ASSESSMENTS OF ENVIRONMENTAL EFFECTS OF MARINE FISH FARMS 1

LITERATURE REVIEW ASSESSMENTS OF ENVIRONMENTAL EFFECTS OF MARINE FISH FARMS By Dr. I.E. Gonenç 1. Introduction and aim The aim of this report is to give a summary of the basic concepts concerning environmental assessments of marine fish farming and to give some practical examples how to use these concepts. The coast of Gulluk Bay is a most important natural resource, but also a zone of conflicts for many parties. There are great differences coasts in Turkey concerning topografhical, hydrodynamical and biological character, and in suitability for fish farms. In Asia, southern Europe and many other parts of the world, aquaculture, like agriculture, has a long history and is of great importance for the economy of individuals and the society. Hitherto, aquaculture has progressed very differently in Turkey. The development of marine and brackish water fish farming during recent years has been very rapid, without considering their environmental effects. The focus of this report is the environmental problems connected with one form of aquaculture – fish cage farms in coastal areas. Today, there is a lack of commonly accepted and scientifically tested guidelines and concepts to evaluate the environmental impact of fish farms in marine environments. This situation has permitted quite a lot of free speculation on the possible negative / positive effects of fish farms. The basic aim of this report is to try to give a survey of the factors, concepts and causal relationships collected from the recent literature that may be applied today in evaluations of environmental effects of fish farms in coastal environments. I have also regarded it as a good start in a project with participants from the UK (Arup, Locum, WATG) to try to identify “ the present state of the art”, to cordinate the concepts and to summarize this in a report for a wider readership. BASIC CONCEPTS The famous Vollenweider diagram (see Fig.1), and the analysis behind this load diagram, constitutes a fundemental base for practically all environmental assessments of phosphorus in limnic environments. For marine areas there is no such load diagram available today. In practice, this makes it impossible to make relevant evaluations of potential problems of fish farms on a regional basis and, for example, to calculate how large a farm could be in a given coastal area before defined negative ecologicial effects, like low oxygen concentrations (below 3 – 4 mg/l) or decreased transparency, would appear. It is not possible to make simple adjustments of the Vollenweider diagram to marine areas, because for such areas nitrogen is often the most limiting nutrient for bioproduction and because the physical conditions in the sea often have a profound impact on the biologicial conditions in a given coastal area. At present there is a fairly agitated debate in progress concerning the most limiting nutrient, N or P, in contexts of marine eutrophication. Most people, though, tend to agree that the most simplistic arguments in favour of an either /or sitiuation must be replaced with a more diversified picture where both elements are recognised as important, as are the forms of the nutrients (e.g dissolved or particulate), the season of the emission, the site of the emission and the geographical scale of the impact area and the effects. This is, anyhow, our viewpoint, and it will be stressed in this report. The relationship between nutrient dose and eutrophiciation effect may be clarified if the following factors are accounted for. The effect (E) is a funtion of the dose (D) and the environmental sensitivity (Wi) where also R = the residual term is defined: E = f (D, Wi) + R (1)........................................... THE DOSE TERM To calculate the potential increase of nutrient concentration in coastal waters caused by a fish farm, one must know the background concentration of the nutrient in the water mass as well as be able to quatify the emission of nutrients from the fish farm per time and the retention time of the water in the given area. The nutrient load can be divided in two posts: Particulate load and dissolved load. The particulate material, mainly feed loss and fish faeces, will primarily be deposited beneath the cage, but may also be spread to neighbouring areas: The stronger the currents, the greater the distribution. A feed conversion rate of 1.6 means that 1.6 kg feed is required to produce 1 kg fish. On a well – run farm, with a minumum loss due to high quality feed (i.e high energy content), the feed conversion rate may be lower than 1.5 Needless to say, it is in the interests of both the producer and the authorities to keep the feed loss and the feed conversion rate low. In Fig. 5 I have question marks at two, important environmental factors, the water exchange and the bottom dynamic conditions. Besides the nutrient load, it is also important to know the load of organic matter; the organic load from a fish farm should be related to the natural load and to other types of antropogenic load, e.g. from agriculture and industries. The flux of organic (and inorganic) material may be measured quite simply with sediment traps. The P – The particulate load generally dominates load from a fish cage on the surrounding waters; the dissolved load dominates the N – load, in contrast. The figures given in Fig. 6 should