[From the U.S. Government Printing Office, www.gpo.gov]
Coasal Zone Information Center COASTAL ZONE INFORMATION CENTER JUL 20 1977 POLLUTANTS IN THE AQUATIC ENVIRONMENT: Detection, Measurement and Monitoring Lectures presented at the Second FAO/SIDA Training Course on Marine Pollution in Relation to Protection of Living Resources GC 1085 AND AGRICULTURE ORGANIZATION OF THE UNITED NATIONS .F6 1976 .F6 1976 SWEDISH FUNDS-IN-TRUST FAO/SIDA/TF 108 SUPP1. 1 TF-INT 86 (SWE) COASTAL ZONE INFORMATION CENTER LECTURES PRESENTED at the SECOND FAO/SIDA TRAINING COURSE ON MARINE POLLUTION IN RELATION To PROTECTION OF LIVING RESOURCES Methods for Detection, Measurement and Monitoring of Pollutants in the Aquatic Environment Gothenburg and Stockholm, Sweden 31 July -.9 September 1973 (SUPPLEMENT TO THE REPORT) US DEPARTMENT OF COMMERCE NOAA COASTAL SERVICES CENTER 2234 SOUTH HOBSON AVENUE ChARLESTON SC 29405-2413 FOOD An AGRICULTURE ORGANIZATION OF THE UNITED NATIONS Rome, 1976 Property of CSC Library Bibliographic reference FAO/SIDA Training Course on Karins Pollution in Relation to Protection of Living Resources. Gothenburg and Stockholm Sweden, 31 July - 9 September 1973 (1976). Rome, FAO/SIDA/TF 108 Suppl. 1:109p. Lectures presented at the Second FAO/SIDA Training Course on Marine Pollution in Relation to Protection of Living Resources. Methods for Detection, Measurement and Monitoring of Pollu- tants in the Aquatio Environment. Gothenburg and Stockholm, Sweden 31 July - 9 September 1973. (Supplement to the report) Training Centres, Narine Pollution, Oil Pollu- tiong Primary production, Water analysis (chemical), Pollution effects, Bioassays, Pollution monitoring. PREFACE Following the FAO Technical Conferenoeon Marine Pollution and its Effects on Living Resources and fishing (Rome, December 1970), collaboration between the Swedish International Development Authority (SIDA) and FAO was extended to aspects of training in the field of marine pollution and plans were developed for a series of Training Courses on Marine Pollution in Relation to Protection of. Living Resources.. So far, four courses of this series have been held. The first course (G othenburg,Swden,May-June 1972) wax arranged for senior scientists and research managers from developing countries in charge of planning investigations or developing of control or monitoring programs. The course provided a broad review on pollu tants known to be hazardous to marine life and fisheries, their fate and effects. Related research and monitoring problem and scientific and administrative elements of importance for management measures were reviewed. The second oourse hold in Gothenburg and Stockholm in July-September 1973, was arranged for research workers from developing countries (biologists, ohemists, technicians, etc.) and me oriented towards planning and conduct of research and monitoring activities giving training in techniques and equipment currently used for the- detection and measurement of pollutants in the aquatic environment. The third course (Lima, Peru,February-March 1975) wan arranged for participants from Latin American countries with the purpose of providing a scientific basis for the protection of living resources from pollution and for management of the quality of the aquatic environment. The fourth course held in Lysekil, Sweden, in October-November 1975 was arranged for scientists from developing countries providing training in, and demonstration of, bioassays and toxicity testing techniques currently used for biological monitoring and establishing of water quality criteria. This volume contains a selection of the lectures given at the second course. It was completed by the staff-of the PAO Fishery Resources and Enviromment Division, particularly Mr. A. Wenblad. Editorial assistance was given by Dr. R. Vaz, SNU Special Analytical Labo- ratory, Wallenberg laboratory, Stockholm, Sweden. The views expressed are those of the indi- vidual lecturers and are, not necessarily those of FAO or SIDA. Used on papers on analytical techniques for pollution measurements, presented during the second course,the Manual of Methods in Aquatic Environment Research, Partl