Cover Image
close this bookGuidelines for the promotion of environmental management of coastal aquaculture development (1992)
View the document(introduction...)
View the documentPreparation of this document
View the document1. Introduction
View the document2. Guidelines for the promotion of environmental management of coastal aquaculture development
View the document3. Coastal aquaculture and the environment: The context
View the document4. Factors influencing environmental performance of coastal aquaculture
View the document5. Assessment of environmental hazards and impacts of coastal aquaculture
View the document6. Options for environmental management of coastal aquaculture development
View the document7. References
Open this folder and view contentsAnnexes

5. Assessment of environmental hazards and impacts of coastal aquaculture

92. In this section, general issues of marine pollution assessment are addressed (5.1). Selected aquaculture-specific methods for pollution assessment are then presented (5.2). Last, the role and functions of environmental impact assessment (EIA) are described (5.3).

5.1 General Considerations on Marine Pollution Assessment

93. Within the context of coastal aquaculture development and environment, it appears to be important and useful to address definitions and concepts related to marine pollution assessment and prevention.

Marine pollution

94. GESAMP gives following definition of marine pollution (GESAMP, 1991b):

“Marine pollution means the introduction by man, directly or indirectly, of substances or energy into the marine environment (including estuaries) resulting in such deleterious effects as harm to living resources, hazards to human health, hindrance to marine activities including fishing, impairment of quality for use of sea water and reduction of amenities.”

5.1.1 Environmental capacity

95. GESAMP further advocates the environmental capacity concept (GESAMP, 1986; Pravdic, 1987). The environmental (also known as receiving, absorptive or assimilative) capacity is defined as a property of the environment, a measurement of its ability to accommodate a particular activity or rate of an activity, such as the discharge of contaminants, without unacceptable impact: it is the “ability of a receiving system or ecosystem to cope with certain concentrations or levels of waste discharges without suffering any significant deleterious effects” (Cairns, 1977, 1989).

96. An important characteristic of both, the above marine pollution definition and the environmental capacity concept, is the discrimination between “contamination”, meaning increased presence of substances in the environment as a result of human activities but with no significant adverse effects, and “pollution”, signifying the occurrence of adverse effects. The distinction between the terms is important since it implies that environmental change resulting from human activities may, or may not, be judged to have adverse effects. The boundary between these two regimes requires a definition of “acceptability”. Irrespective of where this boundary is drawn, the concept of acceptable change remains valid.

97. The environmental capacity approach works well as an interactive environmental management strategy. Other traditionally-used complex strategies, based on environmental quality objectives or simple but readily enforceable strategies - such as those based on uniform emission standards, maximum allowable concentrations in effluent, the black/grey/white lists or the application of principles of best practicable means available - are considered as simple components of this adaptive, interactive strategy. A short description/discussion of traditionally-used strategies is given in Annex 3.

98. Assessing the environmental capacity is a scientific approach which requires technical and socio-economic inputs as parallel, interactive and complementary activities in decision-making in integral, environmentally compatible, development planning. It emphasizes the objectivity and independence of technical inputs and their influences on decisions related to socio-economic feasibility. It also emphasizes that the acceptability of environmental impact rests on much more than political considerations. Such acceptability can be determined scientifically, assuming that the environmental capacity can be quantified. The environmental capacity approach seeks to define the critical load and in its application seeks to keep actual inputs as far as practicable below this. Once the environmental capacity of a given substance is determined, it can be apportioned for various resource uses and needs.

99. The methodology for the assessment of the environmental capacity, which is site- and contaminant-specific, uses critical pathway analysis for both conservative and non-conservative contaminants and establishment of environmental quality objectives, criteria and standards. Faced with the inevitability of several sources of uncertainties in real situations, a probabilistic approach is used as an alternative to deterministic analysis. The approach proposed is decision analysis. The methodology recommended (GESAMP, 1986) consists of three decision-stages (see Figure 8). Socio-economic goals (priorities and objectives) are assessed in the planning stage, considering present and future use of resources. In the preliminary scientific assessment stage the environmental capacity is derived and quantified, resulting in the setup of allowable inputs. Finally, monitoring provides a continuous test of whether the environmental capacity is balanced, exceeded or underutilized. Consequently, adaptive and/or corrective measures may be required.

100. In this context, the role of science is to assess and predict man-induced environmental change. The key scientific issues and disciplines required can be grouped into two categories:

(a) the sources, transport, transformation, and fate of substances introduced to the marine environment

In this category, the distribution of substances is related to their sources, which involves physical, chemical and biological science with major emphasis on physical and chemical oceanography. Readers interested in the state-of-the-art of coastal modelling in relation to transport, dispersion and fate of contaminants disposed of in the coastal environment are referred to GESAMP (1991a).

(b) the effects of these substances on organisms, including man, and resources and amenities in the marine environment

This second category involves the translation of resulting exposures into their effects on organisms, man and amenities, which are covered primarily by studies on toxicology and biological effects.