not be applied uncritically. ENVIRONMENTAL SENSITIVITY Different coast respond differently to one and the same nutrient dose. Besides the characteristics that have to be known to determine an increase in concentration from an emission from a point source (i.e. coastal volume and water retention time), it is also great importance to know the bottom dynamic conditions, i.e. if areas of erosion, transportation or accumalaiton prevail in the given area. This is illustrated in Fig. 6. If the fish cage is placed at a site or within a coastal area dominated by erosion and/or transportation processes, the possibilities for the oxygen (O2)- demanding organic material to accumulate beneath the cage will be minimized and the material from the cage will be spread to surrounding waters and sediments. If the cage is placed within an area dominated by fine sediments and accumulation processes, the spread will be smaller. Possible O2 problems will be accentuated if the bottom water exchange is small. If O2 poblems do arise in the sediments below the cage, this may lead to formation of lethal H2S – gas. Areas of erosion, i.e. bottoms dominated by sand and harder materials, are generally well exposed to wind – generated waves and currents and could therefore, for practical reasons, be less suited for fish farms. But the technical development towards solid offshore constructions is rapid. Resuspension of fine material can only, according to defination , occour in areas or erosion and transportation. During storm events, when base is lowered and the areas susceptible for resuspension are comparatively large, there is generally a large input of oxygen (from oxygen saturated surface water) cand also enhanced oxygen consumption from resuspended material. The internal loading of nutrients is also enhanced by resuspension. Thus, an area – typical sensitivity factor, which is quantitative terms should describe how a given coast responds to a given nutrient dose, must be regarded as a function of a several variables, like coastal volume, water retention time, mean depth and areas of accumulation, transportation and erosion as a measure of the potential for resuspension. The last – mentioned variables are not the least important. Data from various coastline exemplify that in coastal areas dominated by areas of erosion an transportation, a primary dose of nutrients may be used not only once for bioproduction, but several times due to internal loading when the pool of nutrients from the bottom sediments, or rather from the interstitial waters of the sediments, is brought up to the productive surface water. THE EFFECT TERM In many connections where environmental hazards are discussed, the term ecological effect is used very loosely. What is actually meant by ecologicial effect? It is important to give a clear definition of the effect term. In the Vollenweider model, the effcet term is defined by concepts like hypertrophy (extremely productive lakes), eutrophy (very productive), mesotrophy (moderately productive) and oligotrophy (low productive lakes), which are related to defined phosphorus concentrations. Many other effect terms are discussed in contexts of eutrophication, e.g. focussed on the bottom fauna community, algae, fish community, primary productivity, transparency or the oxygen conditions. From the perspectives of both the fish farmer and the environmental authorities, one may advocate that the oxygen status of the water is a critical effect term, which is also quite easy to measure. But then it is not primarily the mean O2 – concentration of the water mass that is requested, but rather the extremely low O2- concentraions. Crucial questions are : Where, when, and how often will low O2concentrations (<3 – 4 mg/l) appear in the bottom zone? There are, however, many reasons why the O2-concentration can not be be used as an effect term uncritically, e.g. because low O2-concentration may not appear until the situation is very serious, because low O2-concentrations may appear naturally in many deep holes, e.g. during summer stagnation and because thew O2-concentration of the surface water rarely decreases below the critical limit of 3 – 4 mg O2 per liter. The focus ought to be on the oxygen status of the deep water during the period with temperature stratification of the water mass (i.e. during summer). The O2concentration in the deep-water decreases in most stratified coastal areas as a result of O2consumption by degradation of settling organic matter. For coastal areas, a parameter called the deep-water oxygen depletion (DOD) can be used as a tool to specify the oxygen status. To determine DOD one must know: (1) The oxygen concentration, (2) the temperature and (3) the salinity of the deep water just after the onset of a temperature stratification and one or two times before the stratification breaks down. One would also require (4) the volume and (5) the retention time of deep water. With sediment traps, it is possible to determine the organic load to the deep water. Such data would make it possible to calculate the potential oxygen deficit measurements of the O2-concentration. Other practical effect terms are, e.g. the Secci-disk transparency, as a general indicator of water quality, and the chlorophyll content, as a measure of total phytoplankton biomass, which is an indirect measure of primary production. THE UNACCOUNTABLE RESIDUAL TERM