Methods for Detection, Measurement and Monitoring of Water Pollution has been published (FAO Fish.Tech. Pan. 137). -V- CONTENTS page PRIMARY PRODUCTION AND NUTRIENTS IN SEA WATER 1 by S Fonselius REDOX MEASUREMENTS IN NATURAL WATERS AND SEDIMENTS 8 by L.E. Bagander INTERLABORATORY STUDY OF METHODS FOR CHEMICAL ANALYSIS OF WATER 16 by G. Ekedahl and B. Rondell OIL AND OIL DISPERSANTS 28 by S.R. Carlberg CASE STUDY CHEMICAL ANALYSES OF A SEA AREA POLLUTED BY MINERAL OIL 36 by S.R. Carlberg ECOLOGICAL EFFECTS OF MARINE POLLUTANTS 47 by A.D. McIntyre INVESTIGATION OF ACUTE POLLUTION PROBLEMS AFFECTING FISHERIES IN ESTUARIES AND 56 COASTAL WATERS by J.S. Alabaster TOXICITY TESTING AT KRISTINEBERG ZOOLOGICAL STATION 65 by M. Swedmark, Granmo and S. Kollberg TOXICITY TESTING IN A CONTINUOUS FLOW SYSTEM 75 by I. Berg and and A. Granmo BIOASSAY METHODS USED BY THE RESEARCH LABORATORY AT THE SWEDISH NATIONAL 81 ENVIR0MENT PROTECTION BOARD by T. Hasselrot MONITORING OF AQUATIC POLLUTION 92 by Go Tomczak PRIVIART PRODUCTION AND NUTRIENTS IN SEA WATER by S, FONSELIUS Royal Fishex-y Board Gothenburg,, Sweden Contents Ltle 1. Introduction 2. Utilization of nutrients for primary production 1 3. Some factors limiting primary production 2 3.1 Nitrogen and. phosphorus 2 3.2 Other constituents 4 4- Potential fertility of sea vater 4 5. Anaerobic decomposition of organic matter 5 6. References 6 1. INTRODUCTION The,bio.logioal activity inosea and lake water is regulated by light, transparency of the. water, temperature, and availability of nutrients, vitamins, hormones, trace elements and other growth-rogmlating factors. The availability of nutrient salts often limits the growth or organic production. Nutrient salts are necessary for the organisms and are taken up by them. When the orgamsms -die, they are decomposed through bacterial breakdown or oxidation. The nutrients are thus .brought back into the water in inorganic form-. Because the organisms sink when they the the nutrient salts.are removed from the surface water and accumulated in the deep water. They are also taken down in convergence areas or through the sinking of cold water during winter. .In upwolling areas nutrients are brought back to the surface. In some areas the winter convection may also bring up nutrient-rich water to-the surface. 2* UTILIZATION OF NUTRI1MTS FOR PRnUff PRODUCTION According to Liebig's law of minimi=j the growth and yield of a crop is dependent on that nutrient which is present in a minim=. The nutrients.whiob. W limit the production in water are phosphorus and nitrogen. Both may reach concentrations close to zero in natural waters. In order to be available for the phytoplankton they have to be in inorganic or very simple organic form, the phosphorus as-phoophate and the nitrogen as nitrate, nitrite, ammonia, urea or some amino acids. In fresh water the availability of phosphorus is generally the nutrient factor that limits production. In the oceans the nitrogen supply is considered to be most limiting. But after a high phosphorus increase, nitrogen and not phosphorus may act as the limiting factor in inland and brackish waters. Inorganic carbon may act as the limiting factor for the phytoplankton production in extremely productive and alkaline waters. Many metals (e.g. iron, manganese, cobalt, molybdenumt zino and o"ometimes.also magnesium and chelating substances have such low concentrations (especially in nutrient-poor mtersi that they limit the growth of algae. Elements are removed from the water phase by formation of organic matter through photo- synthetic assimilati,on. These elements are utilized in constant proportions for forming phytoplainkton, and also to a smaller degree for higher levels of the food chain. The 2 Carbon, alementary composition of living organic matter is for every species quite constant. hydrogen and oxygen form the main part of organic-matter, roughly in the proportion CH 20 (12:2:16). There are also other elements, important for the organism, present in orgafiic matter in constant proportions. Ana),yeis of large plankton samples which represent the main biomass in the open sea have given quite constant proportions