Parameters

101. Any evaluation for a potentially harmful chemical substance must take into account two types of factors, the first intrinsic with the substance, the second related to the extrinsic conditions and their reciprocal interactions. Basically, the necessary scientific parameters are:

Quantity:

production, uses, discharge patterns, loads, sources

Distribution:

physico-chemical characteristics, affinity for environmental compartments

Persistence:

kinetics of hydrolysis, photolysis, biodegradation

Bioaccumulation:

n-octanol/water partition coefficient, metabolic pathways in different organisms

Toxicity:

measures of biological activity of the substance (ideally from cells to ecosystems)

Ecosystem typologies:

biotic and abiotic characteristics and structure and functions of ecosystems

Time-scale of events



Figure 8: Methodology for the assessment of the impact of pollutants on the marine environment (from GESAMP, 1986)

Impact prediction

102. Within the environmental capacity approach, hazard assessment (Landner, 1989, 1988; Bro-Rasmussen and Christiansen, 1984) is a key scientific tool for predicting potentially adverse effects of the discharge of pollutants on the basis of both their inherent properties and the probability of exposure to such substances which determines environmental damage to organisms. In other words, hazard assessment is based on the relationship of the expected environmental concentration of a chemical substance (to which target organisms are potentially exposed) and the toxicological properties of the substance, i.e., the predicted concentrations with adverse biological effect (Cairns et al., 1978). The prediction of the environmental concentration starts with the determination of exposure-related data, which refers to the rate of chemical substance input, the properties of the substance and the environment. The persistence and the distribution of the substance are evaluated from data on physico-chemical characteristics, biogeochemical behavior, biodegradability, bioaccumulation potential and bioavailability. Biological effects are predicted on the basis of acute and chronic toxicity studies or are calculated on the basis of quantitative structure activity relationships (QSARs) (Könemann, 1981; Halfon, 1989; Boudou and Ribeyre, 1989; Lloyd, 1991a, 1991b).

103. In summarizing, the hazard assessment process should yield clear predictions of (1) the transport pathways and rates, (2) the likely environment compartment (water, sediment and/or biota) where the substance/material is likely to accumulate and (3) the likely effect of the substance/material at a given target site or on a given organism or set of organisms. However, predictions should be made not only of chemical concentrations but also of acceptable biological effects. Limits to changes in either a stress response of a given species and/or a community response should be clearly stated. Regulatory measures may then be adopted upon comparison of the predicted environmental concentration (in water, sediments and organisms) and the information on lowest concentration where adverse biological effects can be expected.

5.1.2 Monitoring

104. Through monitoring, experts can provide coastal resource users and administrators involved in environmental management with information on:

(a) whether the condition of the environment is improving or deteriorating;

(b) whether any contemplated management activity, be it for development of a sector other than fisheries or fisheries or aquaculture itself, has had an impact on the environment;

(c) whether individual operators are complying with regulatory requirements.

105. The design of any monitoring programme should be based on clearly-defined objectives and the formulation of testable hypotheses. Monitoring is essential to verify impact predictions in any hazard assessment exercise. It must also be recognized that monitoring in most cases has to be a long-term exercise to be useful and efficient. Since monitoring can be very costly, data should only be collected that:

(a) are required to satisfy the objectives;
(b) are amenable to meaningful interpretation; and
(c) have known precision and accuracy.

106. Otherwise, technical and financial resources will be wasted and, in the case of compliance monitoring, the production of data of doubtful quality may limit their legal acceptance.

107. More specifically, adequate ecological monitoring requires:

(a) the measurement of (i) contaminant levels, (ii) extent of physical modification and/or (iii) related effects on the environment;

(b) the measurement of the rate of contaminant input or of frequency and dynamics of physical modification;

(c) the measurement of effects on identified target(s) exposed to environmental change.

108. Monitoring programmes should include both physico-chemical and biological aspects so that where possible measured biological responses can be related to specific chemical doses and/or physical modification.

109. The success of monitoring depends on the implementation of a strong quality assurance programme. A quality assurance programme has two basic components. The first, quality control, includes activities designed to ensure that the sampling and analytical techniques are adequate for the intended purpose. This forms the basis for deciding if a monitoring programme is adequate to significantly detect assumed (predicted) changes. The second, quality assessment, provides an ongoing basis by which the quality of data is maintained at the required level. This is accomplished through the use of standardized procedures, standard reference material and inter-laboratory comparisons.

110. Biological monitoring can provide a measurement of the direct effect of adverse water and sediment quality on organisms by assessing the extent to which a specific biological response deviates from a normal value. Measurements of effects the individual organism level include reduced growth rates, susceptibility to disease, or mortality of sensitive life-stages. Measurements of community structure changes can be readily equated with harm, particularly if species of commercial or conservation value have been reduced in number. Biological monitoring should be integrated with chemical monitoring, so that the extent to which the measured effects can be ascribed to specific chemicals can be established. This is true also of the monitoring of those chemicals that can be accumulated in the tissues of organisms, particularly those harvested for human consumption. The main objective here is to evaluate the concentrations found in the context of relevant acceptable daily intakes. But the data should also provide information on the potential harmfulness of such concentrations to the organisms themselves. At present, many concentration-effect relationships established for organisms are based on levels of chemicals in the ambient water and not on the amounts accumulated in the tissues.