This term is included here to emphazise the importance to quantify the relationships between dose, sensitivity and effect, and the impossibility to know everything to 100 %. The R – term may be given by the statistical measure called the degree of explanation ( r², which is defined as the square of the corelation coefficient, r). It can be used to quantify how well a model can predict empirical data. If the degree of explanation is 100% (r² =1), then there is perfect agreement between model data. Such models are not obtainable in ecological contexts. It is, however, always important to try to declare how well a given model can predict the reality, since there is a great difference between models that yield a 50% and a 80% degree of explanation. (POD) in the deep water without ECOLOGICAL EFFECTS The aim of this section is to give a brief discussion on concepts linked to ecological effects in general and to ecological effects linked to marine fish farms in particular. The most rational way to describe relationships between the state of the water system and factors influencing it is perhaps the use of different models. There are models of various levels of complexity. Ecological models describing water systems generally include interactions between biological, physical and chemical factors. The reliability of these models depends on how accurately they describe these factors and interactions. A necessary condition for a proper application of models is that the user understands the phenomena included in the model also the prerequisites and limitations of the model. One obvious difficulty in the practical application of ecological models, e.g. for planing purposes, is the very sophisticated nature of such models. To be meaningful, an ecological model must contain enough relevant biological knowledge. But this makes the application difficult: The amount of basic knowledge, which must be acquired, is high and the model might become too complicated for a user without special knowledge of ecology and mathematics (see the following figure). Using indicator species may partly solve this problem. The use of indicator species for describing the state of the community, or the state of the enviroment, presupposes an understanding of population ecology of the species and a detailed knowledge about autecology of the species. Solar energy fisheries OPEN SEA wind phytoplankton rainfall zooplankton runoff MF Inputs from the open sea mf IB Benthic fishes Benthic system Pelagic fishes migrations Pelagic system Human activities dredging pumping Coastal works nutrients detritus MF = macrophytobenthos mf = microphytobenthos IB = benthic invertebrates PRIMARY PRODUCTION AND NUTRIENTS A most important parameter for describing environmental effects of fish farms in marine areas is the primary production. It is important to establish quantitative relationships between nutrient loading, phosphorus and nitrogen concentrations of the waterbody and primary production. It must therefore be an important task to determine the critical nutrient load for different responses in the primary production. An increased load of nutrients may lead to the following responses in the phytoplankton community: Phytoplankton species composition, biomass and primary production are retained in original balance. Because of the natural self- purification ability, the structure and function of the whole ecosystem and the possibilities to use the marine environment are retained as before although extra nutrients are entering the ecosystem. The increase in nutrient load causes essential changes in the primary production. Structural changes in the phytoplankton community may disturb the functional balance of the system, until the ecosystem is changed to a new phase of succcesion, and at the end to a new balance. The loading is further increased and the consequence is an accelerated phase of eutrophication. The primary production of the planktonic algea is very high. The blue –green algae, which are of little value as nutrition for zooplankton, are dominating. The production of zooplankton and bottom animals is decreased in anoxic habitats and the balance of the whole ecosystem is seriously disturbed. Mostly producers and decomposers form the nutritional chains, while consumers, like fishes, are reduced in number. In areas where nitrogen is the limiting nutrient in the summer, blooming of blue – green algae may occour. These algae are able to bind nitrogen from the atmosphere and in this way enhance eutrophication. 2. The dose from the fish farm BACKGROUND It is customary when discussing environmental effects to calculate and compare the total dose of nutrients from various sources, e.g. from municipal water treatment plants, agriculture and fish farrns. We will try to demonstrate that such simplistic calculations may give a rather poor picture of the potential environmental effects linked to the different sources. As a background to the subsequent discussions, we would, however, like to give some examples of such general data on nutrients from various sources. The total discharge of nutrients from a fish cage is generaliy in the order of 10 – 20 kg P and 75 – 95 kg N per ton of produced fish and year. The lower values apply to smail fish, which foremost are produced in lakes. The higher values are more relevant for marine areas, where most of the large rainbow trout and salmon (> 1.5 kg) are produced. The value is based on today’s food types and is given per production quantity and not per maximum amount of fish in a cage at a given time. The values given are based on food coefficients, i.e. added dry fodder divided by fish growth in kilos in the range of 1.6 - 2.2. From these figures one may roughly say that the total discharge of nutrients from a fish farm of 50 tons/year would correspond to discharges from a purification plant from a town of about 7 000 inhabitants (assuming a 90 % reduction in the plant of phosphorus). One important question is now: “Is it relevant from an ecological point of view to compare total doses like this? If not - why?” FEED Dry feed as well as moist pellets is used in the production of salmonid fish species. In some countries the use of moist pellets and wet feed is prohibited. Composition of the feed In commercially manufactured feed the contents of protein, fat, carbonhydrates, nitrogen and phosphorus, as well as the energy content, are declared on the packing. Table 2 presents some values for moist pellets, high energy dry feed and low energy dry feed. It shouıd be noted that there are big differences between the various types of feed, a fact becoming apparent when running through brochures from feed manufacturers. Feed is continuously being developed and improved in order to achieve better growth rates and minimize the losses to the environment through excretion and faeces. Improvements are achieved through reguirements for raw material (fish meal and fish oils) of higher quality and by employing new technigues in the processing of the feed, e.g. eztruded feed. Table 2. Feed composition. Moist pellets Dry feed low energy 900 500 120 150 80 15 4.6 Dry feed high energy 900 450 240 100 72 10 5.2 Dry matter (g/kg) Protein (g/kg) Fat (g/kg) Carbonhydrate (g/kg) Nitrogen (g/kg) Phosphorus (g/kg) Gross energy (Mcal/kg) 325 170 60 50 27 4 1.3 Compoosition of the fish Trout and salmon produced in fish farms have high contents of fat (12 – 20 % ) deposited intramuscularly and as visceral fats. The composition of the fish (Table 3) will vary according to nutritional and physiological condition (e.g degree of sexual maturation). From Table 3 it should be noted that high values of fat, dry matter and energy are linked to each other. Table 3. Composition of Rainbow trout at stocking and harvest respectively. Fish for stocking Produced Fish < 1 kg Dry matter (g/kg) Protein (g/kg) Fat (g/kg) Nitrogen (g/kg) Phosphorus (g/kg) Gross energy (Mcal/kg) 250 – 300 150 - 160 50 – 100 23 – 28 3.5 – 4.5 1.5 – 2.5 > 1 kg 300 - 400 160 - 180 100 - 200 25 - 30 3.5 – 4.5 2.0 – 3.0 LOSS OF N AND P AS A FUNCTION OF FEED CONSERVATION AND FEED COMPOSITION The load of nitrogen and phosphorus is directly proportional to the feed content of nitrogen and phosphorus as well as to the feed conversion rate: Load (N, P) = Feed conversion rate * Feed (N, P) - Fish (N, P) In Figs. 11 and 12 this relation is shown using values of nitrogen (protein/6.25) and phosphorus normally found in commercial feed. As expected, by using feed with a low content of N and P in situations where a low feed conversion rate can be achieved, it is possible to reduce the loss of N and P to the surrounding waters considerably compared to values normally cited from sea farm production of salmon and trout. ANNUAL VARIATION OF LOAD The production and load from a fish farm varies considerably throughout the year with a high bad in the summer and a low load in the winter. The seasonal differences are mainly due to the variation in temperature. In some fish farms feeding is possible throughout the year and the fish can survive in the winter at low feeding rates. NITROGEN LOSS AND VALUE OF PRODUCTION There have been, and in the years to come there will be, significant resources spent to bring down the discharge of organic matter, nitrogen and phosphorus from agriculture, industries and public water treatment plants. It may therefore be very difficult to convince authorities, as well as the public opinion, that marine fish farms should be permitted to increase in numbers and in size when other sectors have to reduce their discharges. Political considerations will decide the allowed discharge levels of nitrogen and phosphorus from various polluting enterprises. These considerations should include a comparative estimation of the value of the production, the amount of discharge and the costs of water treatment. Exemplifying these kinds of considerations, the gross income (production value/raw material used/other materials used) from the Danish sea farm industry can be calculated. In the calculation given below variable costs are estimated for the production of 1 kg rainbow trout, size 2.5 - 3.5 kg, in 1985 (Table 8A). Similar calculations have been carried out for Danish agriculture (Table 8B) . The supply of nitrogen in commercial fertilizers and imported feed is compared to the output of nitrogen in the production sold (Miljöstyrelsen, 1984). Table 8A. Sales price from farm Stocking material Feed Other variables Gross income Production value : 32.00 DKK/kg : 5.50 DKK/kg : 7.25 DKK/kg : 4.00 DKK/kg : 15.25 DKK/kg : 150 DKK/kg : 380 000 tons N/year : 180 000 tons N/year : 560 000 tons N/year : 100 000 tons N/year : 460 000 tons N/year Table 8B. Commercial fertiliziers Imported feed Total supply Production sold Loss to the environment