for carbon, nitrogen and phos- phorus. Table I Atomic Ratios of the Principal Elements in Plankton (Fleming, 1940) C I P Zooplankton 103 16-5 1 Pbytoplankton 108 15-5 1 Average 106 16 1 The free oxygen which is prodneed in the synthesis or is utilized in the decomposition of the plankton should be 212 atoms for each phosphorus atom if only carbon is oxidized. We assume according to Richards (1965) that the biomass has the following general composition and that it decomposes as shown below: (CH20)106(NH3)16 H3PO4 - 106 CH 20 + 16 NH 3 + H 3PO4 106 CH2o + 1o6 02 - 1o6 C02 + io6 H20 16 NH 3 + 32 02 ' 16 MM 3 + 16 H20 '@'(CH20)106(KH3)16H3PO4 + 138 02 - 1o6 C02 + 122 H20 + 16 HNo 3 + H 3PO4 As can be seeng four atoms of oxygen are utilized in oxidizing each atom of nitrogen and the total oxygen consumption is 138 molecules. The O:P ratio will then be 276:1. These ratios make it possible to estimate the average proportions in which the concen- tration of nutrients in sea water may be expected to change due to biological activity. The changes observed in the composition of the nutrient concentrations seem to confirm these theoretical considerations. The composition of plankton has here been assumed to be constant, which of course is not always true. The composition of the different species in the biomass is variable and dependent on time and place. Each species may be expected to have a somewhat different chemical composition. 3- SCME FACTORS LIXITING PRDEARr PRODUCTION 3.1 Nitrogen and ]2hosphorus It has been shown through culture experiments that the chemical composition of unicellular algae changes when the composition of the medium in which they grow in altered. If the conoen- tration of one element is very low in the medium, relative to the need of the organism, cell growth and division can continue for some time. The cells which are produced contain smaller amounts of the deficient elements than the normal cells. When there in an excess of an element in the medium, the cells can increase their content of this element (lwmry consumption). 3 When plankton growth in the sea water increases, available nitrogen and phosphorus dis- solved in the water are greatly reduced. In Long Island Boundt according to Riley and Conover (1956), nitrogen seems to be the element available in minimum proportions relative to the needs of the phytoplankton. Its ratio to phosphorus varied there from almost zero to 8:1. The N.,P ratio in the plankton was found by Harris and Riley (1956) to be on an average 16-7:1- In spite of the abnormal ratio of N:P in the water, the plants seem to assimilate the elements in the normal ratio of 16:1 until the nutrient concentrations reach extremely low values. When the nitrogen is nearly exhausted from the water phase the phytoplankton seem to be able to continue to utilize phosphorus, which is present in surplus proportions. Most of the common algae will obtain comparable growth rates with nitrate, nitrite or Wmonia. When all three nitrogen sources are present simultaneously in the water, many forms seem to prefer ammonia. According to Vaccaro (1965) this may depend on a toxic effect of the enzymatic reduction of nitrate. It has been shown that starved Chlorella cells given equiva- lent amounts of nitrogen recovered faster with ammonia than with nitrate. It has also to be pointed out that ammonia is often the only form of nitrogen available in the surface water in measurable amountB during high productive seasonso It is difficult to explain how phytoplankton can continue to form cells of normal chemical composition when they grow in sea water from which an essential nutrient may be almost absent. According to Redfield et al. (1963), this could be explained in the following my. Mixing processes may transpoxT_zZ_rients from'deeper'layers up to the euphotio surface water, e.gs eddy diffusion may deliver phosphorus and nitrogen to the surface layer in a higher ratio than normally occurs in this layer. When dead plankton cells decompose, phosphorus and nitrogen are also regenerated in a different ratio than their composition indicates. There is also the possibility that phosphorus for examplej is