111. Surveillance differs from monitoring in that predictions are not tested but target sites or organisms are surveyed to ascertain whether or not there are detectable differences between the surveyed site and control site.

112. However, the systematic collection of first-hand information on apparent pollution events or other visible environmental deterioration by fishermen and aquaculture operators may - in some cases - prove very useful for various purposes such as:

- early detection/early warning schemes (e.g., on oil spills, phytoplankton blooms)

- preliminary demarcation of pollution-exposed areas (e.g., changing patterns of distribution/expansion of contamination)

- record-keeping of chronology/history of events (frequency, duration, time of year, etc.)

113. Such activities, which may be better termed “regular observation” or “watchkeeping” of the environment, could eventually support ecological surveillance and/or monitoring programmes.

5.2 Selected Aquaculture-Specific Pollution Assessment Methods

114. When assessing environmental change resulting from aquaculture, it is important to distinguish between:

(a) output (waste load) and consumption (e.g., of oxygen, paniculate material, phytoplankton) by culture practices or organisms; and

(b) related ecological effects.

115. For example, the quantification of the total suspended solids that might be generated by an aquaculture operation does not equate to the effect that such output might have on the ecology of the farm site. Unfortunately, this interaction is often discussed as though the two issues were the same.

116. Selected aquaculture-specific pollution assessment methods are described below. It is emphasized that environmental impact assessment methods are being improved (see Silvert, in press).

5.2.1 Mass balance estimates of waste production

117. So far, most estimates of waste production have been published for temperate carnivorous fish species, in particular salmonids. A graphic illustration of the fate of waste material released from intensive fish farming is given in Figure 9. It is possible to estimate the amounts of uneaten food, faeces and excreta produced from data on feed qualities and quantities, food conversion ratio (FCR) values, digestibilities and faecal composition, and to produce mass balance equations for various waste parameters, such as nitrogen, phosphorus or carbon, solids or biological oxygen demand (Beveridge et al., 1991).

118. Some general equations follow (modified from Iwama, 1991) which enable the estimations to be made of the output of the total mass of particulate organic matter deriving from uneaten food and faeces.

Given:

UW

=

percentage uncaptured feed waste/100 (i.e., ratio of total food uncaptured to total food fed)

F

=

percentage faecal waste/100 (i.e., ratio of total faecal waste to total food eaten)

FCR

=

food conversion ratio (weight of food fed/weight gained)

PD

=

production (increase in fish biomass)

O

=

total output of particulate organic matter

Then:

TF

=

total food fed

=

PD x FCR

TU

=

total food uncaptured

=

TF x UW

TE

=

total food eaten

=

TF - TU

TFW

=

total faecal waste

=

F x TE

O

=

TU + TFW



119. The total amount of food fed (TF), if unknown, can be derived from estimates for production (PD) and FCR values. The significance of FCR values (and water content in feeds and fish) for waste production estimates are discussed by Hopkins and Manci (1989). FCR values from cages appear higher than those from ponds, possibly indicating higher food losses (Beveridge, 1984). There have been few attempts to estimate directly the proportion of uneaten food (UW), partly because it is difficult to distinguish between the food and the faecal components of solids collected. Estimated values for UW can range from 1 to 30% or more. Faecal waste production can be estimated from studies on the digestibility of main diet components. Dry weight based estimates for whole diet digestibilities seem to compare well with values for F, which range from 25 to 30%.

120. With the total content of carbon, nitrogen and phosphorus in feeds and faeces known, it is possible to estimate the output of each of these components in the uneaten food and faeces fractions using following general equations:

UM

=

mass of C, N, or P output from uncaptured food

DM

=

TF x UW x K

EM

=

mass of C, N, or P output from eaten food

EM

=

(TF - TU) x K x E

TM

=

total mass of C, N, or P output from both uncaptured and captured food

TM

=

UM + EM

where:

K

=

percentage of each component in food/100

E

=

percentage of each component in faeces/100


Figure 9: The fate of waste material released from intensive fish farming (from Gowen et al., 1990)


Figure 10: Phosphorus and nitrogen load from cage fish farming, expressed in kg per ton produced fish per season. The feed coefficient used is 1.5 and the content of phosphorus and nitrogen in the feed is considered to be 0.9% and 7.2% of wet weight, respectively. The desorption from the sediment is considered to be 50% of the sedimented (particulate) phosphorus and nitrogen (Enell and Ackefors, 1991)

121. Given the total content of a component in the fish, the output of this component can also be estimated as being equivalent to the difference between the content of the component in the food and that retained in the fish. Wallin and Hakanson (1991) give following equation to estimate the nutrient load from fish farms:

L

=

P x (FC x Cfeed - Cfish)

where:

L

=

nitrogen and phosphorus load (kg tot-N & tot-P/year)

P

=

fish production (kg wet weight/year)

FC

=

feed coefficient (kg wet weight feed/kg fish production)

Cfeed

=

nitrogen and phosphorus concentration in feed (% wet weight)