It can be seen that the total loss of nitrogen to the environment as a result of Danish agricultural production is in the magnitude of 460 000 tons/year. The gross income, as mentioned above (production value/raw material used/other rnaterials used), for Danish agricultural production in 1985 was about 25 biılion DKK (DDL, 1986). consequently, the value added in agriculture compared to the loss of N to the environment was about 50 DKK/kg N loss. The nitrogen outlet into Danish seas is not 460 000 tons /year. The exact amount of nitrogen due to atmospheric precipitiation and runoff is hard to quantify but it should be in the magnitude of 100 – 200 000 tons/year. The value added in agriculture, compared to the amount of nitrogen discharged, thus becomes abaut 125 - 250 DKK/kg N loss, i.e. in the same order of magnitude as seen for sea farms. In a comparative analysis, however, it would not be reasonable to use this figure alone as the agricultural discharge has supplementary undesirable environmental effects which are not seen in sea farming. One of the major inconveniences is the flow af nitrogen-rich waters into the ground water reservoirs, thereby increasing the nitrate concentration in drinking water supplies to unacceptable levels. The runoff of nitrogen-rich surface waters also creates undesired floristic and faunistic changes in lakes and streams. The evaparotian of ammonia from Danish agriculture is considerable, presumably in the magnitude af 100 000 tons/year, thus contributing to acid rainfall and acidificatian af surface waters. Taking this into consideration, it seems reasonable to assume that the value added compared to the nitrogen loss is higher in sea farming than in agriculture. Reducing the discharges from agriculture by 1 kg N would cost about 10 DKK (interest, depreciation, operational costs) should it be obtained by better use of manure, or 20 DKK should it be obtained by reducing the use of commercial fertilizers. Treatment of waste water from towns and industries costs about 15-20 DKK/kg (PFF, 1987). It can be seen that the value added in sea farming is considerably greater than the costs involved in removing 1 kg of nitrogen in agriculture or in public or industrial waste water plants. Based on socioeconomic considerations, it could be argued that it would be more advantageous to reserve the allowable nitrogen discharge for the kind of production achieving the highest value per unit of nitrogen lost to the environment. Following this line of thought, it would make good sense to keep or even increase the production from sea farming at the same time as limitations should be laid on productions in which nitrogen loss is greater compared to the value of the products. 3. Coastal Water Sensitivity The objective of this section is to give a brief general presentation of different parameters that may be used as sensitivity parameters in assessments of environmental effects from fish farms in coastal areas. Any coastal area may be described and classified from many perspectives and with a variety of biological, chemical, sedimentological and hydrodynamical methods. One and the same thing, e.g. the bioproductivity; may be defined and measured in difterent ways. The philosophy behind the following sensitivity parameters is that they must be easy to determine and apply in practical situations and that they are scientifically relevant and applicable to problems concerning environmental effects of coastal fish farms. COASTAL MORPHOMETRY Morphometrical key parameters Morphometrical parameters (Table 9) may be classified into the following three categories : Table 9. Frequently used morphometrical parameters. Size parameters Shoreline length Max.length Min.depth Water area Total area Section area Volume Form parameters Mean depth Mean width Mean coastal width Shore irregularity Mean slope Special parameters Exposure Fetch Effective Fetch Size parameters, such as shoreline length, maximum depth, water surface, total area (water area plus area of islands), section area and water volume. Form parameters, generally determined from size patameters, e.g. mean coastal width, mean depth and mean slope. The exposure (E), or topographical openness, which describes the openness of a given coast towards the sea. SEDIMENT TYPES The purpose of this section is to section is to discuss certain sedimentological concepts of interest in assessments of fish farms, especially the significance of areas of erosion, transportation and accumulation (see Figs 6). In the first place, it is important to distinguish metarial emanating from the fish farm from metarial of allogenic and autogenic origin. The allogenic material comes from external sources, like rivers. The autogenic material is produced in the coast, e.g. plankton and algae. There are the very important link between the organic load to the sediments, the character of the benthic community and the oxygen conditions; the higher the organic load, the greater the oxygen consumption, the lower the oxygen concentration and the fewer the animals. In defining the bottom dynamic conditions (erosion, transportation and accumulation), the focus is on the finer materials most easily set in motion / resuspension and having the highest capacity to bind pollutants. 1. Areas of erosion (E) dominate where there is no apparent deposition of fine materials (i.e. bottoms of sand, gravel, consolidated clays and/or rocks). 