regenerated faster and that nitrogen is retained in the marine humus, which may be quite stable. Accordingto Redfield et al. (1963) the nut- rients present in the water represent only the residue of elementst wlilcb7are not required for the algal cell composition. There does not seem to be any proof that natural algal populations may form cells with an abnormal composition. In laboratory experiments with algal cultures it has, however, been possible to"grow cells with nutrient deficiencies. Redfield et 11. (1963) have also established a norm for surface water regarding nutrients* .Thus deviatior;_f7rom the norm can be exp'ressed. As the norm for phosphorus a value based on the conditions in the deeper water of the Antarctic Ocean me used. They derived a nitrogen norm from this value using the ratio 1521- Por the carbon concentration the value given by Sverdrup Lt &1L (1942) was uled. The oxygen value was taken as the oxygen saturation value for normal ocean water at +2 C. Table 11 shows the ratios in which these elements are available in "normal" sea water and those in %hich they are,used for the formation of organic matter. The ratio of availability to utilization is also given. It can be seen that nitrogen and phosphorus ooo=Ting in sea water have almost exactly the proportions in which they are utilized. This was first observed by Harvey (1926) in the English Channel. During the growing season both nutrients disappeared from the surface water. We can also see that carbon is present in a large excess (ten times the quantity which can be utilized) if phosphorus and nitrogen limit the production. Therefore carbon cannot be a limiting factor in normal sea water., There will alwaym be enough carbon for building up organic matter. Often there are small residues of nitrogen or phosphorus left in the waterg depending on the nutrient that has been completely utilized. It is of course clear that local variations may occur in the surface water resulting in differences in the composition of the water. -4- Table II Availability of ?Arient Elements in "Average" Sea Water (S - 34.7f. T = 2 C) and the Ratios of their Availability and Utilization by Plankton (Redfield et al. 1963) Availability in Utilization Ratio "average" sea by Of water Plankton availability pg-at/l, ratio ratio to utilization Phosphorus 2.3 1 1 1 Nitrogen 34.5 15 16 0.94 Carbon 2 340 1 017 106 9.6 Oxygen saturation 735 320 276 1.16 value 3.2 Other constituents In nat ure, there are often several factors which may influence growth rate. In sea water we have for example, trace metals such as iron, manganese, copper, zinc, molybdenum and cobalt present in such small amounts that they may limit production. Menzel and Ryther (1961). have shown that addition of iron to water where the natural plankton growth has come to an end due to lack of nutrients can stimulate growth. They used surface water from the Sargasso Sea. They did not get the same effect with other trace metals. Harvey (1947) showed that manganese had the same effect for Chlamydomonas in water from the English Channel. Hormones, vit-hin and other organic compounds may have similar effects. The -in limiting factors for phyto- plankton production in the sea seem, howevert to be nitrogen and phosphorus. Other factors may locally be limiting, especially in coastal areas, but are not of general importance. 4- POTMMAL FERTILITr OF SEA WATER Harvey (1947) has proposed that the concentration of total phosphorus or total nitrogen in the water may be used to determine its potential fertility. The potential fertility is defined as the quantity of organic matter which can be produced by photosynthesis in a unit volume of sea water if it is illuminated until the limiting nutrients are exhausted. Redfield et al. (1963) applied this on the values in Table II. They assume that the carbon of organic Mft-er is 50 percent of the dry weight. When all nitrogen has been exhaustedt organic matter with a dry weight of 5.48 mg/1 and containing 2.74 mg C/1 has bee 'n formed* If we assume the dry weight/wet veight ratio is 0.2, the formed plankton would be 28 mg11 or 28 g/ton. It should be stressed thwk this is the maximum amount that could be produced in the photic zone if the conditions are ideal. In reality such amounts are never formed. Fjrther (1966) estimated that the plankton concentration in ocean water is normally onlY 10 000 000 expressed as wet weight. He assumed that all plankton2 is formed in*the V4*w 100 metres of the water and that there are 3 g plankton carbon per m of sea surface. From these estimates we can see how low the real fertility of ocean water is compared to the potential fertility, There is only 0.3 g of living plankton matter in I *on of ocean water. 