing contribution to surface water pollution.
The observations reported by Quirós (see Box 5) for the La Plata basin are indicative of the difficulty in segregating the effects of agriculture from other sources. In the Great Lakes of North America some $10 million dollars were spent between 1970 and 1980 to quantify the relative impacts of point versus non-point sources. That exercise proved enormously successful and specific policies were adopted for nutrient control in each lake basin that reflected the relative contributions from each type of source.
The investigation of eutrophication of surface water by agriculture must adopt a pragmatic management perspective. Of value to agriculture is the perspective adopted by the OECD study of eutrophication. That study focused on the following aspects:
· The qualitative assessment of the trophic state of bodies of water in terms of a few easily measured parameters.· The dependence of this state on nutritional conditions and nutrient load.
· Translation of these results to the needs of eutrophication control for management.
The progression of these aspects is interesting in that the focus is on easily measurable state of the water body, followed by a determination of the extent to which the state is a product of nutrient loads, then the degree to which loads may be manipulated to achieve a desired trophic state that is determined by water use.
Prediction of water quality impacts of fertilizers and related land management practices is an essential element of site-specific control options and for the development of generic approaches for fertilizer control. Prediction tools are essentially in the form of models, many of which are contained in Table 7.
BOX 5: SEGREGATING AGRICULTURAL FROM INDUSTRIAL IMPACTS ON WATER QUALITY OF THE LA PLATA BASIN, SOUTH AMERICA In his report on impacts on the fishery in the la Plata river system, Quirós (1993) provided a comprehensive summary of observed symptoms. He admitted to the difficulty in providing evidence for cause and effect in this large river system. Nevertheless, he concluded that the evidence was consistent with that of a regulated river-floodplain system impacted from toxic substances used in agriculture and industry. Observed Symptoms Fruit and seed eater species of the genera Colossoma and Brycon and the big catfish Paulicea lutkenii have practically disappeared from the commercial catch in the lower Paraná river, and also from the catches in the La Plata and Uruguay rivers. Fish species of marine lineage of the genera Basilichthys and Lycengraulis, usually moving upstream from the estuary in winter, have practically disappeared from the commercial catches in the middle Paraná. The commercial catches of the pelagic top predator Salminus maxillosus have been decreasing since the late 1940s in all the lower basin, though its commercial catch has been highly restricted. Populations of most of the migratory fish species are severely diminished in the middle and upper Uruguay river. Relatively high levels of agricultural pesticides and heavy metals were detected in fish tissues. Periodic massive fish mortalities were reported in the lower Paraná delta and the La Plata river. Low water oxygen levels and massive fish mortalities were detected in the lower Paraguay river, and discharges of high organic matter content effluents from the agricultural industry have increased in the upper basin. The exotic Cyprinus carpio was the most important species in biomass in the experimental catches in the La Plata river, and its catch has been increasing in the middle Paraná. Maximum size of catch of the big catfish of the genera Pseudoplatystoma has been decreasing for the last three decades in the lower middle Paraná. The conflicting situations between recreational and commercial fishermen have been increasing, and the trophy size of Salminus has been decreasing at the confluence of the Paraná and Paraguay rivers, though total fishing effort seems not to have increased. Source: Quirós (1993). |
The response to the need to control leaching and runoff of nutrients and contamination of soils and water by heavy metals has been variable in Europe. Control measures are part of the larger issue of mineral and organic fertilizer usage. FAO/ECE (1991) summarized the types of voluntary and mandated controls in Europe that apply to mineral fertilizers as:
· Taxes on fertilizer.· Requirement for fertilizer plans.
· Preventing the leaching of nutrients after the growing season by increasing the area under autumn/winter green cover, and by sowing crops with elevated nitrogen
· Promoting and subsidizing better application methods, developing new, environmentally sound fertilizers, and promoting soil testing.
· Severely limiting the use of fertilizers in e.g. water extraction areas and nature protection areas.
In any location where intensive agriculture and/or livestock farming produces serious risks of nitrogen pollution, Ignazi (1993) recommended the following essential steps that are taken at the farm level:
1. Rational nitrogen application: To avoid over-fertilization, the rate of nitrogen fertilizer to be applied needs to be calculated on the basis of the "crop nitrogen balance". This takes into account plant needs and amount of N in the soil.2. Vegetation cover: As far as possible, keep the soil covered with vegetation. This inhibits build-up of soluble nitrogen by absorbing mineralized nitrogen and preventing leaching during periods of rain.