Cfish

=

nitrogen and phosphorus concentration in fish (% wet weight)

122. A slightly different and perhaps more detailed description of these relationships is presented by Ackefors and Enell (1990):

The equation for the phosphorus load is:

kg P

=

(A x Cdp) - (B x Cfp)

The equation for the nitrogen load is:

kg N

=

(A x Cdn) - (B x Cfn)

where:

A

=

wet weight of dry pellets used per year (normal water content in dry pellets is 8-10%)

B

=

wet weight of fish produced per year

Cd

=

phosphorus (Cdp) and nitrogen (Cdn) content of dry pellets, expressed as % of wet weight

Cf

=

phosphorus (Cfp) and nitrogen (Cfn) content of the fish, expressed as % of wet weight

123. In summarizing, these general mass balance equations serve to roughly estimate waste production from fish farms (see Figures 10 and 11). It must be noted, however, that they are based mainly on a number of assumptions (e.g., when accounting for feed losses) and results from laboratory studies (e.g., estimations of digestibilities). Additional aspects need to be considered such as diurnal and seasonal variability, variations between different species, effects of temperature, body size, health, feeding rates, quality of nutritional components, synergistic/antagonistic effects of one dietary component on the digestibility of another, influence of processing and non-nutritional constituents on digestibility. Further, culture system and management factors will also influence waste load composition, for example, in terms of suspended or settleable solids, and nitrogen or phosphorus species.


Figure 11: Mass balance for nitrogen and phosphorus flow in a fish farm (from Wallin and Hakanson, 1991)


Figure 12: The diagrammatic representation of the horizontal displacement of organic waste showing the relationship between water depth, current velocity, and settling velocity of waste particles (from Gowen et al., 1989)

5.2.2 Modelling organic enrichment of the benthos

124. The dispersal and loading of paniculate organic waste to the sediment will depend on the amount of waste produced, surface area of the farm, water depth, current velocity, and the settling of the waste particles. Estimates for the waste-loading over a potential area affected can be derived (Gowen et al., 1989) from following equation (see also Figure 12):

d

=

D x Cv/V(1 or 2)

where:

d

=

distance dispersed (horizontal distance travelled by particle)

D

=

water depth

Cv

=

current speed

V

=

settling velocity of waste particles (uneaten food and faeces)

125. It is important to distinguish between uneaten food and faeces particles since the settling velocities of the two are different, in practice, however, it is likely that there is a broad size and density spectrum of waste particles and hence settling velocities. Particles may also break down into smaller particles. Variations in current speed and direction are accounted for in the model by Gowen et al. (1989). Other limitations are due to lack of consideration of various aspects such as possible consumption of uneaten food by wild fish, possible resuspension of sedimented material, differences in bottom characteristics, effects of benthic organisms and other microbiological and chemical processes on the deposited organic particulate matter (Holmer, 1991; Holmer and Kristensen, 1992).

5.2.3 Assessment of effects on the benthic ecosystem

126. There is a variety of benthological methods to assess physico-chemical changes and biological responses resulting from organic enrichment (see for example Viarengo and Canesi (1991); Frid and Mercer (1989); O’Connor et al. (1989)).

Bottom types

127. The occurrence of organic enrichment effects will depend on conditions of bottom dynamics and water exchange, which determine sediment types. Hakanson et al. (1988) give methods for determination of bottom dynamics conditions and suggest the following classification of bottom types which reflect the influence of wave action and prevailing currents in coastal areas:

(a) 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, so-called erosion bottoms);

(b) areas of transportation (T) prevail where fine materials are deposited periodically (i.e., bottoms of mixed sediments, so-called transportation bottoms);

(c) areas of accumulation (A) prevail where fine materials are deposited continuously (i.e., soft bottoms; so-called accumulation bottoms).

128. Accumulation bottoms are often characteristic of sheltered coastal areas where many aquafarms are located. On A-bottoms, where the organic matter is deposited over a rather long period, effects of oxygen depletion are frequent, and it is to be expected that the bottom fauna is more severely affected than it would be on other sediment types. Organic matter released on T-bottoms spreads further, and the effect, although likely to be less severe, is dispersed over a larger area.

Sediment metabolism

129. Physico-chemical parameters measured in investigations included current speeds, sedimentation rates using sediment traps, sediment thickness and density, sediment particle size, sediment chemistry of the solid phase and pore waters (water content, dissolved oxygen content, content of inorganic and organic, soluble and particulate components, alkalinity, redox potential) and near-bottom water chemistry (dissolved oxygen content, release of nutrients such as ammonium, nitrate, phosphate and gases such as hydrogen sulphide and methane).

130. Based on physico-chemical parameters measured, sediment metabolism patterns have been analysed which indicate complex chemical processes combined with aerobic and anaerobic microbial activities. It can be stated that high, localized deposition rates of paniculate organic matter are associated with metabolically very active sediments. The microbial activity is stimulated and the demand for oxygen in microbial processes and for re-oxidation of reduced mineralization products increases to the extent of oxygen depletion resulting in a net production from the sediment of the by-products of anaerobic metabolism (ammonium, hydrogen sulphide and methane). The reader interested in further discussion on related effects and methodologies is referred to Holmer (1991); Holmer and Kristensen (1992); Weston and Gowen (1988); Kupka-Hansen et al. (1991); Kaspar et al. (1988); Lumb (1989).