5 5. ANAEROBIC DIDCCNPOSITION OF ORGANIC MATTER From Tablell we can see that the oxygen reserve of the ocean water is not very large. When there is considerable accumulation of organic matter in an area the dissolved oxygen in the water may be completely used up. This will lead to anoxic conditions in the water. The oxidatiou of organic matter may however still'oontinue through anaerobic.bacterial processes. These processes begin with denitrification where nitrate is transformed into nitrite and then into nitrogen gas or ammonia. When this process is completed, sulphate ions may serve as oxygen donors being transferred into hydrogen sulphide.- Finally carbon dioxide may be trans- ferred into methane losing'its oxygen to the oxidation process. The reduced products (nit- rogen gas, ammonial hydrogen sulphide and methane) accumulate in the water together with the oxidationproducts of the organic matter-(@phosphatej ammonia and carbon dioxide). This accumulation of ions and gases will change the normal proportions of the.component.s in sea water, giving it different characteristics.. Redfield et al. (1963) suggest the following reactions.for oxidation'of organic matter in the presence and absence of dissolved oxygen with CH 2 representing organic matter. Oxidation by.oxygen (dissolved)' CH20 + 0 2 Co 2.+ H20 NH + 20 HNO + H20 3 2. 3 Denitrification 5CH20 + 090 3 5CO2 + 2X2 + 7H20 5NH3 + 3HIFO 3 4N2 + 9H 20 Sulphate reduction 2CH20 + H2SO 4 20 02+ H2.S+ 2H20 NH3 is not oxidized It is also possible to assume that phosphate may be reduced to phosphine. Phosphate reduction 2CE20 + H3P04 2CO2 + 2H@O + PH3 According to Soviet scientists phosphine is found in the bottom water of the Sea of Azov. ,.Finally we have the possibility of carbon dioxide reduction in which methane may be formed. Carbon dioxide reduction 2CH20 + H2CO3 2CO2 + H20 + CH 4 Methane has been found in the deep water of the Black Sea and in some anoxic fjords in -This formation seems to begin before all sulphate is used up. Methane is generally formed in the sediments of stagnant oxygen-free lakes. It is also formed in the bottom mud of shallow salt water lakes or lagoons if the water.has a heavy organic load. The free energy determining the different oxidation steps decreas6s in the order; oxygen, (dissolved) nitrate, sulphate, carbonate, when these are the hydrogen acceptors. Therefore the free oxygen is utilized first when it is available in the water. Then the nitrate is -6- utilized. When the nitrate was been exhausted the oxidation continues with reduction of sulphate. Apparently there is not sufficient free energy left to oxidize the ammonia to nit- rogen gas in this step. Therefore ammonia is accumulated during the sulphate reduction. As an intermediate step in the nitrate reductiong nitrite is formed and it is also formed in the oxidation of ammonia to nitrate. Aooording to Codispoti (1973) denitrifioation begins when the oxygen concentration is below 0.1 ml/i. This has also been shown for lake water. Nitrite is found in relatively high concentrations in the oxygen minimum layer off the Mexican coast in the Pacific and also in anoxic basins in the border layer between oxygen and hydrogen sulphide. The end product of the denitrification process is, according to Codispoti (1973) most probably free nitrogen gas. The deep water of the Black Sea is supersaturated with free gaseous nitrogen. Richards and Benson (1961) have examined the accumulation of free nitrogen in the anoxic waters of the Dramsfjord and the Cariaco Trench. They compared the observed nitrogen/argon ratios with the ratio in sea