3. Manage the period between crops: Organic debris produced by harvesting is easily mineralized into leachable N. Steps to reduce leachable N includes planting of "green manure" crops, and delaying ploughing of straw, roots and leaves into the soil.
4. Rational irrigation: Poor irrigation has one of the worst impacts on water quality, whereas precision irrigation is one of the least polluting practices as well as reducing net cost of supplied water.
5. Optimize other cultivation techniques: Highest yields with minimum water quality impacts require optimization of practices such as weed, pest and disease control, liming, balanced mineral fertilizers including trace elements, etc.
6. Agricultural Planning: Implement erosion control techniques (see Chapter 2) that complement topographic and soil conditions.
Voluntary and legislated control measures in Europe are intended to have the following benefits:
· Reduce the leaching of nutrients
· Reducing emissions of ammonia
· Reducing contamination by heavy metals
The nature of these measures varies by country; however FAO/ECE (1991) have summarized the types of voluntary and mandated control as:
· Maximum numbers of animals per hectare based on amount of manure that can be safely applied per hectare of land.· Maximum quantities of manure that can be applied on the land is fixed, based on the N and P content of the manure.
· Holdings wishing to keep more than a given number of animals must obtain a license.
· The periods during which it is allowed to apply manure to the land have been limited, and it is obligatory to work it into the ground immediately afterwards.
· Establishment of regulations on minimum capacity for manure storage facilities.
· Establish fertilizer plans.
· Levies (taxes) on surplus manure.
· Areas under autumn/winter green cover were extended, and green fallowing is being promoted.
· Maximum amounts established for spreading of sewage sludge on land based on heavy metal content.
· Change in composition of feed to reduce amount of nutrients and heavy metals.
· Research and implementation of means of reducing ammonia loss.
Sludge is mentioned here only insofar as the spreading of sludge from municipal wastewater treatment facilities on agricultural land is one method used to get rid of municipal sludge in a way that is perceived to be beneficial. The alternatives are incineration and land fill. FAO/ECE (1991) include sludge within the category of organic fertilizers but note that sludge often contains unacceptable levels of heavy metals. Pollution of water by sludge runoff is otherwise the same as for manure noted above.
Nutrient loss is closely associated with rainfall-runoff events. For phosphorus, which tends to be associated with the solid phase (sediment), runoff losses are directly linked to erosion. Therefore the economics of nutrient control tend to be closely tied to the costs of controlling runoff and erosion. Therefore, this will be treated briefly here. In particular, it is useful to examine the economic cost of nutrient runoff which must be replaced by fertilizers if the land is to remain productive.
The link between erosion, increasing fertilizer application, and loss of soil productivity is very direct in many countries. In the Brazilian state of Paraná where agriculture is the base of the state economy, Paraná produces 22% of the national grain production on only 2.4% of the Brazilian territory. Agricultural expansion in Paraná occurred mainly in the period 1950-1970 and was "characterized by short-term agricultural systems leading to continuous and progressive environmental degradation as a result of economic policies and a totally inappropriate land parcelling and marketing system..." (Andreoli, 1993). Erosion has led to extensive loss of top soil, large-scale gullying (Figure 4), and silting of ditches and rivers. The use of fertilizers has risen as a consequence, up 575% over the period 1970-1986 and without any gain in crop yields. Loss of N-P-K from an average erosion of 20 t/ha/yr represents an annual economic loss of US$242 million in nutrients.