Effects on benthic fauna

131. Bottom fauna living in and above sediments can be used as an indicator in aquaculture pollution studies, since the benthos is fairly stationary.

132. Bottom fauna community structures can be disturbed by organic enrichment of the sediment to the extent of complete disappearance of macrofauna. It is generally assumed that a macrobenthic community subject to increased organic loading, either spatially or temporally, will exhibit (Weston, 1990):

(a) a decrease in species richness and an increase in the total number of individuals as a result of the high densities of a few opportunistic species;

(b) a general reduction in biomass, although there may be an increase in biomass corresponding to a dense assemblage of opportunists;

(c) a decrease in body size of the average species or individual;

(d) a shallowing of that portion of the sediment column occupied by infauna;

(e) shifts in the relative dominance of trophic guilds.

133. Several methods can be used to analyse changes in bottom community structures. For example, an area can be divided into different pollution zones using the biomass and abundance values and the occurrence of indicator species, which exhibit specific reactions to organic pollution. Often, so-called SAB curves are used (Weston, 1991), which display the changes in species number, abundance and biomass along a gradient of organic pollution (see Figure 13). A variety of diversity indices is in common usage (Shannon-Wiener index, evenness, Sanders rarefraction technique). Disturbance at individual sites may be determined by plotting the number of individuals in geometric classes of abundance, or by the abundance, biomass comparison (ABC) method whereby the cumulative dominance in terms of abundance and biomass are plotted against a logarithmic scale of species rank.

134. It should be noted that most methods have advantages and limitations in the assessment of aquaculture pollution effects on the benthos. Results from related studies are, in some cases, variable. Quantitative appraisal of parameters of benthic community changes may, in some cases, be difficult. And, changes, when occurring, are not always attributable to organic enrichment. Prediction of ecological effects however appears to be possible in qualitative terms based on empiric evidence. For further reference on benthological studies, see Gray et al. (1992); Lauren-Maatta et al. (1991); Holmer (1991); Weston (1991); Weston (1990); Brown et al. (1987).

5.2.4 Modelling hypernutrification and eutrophication

135. Hypernutrification (nutrient enrichment) and eutrophication (increase in primary production) of open coastal waters due to aquaculture is unlikely, but may occur in semi-enclosed coastal embayments which have restricted water exchange with more open waters.


Figure 13: Trends in areal species richness (S), biomass (B) and abundance (A) of macrofauna with distance from two salmon net-cage mariculture sites. Data in upper figure from Loch Spelve, Scotland, from Brown et al. (1987). Data in lower figure from Puget Sound, northwestern United States, from Weston (1990). (from Weston, 1991)

136. Attempts to model hypernutrification and eutrophication at present appear to be considerably less successful in coastal waters than in inland waters. The same basic principles of relating nutrient concentration to phytoplankton growth together with a dilution or water body-flushing term may apply. However, difficulties in modelling coastal ecosystem responses to nutrient enrichment are generally related to the influence of salinity stratification and tidal mixing, particularly in embayments and estuaries. Also, the boundaries of the affected area are often difficult to define. It is emphasized that, as a consequence of the complex linkages between biological, chemical and physical processes, these models are area-specific and as such their wider use is limited. Despite the range of limitations and the assumptions to be made, such models may be used to provide ‘best’ and ‘worst’ case scenarios.

Hypernutrification

137. The extent of hypernutrification depends on the size of the farm and the hydrography of the water body within which the farm is located. Thus the volume of the water body, its rate of exchange with the adjacent sea, the onset and duration of vertical stratification and the extent of horizontal advection, all have an important bearing on the level of hypernutrification.

138. An example for an approximative assessment follows. Assuming complete dispersion of waste nitrogen throughout a semi-enclosed water body, Gowen et al. (1989) presented the following approach to estimate the equilibrium increase of soluble nitrogen concentrations:

Ec

=

N x F/V

where:

Ec

=

equilibrium rise in concentration (level of hypernutrification)

N

=

daily output of soluble nitrogenous waste

F

=

flushing time of the waterbody in days

V

=

volume of the waterbody

Two methods were used for estimating flushing time F or dilution rate D, whereby D = 1/F:

(a) Tidal exchange method:

D

=

(Vh - VI)/T x Vh

where (Vh -VI) is the volume exchanged every tide, and:

Vh

=

high water volume of waterbody

Vl

=

low water volume of waterbody

T

=

tide period, in days

139. This method assumes that the mean volume of the water body is greater than the tidal volume, which is in turn greater than the volume of river inflow per tide. The method also assumes that there is complete mixing and that none of the water which leaves the basin on the ebb tide returns on the flood, which means that it fails to account for incomplete exchange.