water equilibrated with air. The excess quantities of nitrogen found in the anoxic waters corresponded approximately with the amounts estimated to have been produced by the oxidation of organic matter.' The changes in the oxygen, sulphide and phosphorus content of the water also supported these results. They-found that nitrate and nitrite were present only in trace amounts or were absent from the anoxic water. Ammonia, on the other hand, was present and varied in proportion to the accumulation of sulphide. This shows that sulphate probably does not act as the hydrogen acceptor in oxidizing the ammonia which is liberated from organic matter during the reduction of sulphate. The most striking change in the ionic composition of anoxia sea water is the occurrence of sulphide ions. Hydrogen sulphide is a poisonous gas which kills all higher life, and only certain bacteria can live in such an environment. It has been established (Skopintsev et &I., 1958) that the principal source of the sulphide sulphur in sea water is the reduction of sul- phate and not the sulphur from decomposed organic matter. 6. REFERENCES Codispoti, L., Some chemical and physical properties of the eastern tropical North Pacific 1973 with emphasis on the oxygen minimum layer. Seattle University of Washington, Department of Oceanography. Ref.M73-64. Oct1973 Fleming,R.H., The composition of plankton and units for reporting population and production. 1940 Proc.Pac.Sci.Cong., 6(3) Harris, E. and G.A. Riley, Oceanography of Long Island Sound 1952-1954. .8. Chemical oompo- 1956 sition of the plankton. Bull.Bingham Oceanogr.Collect., 15:315-23 Harvey, H.W., Nitrate in the sea. J.Mar.Biol.Assoc.U.K., 14:71-88 1926- Manganese and the growth of phtoplankton. J. Mar.Biol.Assoo. U.K.,26(4):562-79 1947 Menzel, DW and JH. Ryther Nutrents limiting the production of phytoplankton in the 1961 Sargasso Sea with special reference to irons Deep-Sea Res., 7(4): 276-81 Redfield, A.C.,B.H. Ketchum and P.A. Richards, The influence of organisms on the composition 1963 of sea water. In The sea Vol.2. Compotition-of sea water, edited by N.N. Hill New York, Intersoienoe Publishers, Richards, F.A. Anoxic basins and fjords. In Chemical oceanography, edited by J.P. Riley and 1965 Go Skirrow, London, Academic Press, vol.1, pp. 611-45 7 Richards, FA and B.B., Bensonj Nitrogen/argon and nitrogen isotope ratios in t wo anaerobic 1961 environments, the Cariaco Trench in the Caribbean Sea and Dramsfjord, Norway. Deep-Sea, Res., 7(4)254-64 Riley, GA and A. Conovert Oceanography of Long Island Sound 1952-1954. 3. Chemical oceano- 1956 graphy. Bull.Bingham Oceanogr.Collect., 15:47-61 Ryther, JH Organic production by planktonic algae and its environmental control. In The 1966 Pymatunio Symposium in Ecology. Spec.Publ.Pymatunio Lab.Field Biol.Univ.Pittsburg (2) Skopintsev, B.A. et al., Content of the main components of the salt-water of the Black Sea 1958 and the problem of water exchange. Tr.Akad.Nauk.SSSR.Morsk Gidrofiz.Inst., 13 (in Russian) Sverdrup, H.U., M.W. Johnson and R.H. Fleming, The oceans. London, Prentice Hall 1942 Vaocaro, R., Inorganic nitrogen in sea water. In Chemical oceanography, edited by J.P. Riley 1965 and Skirrow. London, Academic Press, vol.1, pp. 365-408 8 REDOX MEASUREMENTS IN NATURAL WATERS AND SEDIMENTS by L.E. BAGANDER Department of Geology University of Stockholm. contents Page 1. Redox potential - theory 8 2. Redox measurements 9 2.1 Equipment 9 2.2 Calibration 11 2.3 Measurements 11 3. References 14 1 REDOX POTENTIAL- THEORY A chemical process in which electrons are lost is called an oxidation. The opposite, a process in which electrons are taken up, is called a reduction. In a ch mical reaction an oxidation process is always coupled to a process in the opposite direction, a reductio and vice versa. Fe 2+ Fe 3+ + a (1) The reaction above is called a redox reaction. Fe (3) is the oxidized form and Fe(2) in the reduced form. When an Fe(3) ion is reduced for every Fe(2) ion that is oxidized the reaction is at equilibrium. The redox couple then develops a very characteristic redox potential (E). The voltage scale is relative and the reaction H2 2H + 