FIGURE 12: Water-based aquaculture in the Lakes Region of southern Chile
Analysis by Elwell and Stocking (1982) of nutrient loss arising from erosion in Zimbabwe shows similar significant economic losses in African situations. Stocking (FAO, 1986), applying data collected in the 1960s by Hudson to the soil use map of Zimbabwe, calculated an annual loss of 10 million tonnes of nitrogen and 5 million tonnes of phosphorus annually as a consequence of erosion (cited by Roose in FAO, 1994a). Roose (FAO, 1994a) also cites losses of 98 kg/ha/yr of nitrogen, 29 kg/ha/yr of phosphorus, 39 kg/ha/yr of lime and 39 kg/ha/yr of magnesium from soils of lower Côte d'Ivoire as a result of erosion. This loss is so severe that compensation requires 7 tonnes of fresh manure annually, plus 470 kg of ammonium sulphate, 160 kg of superphosphate, 200 kg dolomite and 60 kg of potassium chloride per hectare per year. Roose notes that it is not surprising that the soil is exhausted after only two years of traditional agriculture. Furthermore, these losses do not take into account additional losses due to harvesting and runoff. Roose summarizes by stating that action against soil erosion is essential in order to manage what he describes as a "terrible" chemical imbalance in soils caused by soil erosion. Estimates of phosphorus loss by erosion in the Republic of South Africa (Du Plessis, 1985) are R26.4 M/yr (US$ 10.5 million).
Economic losses tend to be higher in tropical countries where soils, rainfall and agricultural practices are more conducive to erosion, and reported rates of erosion are much above average. The World Bank (1992) reported that extrapolation from test-plots of impacts of soil loss on agricultural productivity, indicates some 0.5-1.5% loss of GDP annually for countries such as Costa Rica, Malawi, Mali and Mexico. These losses do not include offsite costs such as reservoir infilling, river sedimentation, damage to irrigation systems, etc.
Soil fertility is a complex issue and nutrient loss is not necessarily nor always a consequence of erosion. Erosion and soil loss is the end member of a variety of physical, vegetative and nutrient factors that lead to soil degradation. Global patterns of fertilizer application, as reported by Joly (1993), indicate however that rapidly rising levels of fertilizer utilization are required merely to maintain soil productivity from a variety of types of loss, including losses due to erosion and, more generally, to soil degradation.
In a study of 17 agricultural sub-watersheds in the Lake Balaton district of Hungary, Jolankai (1986) measured and modelled N and P runoff from a variety of agricultural land uses. He calculated that a selection of control measures (mainly erosion control) would reduce phosphorus loss by 52.8% at a cost of US$ 2500 per ha in remediation measures (in 1986).
Aquaculture is a special case of agricultural pollution. There are two main forms: land-based and water-based systems (Figure 12). Effluent controls are possible on land-based systems, however water-based systems present particular problems. Aquaculture is rapidly expanding in most parts of the developed and developing world, both in freshwater and marine environments. In contrast, coastal fisheries in most countries are declining.
The environmental impact is primarily a function of feed composition and feed conversion (faecal wastes), plus assorted chemicals used as biocides, disinfectants, medicines, etc. Wastage of feed (feed not taken up by the fish) is estimated to be 20% (Ackefors and Enell, 1992) in European aquaculture. Waste feed and faecal production both add substantial nutrient loadings to aquatic systems.
Additional environmental problems include risk of disease and disease transfer to wild fish, introduction of exotic species, impacts on benthic communities and on the eutrophication of water, interbreeding of escaped cultured fish with wild fish with consequent genetic change in the wild population.
Traditional integrated aquaculture systems, as in China, where sewage-fish culture is practised, can be a stabilizing influence in the entire ecosystem (Rosenthal, 1992). This is recommended, especially in developing countries where water and resources are scarce or expensive.
Eutrophic and hypertrophic lakes tend to be shallow and suffer from high rates of nutrient loadings from point and non-point sources. In areas of rich soils such as the Canadian prairies, lake bottom sediments are comprised of nutrient-enriched soil particles eroded from surrounding soils. The association of phosphorus with sediment is a serious problem in the restoration of shallow, enriched lakes. P-enriched particles settle to the bottom of the lake and form a large pool of nutrient in the bottom sediments that is readily available to rooted plants and which is released from bottom sediments under conditions of anoxia into the overlying water column and which is quickly utilized by algae. This phosphorus pool, known as the "internal load" of phosphorus, can greatly offset any measures taken by river basin managers to control lake eutrophication by control of external phosphorus sources from agriculture and from point sources. Historically, dredging of bottom sediments was considered the only means of remediating nutrient-rich lake sediments, however, modern technology now provides alternative and more cost-effective methods of controlling internal loads of phosphorus by oxygenation and by chemically treating sediments in situ to immobilize the phosphorus. Nevertheless, lake restoration is expensive and must be part of a comprehensive river basin management programme.