(b) Salinity and river flow method:

D

=

R x S0/V(S0-S)

where:

R

=

rate of river inflow

S0

=

mean salinity of seawater flowing into water body

S

=

mean salinity of the outflow

V

=

volume of the waterbody

140. This method assumes a steady state, i.e., uniform river inflow. Direct measurements of ammonium in the water body (in this case a Scottish loch) as well as in the immediate vicinity of the fish farm, however, showed a spatial distribution of ammonium indicating only a very localized hypernutrification around the fish farm.

Eutrophication

141. It is, at present, impossible to predict eutrophication resulting from hypernutrification caused by fish farms. The consequences of hypernutrification in terms of enhanced primary production and phytoplankton standing crop are complex, and respective relationships are still poorly understood. Enhanced primary production could occur without an increase in standing crop, if additional biomass is rapidly removed, for example by grazing. Direct measurements of primary production are therefore likely to be more informative than simple estimates of algal biomass. However, additional nitrogen is not necessarily utilized by phytoplankton. Despite higher levels of hypernutrification, phytoplankton growth might be limited by other factors such as light availability in turbid or deep vertically mixed waters. Hydrographic conditions, such as the flushing time of a water body, can limit the accumulation of phytoplankton biomass, i.e., when algal cells are removed from the source of nitrogen before significant growth occurs. For further discussion on assessment and prediction of potential hypernutrification and eutrophication associated with coastal fish culture, the reader is referred to Gowen and Bradbury (1987); Gowen et al. (1989); Gowen et al. (1990); Gowen and Ezzi (1992); Wallin and Hakanson (1991); Hakanson and Wallin (1991); Aure and Stigebrandt (1990); Enell and Ackefors (1991); Turrel and Munro (1989).

5.2.5 Oxygen depletion

142. The likelihood of large-scale oxygen depletion will clearly depend on size and intensity of the aquaculture operation (i.e., the oxygen demand by both the cultured stock and the wastes released) and the topography/hydrography of the waterbody. An approximative assessment of the holding capacity of the waterbody (i.e., the threshold at which production becomes limited by a non-trophic resource see Rosenthal et al. 1988), can be obtained by comparing the oxygen demand of the stock to the pool of available oxygen and the rate of supply. Models to predict effects of aquafarm operations on the oxygen budget of a waterbody are, however, still being developed (see for example Aure and Stigebrandt, 1990).

5.2.6 Carrying capacity

143. Shore-based aquaculture practices such as the farming of seaweeds and bivalves interact with coastal food webs since they rely on naturally-available food resources. The carrying capacity of a defined area refers in ecology to the potential maximum production a species or population can maintain in relation to (naturally) available food resources within the area (Rosenthal et al., 1988). Hence, the production potential of, for instance, bivalves in a coastal waterbody is determined by the carrying capacity of this water body. Over-stocking would result in reduced production, since the carrying capacity is exceeded. Carrying capacity can be assessed by evaluating historical records of bivalve culture (see Figure 14), by measuring the availability of phytoplankton biomass or by undertaking more detailed studies, e.g., of carbon and nitrogen flows through a bivalve culture unit interacting with the food web (see for example Rodhouse et al., 1985). Methods for estimating the carrying capacity of areas used for oyster and mussel farming are reviewed by Héral (1991).

5.2.7 Visual observation and remote sensing of contamination and degradation

144. There are various possibilities to directly or indirectly detect and assess areal extent of contamination and degradation by visual observation and remote sensing.


Figure 14: Estimated biotic capacity of the Marennes-Oléron basin, France.

145. Direct observation of the sediments in the vicinity of aquafarms by divers will give first indications, e.g., on the degree of siltation, on the coverage or damage to benthic flora, on the lack of typical fauna, on the fall-off of shells, on the colour, thickness and consistency of the sediments. For example, a white mat on the sediment surface, formed by the sulphur-oxidizing bacteria Beggiatoa is often found around farms. Simple Secchi-disc measurements from boats may reveal degree and extent of turbidity increase in the water column around aquafarms.

146. On a larger scale, aerial photography and video filming may prove useful to record extent of water contamination and degradation of land surface. More sophisticated remote sensing techniques including infra-red and multi-spectral photography, multi-spectral scanning and radiometry, may, under certain circumstances, provide information on effluent discharge plumes, sedimentation patterns, phytoplankton concentrations and changes in coastal land use, all of which may be related to large-scale coastal aquaculture expansion and intensification (Meaden and Kapetsky, 1991; Butler et al., 1988; Kapetsky et al., 1987; FAO, 1989a, 1985b).

147. Large-scale spatial and temporal environmental changes in coastal areas may be assessed using geographical information systems (GIS). A GIS is a computerized approach to storing, manipulating, analysing and reporting data by reference to space (geo-referenced information), i.e., data which can be attributed to a location. GIS has proved useful (see previous citations) inter alia: in assessing impacts on aquatic resources and environments from development projects involving land and water use, in aquaculture site selection in relation to ecological and socio-economic variables, in space and resource allocation to conflicting types of use, in aquaculture development planning and environmental impact monitoring.