2e- (0.000 V) (2) has been chosen as the zero point in this scale. The redox potential is therefore referred to as Eh. The reaction (1) has according to this scale a potential of +0.771 V. The Potential is given according to the international convention. It should be noted that in American books, especially old ones, the potentials are given with the saw nunerical values, but opposite signs. The redox potential is measured by using an inert metal electrode (Pt or Au) in combina- tion with a reference electrode (oalomel or silver/silver chloride). When immersed in a solu- tion and joined to a circuit they form an electrochemical cell. The KMF or cell potential then be measured. Each of the electrodes is a half-cell for which the potential is given by the Nernst equations E = E+ RT In ox (3) nF red E constant for the half-cell R gas constant (8.3143 Joules/deg. and mole) F Faraday's constant (96.487 coulombs/equivalents) n number of electrons involved, according to the formula T absolute temperature ox/red activity ratio of the oxidizing and reducing species. 9 The potential of an electrochemical cell can be written: E cell E right-Eleft (4) In this case when using a platinum electrode and a reference electrode we can write: cell "ref What is. really measured is the.difference in potential between the two electrodes. The redox potential is an intensity factor like pH and temperature and gives no informa- tion on either the oxidizing or reducing capacity The potential at the inert platinum- electrode is created by an.excess or lose of electrons At the electrode surface. A reducing solution has a tendency of giving away electrons and an oxidizing solution has a tendency of taking up electrons. The reference electrode is constructed to give a constant potential. A calomel electrode consists of mercury in contact with mercurous chloride (oalomel) which, when immersed in a solution (KCI) with a -constant chloride activity, maintains a stable-potential. A glass body surrounds the electrode and contact with the outer solution is provided through a liquid junction. The electrode reaction may be written: Hg2 Cl2(s)+2e 2Hg(1) + 2Cl- (6) When changes occur at the electrode due to oxidation or reduction, the electrode potential is kept constant by the dissolution or precipitation of Hg2 C12 The silver/silver chloride electrode works according to the same 'principles. 2. REDOX MEASUREMENTS in natural waters and sediments the redox potential is created by a great number of reactions, none of which are at equilibrium and they are seldom reversible in the strict chemical sense. This is mainly due to biological activity. These mixed potentials-are of limited value for quantitative chemical calculation, (Stumm,1967). However, the redox potential gives a lot of information on the kind of reac- tions that take place, both from the chemical and biological point of view. Redox measure- ments, are therefore of great value,, being indications useful for ecology and pollution research. Determination of the level of the redoxolins (gradient in the redox- state) in a water body or sediment is of-ton more important than measurement of the absolute value of-the redox potential. Fig. I shows Eh profiles from two sediment. cores. In core RR1 a distinct redoxcline is situated at the 3.5 cm level. In core R4 the redoxcline is situated in the water above and the sediment is reduced from the top layer. 2.1 Equipment A platnum electrode, oalomel reference electrode and a potentiometrio instrument, usually a pH-meter with a millivolt scale, are used for measurements. Commercial platinum electrodes are expensive compared to easily constructed home-made electrodes. A platinum wire (eq 25 mm long and 0.5 mm in diameter) is fused into the end of a sodium glass tube or cemented into the end of a plastic tube.The electrode area is adjusted to a suitable size by cutting the wire.or enlarging it:with a platinum foil. It is thus very 10 @ediment depth CM. 0.