2. Areas of transportation (T) prevail where fine materials are deposited periodically (i.e. bottoms of mixed sediments), and 3. Areas of accumulation (A) prevail where the fine materials are deposited continuously (i.e. soft bottom areas). Beneath a fish cage, the amount of organic material is often significantly greater than further away from the cage. This is because the spill from the farm settles quite rapidly and also because the cage constitutes a physical obstacle which lowers the turbulent energy of the water mass; the movements of the fish in the cage also create a certain turbulence but this is in a restricted area and of another order of magnitude than what even moderate winds will create. If one would like to compare two alternative localities for a fish farm, then the situation in fig. 35 can be used as an illustration of an interesting principle. If the cage is located in the bay, there is the alternative for enclosure; if it is placed outside the bay, we have an alternative for spread and dilution. The potential ecological effects are closely linked to the fate of the particulate emission from the cage: ın which direction will this material go and where will it end up? Table 11. Typical values of different sediment parameters from the Swedish coastal zone in different bottom dynamic conditions. (mg = 10³, ug = 109 , ws = wet substance, ds = dry substance) Eroision Transportation Accumulation PHYSICAL PARAMETERS Water content (% ws) Organic content (loss On ignition, % ds) NUTRIENTS (mg/ g ds) Nitrogen Phosphorus Carbon METALS Iron (mg / g ds) Manganese (mg / g ds) Zinc (ppm) Chromium (ppm) Lead (ppm) Copper (ppm) VERY TOXIC METALS Cadmium (ppm) Mercury (ppb) < 50 <4 <2 0.3 – 1 4 - 10 50 - 75 >10 10 – 30 0.3 - 1-5 >5 > 50 > 75 < 20 20 - 50 >1 < 10 < 0.2 < 50 < 25 < 20 < 15 < 0.5 < 50 10 - 30 0.2 – 0.7 50 – 200 25 – 50 20 – 30 15 - 30 0.5 – 11.5 50 – 250 > 50 > 30 > 20 0.1 – 0.7 > 200 > 30 > 1.5 > 250 In the alternative for enclosure, high concentrations of nutrients and organic material will appear within a limited area at and around the cage, but the main coastal area will be comparatively little influenced by the emissions from the farm. The inner bay will act as a “purification plant”, a “sink” for the open water areas – provided the bay actually has areas of accumulation. In the alternative for dilution, nutrients and organic matter will be distributed over a larger waterbody; the maximal concentrations will be lower in water and sediments, but the mean concentrations in the entire coastal area will be higher. A fish farm located in a small bay may give rise to significant, visible, negative effects, like reduced transparency and increased growth of attached algae, where the fish farm may be the dominating source of emissions of nutrients and organic material. If, on the other hand, the same farm is located in a larger coastal area, it may be practically impossible to link negative effects to emissions from the farm because of other more dominant emissions and / or because of the character of the coast (e.g. a rapid water Exchange). WATER EXCHANGE It is easy to motivate why it is important to know the water Exchange in coastal areas: No concentrations of nutrients or toxins from point sources, like fish farms, can be evaluated without knowledge of the water retention time. If concentrations cannot be predicted, it is also practically impossible to predict ecological effects. But the water Exchange varies in time and space. It can be driven by many processes, which also vary in time and space. The importance of the various processes will vary with the topographical characteristics of the coast, which do not vary in time, but vary widely between different coasts. The water Exchange sets the framework fort he entire biotic life; the prerequisites for life are quite different in coastal waters where the retention time varies from hours to years. With traditional methods, it is generally very laborious and costly to determine the water retention time. It is important to note that the retention time fort the surface water, i.e. the water above the thermo – and/or halocline, generally is much shorter (10 times or more) than that for the deep water. In certain narrow, deep fjords, the retention time of the deep water may be as long as 5 to 10 years; in certain open coasts, the deep water may be renewed by most storms. The deep water Exchange rate is important in many contexts, since oxygen problems primarily appear in bottom areas with a heavy organic load and a slow water Exchange. There are no simple methods available today to determine the deep water retention time. The aims of this section are: • • to give a brief discussion on basıc hydrodynamical concepts for coastal areas; and to present some practically useful methods in aquaculture to estimate the surface water retention time. Hydrodynamical concepts Many factors influence the water Exchange; all the processes illustrated in fig. 40 can drive and / or mix coastal waters. They vary in importance with coast type and time. From a very basic perspective, one can argue that all driving energy emanates from the sun and fig. 41 gives one example of how the solar energy may be split up into different sub- processes that drive and / or mix coastal waters. The fresh water discharge (Q in m³/s) is the amount of water entering the coast