5.2.8 Surveillance and monitoring in coastal aquaculture environments

148. Most coastal aquaculture practices have no or little significant effects on the environment. It can also be expected that, in most cases, the environmental capacity of coastal water bodies is far from being exhausted by the waste load received from aquaculture operations. At present, it is only under particular circumstances, such as the combination of poor hydrographic conditions and unusually dense aggregation of farming units or highly intensive practices, that severe impacts on the aquatic environment may occur. These considerations are important when deciding whether or not to establish aquaculture-specific reporting, surveillance or monitoring schemes.

149. Nevertheless, existing and future coastal aquaculture activities will certainly benefit from aquatic pollution monitoring programmes assessing environmental changes in coastal areas and, possibly, river basins. Aquaculture-specific monitoring schemes should, where possible, be integrated in existing coastal water pollution assessment activities. Furthermore, the data derived from aquaculture-specific monitoring may also serve as an aid in improving husbandry practice which in itself can be a means of reducing possible ecological effects without lowering production.

150. General principles and requirements for monitoring and surveillance are outlined in section 5.1.2. Considering the variety of coastal aquaculture practices and the diversity of potential ecological effects involved, it is crucial for any aquaculture-specific monitoring activity that its objectives and purposes are well defined. Data collection for monitoring can lead to large and costly programmes. It is essential that monitoring be carefully planned and that the techniques adopted are statistically sound. Therefore, before starting a monitoring activity, a baseline assessment may be required to clearly establish:

- the correct key parameters to be recorded or monitored;
- the duration and the area to be covered;
- the possible achievements and limitations of the approach chosen;
- the skills and equipments required;
- the costs involved; and

- the source for financing the monitoring activity.

151. For monitoring to be an effective means of indicating that ecological change does not exceed a predetermined level it is necessary to identify and where possible quantify the change in a particular parameter. At present, however, there is considerable debate regarding which key parameters should be monitored to assess a given effect (Gowen et al., 1990).

152. The right choice of parameters will be determined by the circumstances encountered, e.g. type of coastal aquaculture practice, production level and the physical and ecological conditions prevailing in the area. A cursory overview of parameters and characteristics for which data may be collected is given in Annex 4.

153. It is important that suitable control stations are located. The frequency and timing of sampling have to be related to the nature of the parameters being monitored, and consideration should also be given to effects of natural variation and seasonality in a given parameter. Standardization of both methodology and description of parameters (in terms of units, dimensions, relationships, ratios, rates, etc.) is emphasized.

5.3 Role and Functions of Environmental Impact Assessment (EIA)

154. Environmental impact assessment (EIA) will be discussed briefly, as it is sometimes viewed as a management tool or as a regulatory mechanism or a policy instrument. There is no “standard” or “ideal” EIA. EIA is controversial and is still being discussed and improved in terms of scope, objectives and procedures. There are many definitions and concepts of EIA, including different terminologies (Jernelov and Marinov, 1990). Means of practical EIA implementation and features of incorporation of EIA in institutional frameworks of countries and international organizations are very diverse (ERL, 1990). This diversity in EIA is a positive feature, reflecting different circumstances and requirements wherever EIA is to be applied.

5.3.1 Purposes of EIA

155. According to UNEP (1988), EIA is a management tool like economic analysis and engineering feasibility studies. EIA (1) predicts the likely environmental impacts of projects, (2) finds ways to reduce unacceptable impacts and to shape the project so as that it suits the local environment, and (3) presents these predictions and options to decision-makers. Bisset (1989) lists the following objectives of EIA:

(a)

to identify beneficial and adverse environmental impacts;

(b)

to suggest mitigation actions which might reduce or prevent adverse impacts;

(c)

to suggest measures which might enhance beneficial impacts;

(d)

to identify and describe the residual adverse impacts which cannot be mitigated;

(e)

to identify appropriate monitoring strategies to “track” impacts and provide an “early warning” system;

(f)

to incorporate environmental information into the decision making process relating to development projects; and,

(g)

to aid selection of the “optimum” alternative (if alternative sites or project designs are being investigated in an EIA study).

156. Aspects of “modern” EIA include the understanding of EIA as a positive, improvement-oriented approach and as an iterative process (and not as a single study), whereby consultation and public participation are included and socio-economic and socio-cultural issues are covered (Driver and Bisset, 1989). An overview of socio-economic parameters which may need to be considered in coastal aquaculture development is given in Annex 5. Aspects of social monitoring and evaluation in aquaculture projects are reviewed by Molnar and Duncan (1989).

5.3.2 The EIA sequence

157. In general, the EIA process would contain three main parts: environmental appraisal, monitoring and evaluation. The environmental appraisal would consist of four key steps: screening, preliminary assessment, scoping and detailed assessment (see also Figure 15).

158. In a first step, projects are screened according to pre-determined criteria to decide whether or not further appraisal is required. Criteria on potential significant environmental impact may include type and size of projects, environmental sensitivity of areas, or combination of both. More flexible case-by-case approaches may also be followed.