from tributaries per time unit. In small bays with large tributaries, the Q-factor may be the most important factor fort he water retention time. Tides . When the tidal variation is larger than about 40 cm, it is an important factor for the surface water retention time. Water level fluctuations always cause a flux of water. These variations may be measured with simple gauges. They vary with the season of the year and are important for the retention time of shallow coastal areas. Thus, the mean depth is a useful coastal parameter. Boundary level fluctuations. Warmer and lighter coastal water is separated from colder and denser water by a thermocline. Salter and denser water is separated from lighter fresh water by a halocline. These boundary layers may be quite stable and act as “bottoms” or “glide surfaces” with a profound effect on water Exchange and water quality. Fluctuations in these boundary layers may be very important for surface and open coasts. Local winds may create a water Exchange in all coastal areas, especially in comparatively small and shallow coasts. Thermal effects. Heating and cooling, e.g. during warm summer days and nights, may give rise to water level fluctuations which may cause a water Exchange. This is especially true in shallow coasts since water level variations in such areas are more linked to temperature alterations in the air than is the case in open water areas. Coastal currents are large, often geographically concentrated, shore-parallel movements in the sea close to the coast. They may have an impact on the water retention time, especially in coasts with a great topographical openness. From this it is clear that the morphometrical characteristics of the coast are important to rank the driving processes for the water exchange. In theory, it may be possible to distinguish driving processes from mixing processes. In practice, this is often impossible. Surface water mixing causes a change in boundary conditions, which causes water exchange, and so on. The boundary conditions of the watermass have a great impact on, e.g. the spread and effect of emissions from fish farms. The water around our coasts is generally stratified / layered. This layering effect is enhanced during summer when the surface water is warm and therefore lighter than the deep water. PROBLEM DIMENSIONING OF ENVIROMENTAL HAZARDS An initial assessment or dimensioning of the enviromental impact of nutrient discharges from fish cages must include consideration to the two question marks in Fig. 5. Subsequently, we will use formulas (2) and (5) to answer the following two questions: what is a probable increase in nutrient concentrations in the water mass from a given fish cage enterprise? Will the spill be spread / diluted or enclosed / concentrated. The second question raises a third; it is evident that the conditions within and just below the cage will be more influenced by the spill than waters at a distance from the cage. The question concerning the geographical definition of impact area will be circumvented here since formulas (2) and (5) are based upon certain premises concerning the areal definition of the coast. So, we will not discuss site-typical information concerning the actual cage site but rather area-typical information concerning the entire coastal area defined by the limitation lines towards open water areas drawn at topographical bottle necks to yield a minimum of topographical openness (E). To determine a possible increase in concentration, one must know the following morphometrical parameters: • water volume of the coastal area (V); • water retention time (T); to determine these values, one must also know: • the topographical openness (E ), and • the mean coastal width (W); which in turn requires data on: • total area (the limitation line should be drawn at the topographical bottle necks at the sea), • coastline length. To determine the potential bottom dynamic conditions dominatinating in the given coastal area, the following additional information is required (from Formula 2): • the mean slope (xm); • the form factor (Vd), which in turn is based on: • mean depth and • maximum depth. It should be noted that all these data emanete from simple bathymetric maps. Concluding remarks Due to the fact that the coastal landscapes are very different in many countries, and because these countries have different legal approaches to marine fish farms, it is neither possible nor meaningful to try to develop common guidelines for all countries on environmental impact assessment of fish farms. However, there is a framework or backbone of generel concepts and methods that may be applied when working with these issues and the focus of this report is on such general concepts. The idea with this report has been to present the basic concepts of one form of environmental hazard analysis, dose sensitivity- effect-analysis for defined coastal recipients. How can dose, sensitivity and effect ebe defined and determined in a practical manner? How is it possible to conduct a probelem dimensioning and figure out if a fish cage farm would cause major or minor environmental problems in a given coastal area? I have tried to answer questions of that kind and we believe that this is one important piece in the puzzle that has to be known when discussing the benefits and drawback of marine fish farms in broadere economic ecological contexts.