159. The preliminary assessment step applies to projects which would require further environmental appraisal. Here, a proposed project would be appraised in terms of: likely environmental impact based on initial impact prediction, purpose, societal need, technological requirements and costs, location, level of resource utilization, people and communities potentially affected, alternatives (including technological improvement and adaptation of project), and possible mitigation measures. This step identifies projects with limited environmental implications that can be readily overcome. Provided appropriate adaptation and mitigation measures are identified and incorporated into the project design, no further assessment would be required.

160. Scoping applies to projects which are likely to have serious environmental consequences and which will need thorough examination and assessment. Here, the scope of a detailed EIA is determined in terms of objectives and major issues, appropriate methodology, data requirements, geographical boundaries, time horizon of analysis, costs involved, expertise required, affected groups, institutions, agencies, and possible alternatives.

161. Detailed assessment of proposed projects may require baseline studies on the current and past state of the environment. Assessment and prediction methods are then used to identify impacts and to predict their magnitude. A monitoring programme may be developed and ameliorative/mitigative measures may be formulated.

162. Monitoring provides an early warning that adverse effects (predicted or not) are occurring. Monitoring can be useful for continued generation of EIA inputs to management (e.g., mid-course corrections, effectiveness of and compliance with recommended mitigation measures, and improvement of predictions). All projects which may cause significant negative environmental effects should include a monitoring component. In cases where the environmental appraisal has confirmed that the effects will not be significant, and that adequate steps have been taken to minimize the identified effects, monitoring may consist of periodic reviews of the main areas of concern.

163. Through evaluation of completed projects the accuracy of predictions and the relevance of recommendations are judged against actual experience. Evaluation may help to identify additional significant effects warranting corrective actions. Evaluation results can prove useful to refine impact predictions for future projects of the same type and magnitude.

164. Figure 16 and Table 14 show when and how an EIA can contribute to the various stages of the project cycle. In order to show the variety in approaches and terminology, examples of other EIA sequences are given in Annex 6.

5.3.3 Methodologies in EIA

165. There are more than 100 techniques and formats for carrying out EIAs including check-lists, environmental evaluation, matrices, networks, environmental indices, cost-benefit analysis, overlay mapping, simulation modelling workshops, which are reviewed elsewhere (ESCAP, 1985; Shopley and Fuggle, 1984; Carpenter and Maragos, 1989). Often, the resources required to carry out an EIA include: (1) establishment of a qualified multi-disciplinary team and analytical facilities, (2) access to EIA techniques, (3) financial and institutional support, and (4) monitoring and enforcement powers.


Figure 15: The EIA process (from Driver and Bisset, 1989)


Figure 16: EIA and the project cycle (from Driver and Bisset, 1989)


Table 14: EIA in the project sequence.

Note: the exact correspondence varies among projects (from Carpenter and Maragos, 1989).

5.3.4 Problems and limitations in EIA

166. In many countries, an inadequate database and lack of trained personnel are likely to be major constraints. Implementation of EIA in many countries faces problems due to:

- poor availability and reliability of data;

- insufficient training/education in EIA methodologies and in the establishment of appropriate legal and regulatory frameworks and institutional arrangements;

- negligence of beneficial impacts in EIA reports;

- lack of consideration of alternative sites, technologies, designs, and strategies;

- insufficient involvement and participation of all interested and affected parties;

- insufficient emphasis on required cost-effectiveness of EIA;

- lateness in implementation and lack of follow-up monitoring and evaluation;

- inappropriate recommendations,

- e.g., mitigation/adaptation measures which are not affordable or feasible in terms of maintenance requirements or operating costs;

- poor presentation of EIA results.

167. Also, scientific knowledge about ecological cause-effect relationships and modelling of ecosystems is still developing which limits the capacity to quantitatively predict all environmental changes. Similarly, techniques for economic valuation of environmental impacts also need to be improved although much progress in identifying and including so-called offsite and future externalities has been made (see for example Dixon, 1989; Bergstrom, 1991; Dixon et al., 1986; Hufschmidt et al., 1983; Dixon and Hufschmidt, 1986; Klaassen and Opschoor, 1991; Pierce, 1988).

5.3.5 Alternatives to EIA

168. Although the implementation of EIA and its integration in regulatory and institutional frameworks probably contribute to the process of sustainable development, alternatives to EIA, which are more suited to the database in the countries concerned and to their levels of personnel and other resources should also be explored and emphasized. Of particular importance in this context are physical planning procedures and rules for land zoning and land use.

169. For example, land planning may have to intervene in the functioning of markets when the existence of externalities or public good result in market outcomes that are unfavourable to sustainable resource use. From the standpoint of legal instruments, this involves, inter alia, the resolution of competing demands for land in relation to the predetermined development goals, criteria and priorities (FAO/Netherlands, 1991b).

170. Generally, there is an increasing trend to combine approaches and methodologies to resolve environmental and socio-economic development issues, which is reflected, for example, in recent guidelines for land-use planning, rural-area development planning and economic-cum-environmental development planning (FAO, 1989b, 1991b; Bendavid-Val, 1990; ADB, 1991 a, 1991b, 1988). Recently, the United Nations Development Programme also produced a handbook and guidelines for environmental management and sustainable development (UNDP, 1991).