- x x RR 1 R 4 x 2-- x x 3-- 4-- x xx 5 x x 6-- x x 7- 91 + + q +500 +400 -t300 +200 +100 0 -100 -200 mV Fig. 1 Eh profiles from two sediment cores easy to design a platinum electrode for a special purpose. As a rulev for greatest accuracy in measurements, the electrode area should be kept as large as possible within the practical limits for each type of measurement. Commercial oalomel reference electrodes are available with either saturated or 3-5 N KC1 solution as the inner solution. Standard potentials for the oalomel electrode are given in Table 1. The 3-5 N K01 solution is recommended for field work. The saturated XCI solution may cause erratic potentials if KC1 is precipitated on the oalomel or at the liquid junction due to low temperatures. Table I Standard potentials of oalomel electrodes (volt) relative to standard hydrogen electrode (Whitfieldt 1971 a) Temperature OC 3-5 M KCl Saturated KC1 10 0.2556 0.2543 15 0.2538* 0.2511 20 0.2520 0.2479 25 0.2501 042444 Calculated by the author 2,2, Calibration The ZoBell solution (0-003 M Potassium ferrocyanidev 0.003 X portassium ferricyanide in 0.1 N K01) is most frequently used for calibrations when meaovrin-z X . The potential of this solution is 4430 mV at 25 OC (ZoBell 1946). thfortunately this aol%Aon with a single well poised redox couple cannot detect if the working area of the redox electrode is-reduced due to adherent contaminants, The calibration procedure is therefore more concerned with'the function of the reference electrode and the instrument. 2,3 Measurements The platinum electrode our face must be kept as clean as possible. Mechanical cleaning using a mild abrasive or a tissue followed by rinsing with distilled water is a method which Is practicable in field work. Chemical cleaning must be used with care biBoause the electrode surface may be damaged. The liquid Junction of the reference electrode may be affected by suspended particles or precipitateag which may change the liquid Junction potential. The reference'electrode should therefore not be inserted into solutions with suspended particles or in a sediment. The uBe of a salt bridge between the sample and the.reference electrode prevents such errors. Samples of natural waters and sediments are.often poorly buffered with respect to redox potential. It therefore often takes some time to get a stable reading Electrode drift in four sediment samples at different redoxotateo are shown in F%g. 2. Intermediate redox values' around the redoxoline (BO 3) are stabilized more'slowly due to the unstable redox situation. Fig. 3 show electrode drift of seven in situ measurements made in a Dutch tidal flat sediment. After 30-minates the electrode was withdrawn 5-10 mm. Stabilization will thereafter be attained at lower values. Stabilization usually takes place within five to ten minutes and an electrode drift of less than 2 mv/Min can be accepted an a good reading. When introduced into a sample the redox electrode always causes some degree of disturbance. Sediment samples are most sensitive because part of the sediment structure is broken and the 12 Mv SR 5 +500-- +400- +300- +200-- 80.3 +100-- 0 R 4 -100-- -200-- F 80 0 5 10 15 TIME OAIN) Fig. 2 Electrode. drift in four sediment samples at different redoxetaten NA 93 0 0 V+ @-j 0 14 interstitial water is pressed away. The shape and size of the electrode must be designed to minimize these effects. Electrode drift due to introduction of air or gas loss when the electrode penetrates the sediment must also be considered. Absolute stabilization never occurs because of the continuous biological activity. In situ Eh measurements are most accurate. because the natural environment is more or less changed by any sampling technique* In situ probes for Eh and PH measurements have been constructed and used by Mortimer (1@4-1)-,Manheim (1961) and Whitfield (1971). In situ probes.are used for profiling in the water colunin and the few uppermost centi- metres of the sediment. If Eh is measured in samples, the measurements should be performed as soon as possible because long storage of the sample may change the redox potential due to oxidation processes or biological activity to values not representative for the respective natural environment. The redox potential is dependent on PH, which controls ionic equilibria, solubilities and the formation of complex ions, etc. The Eh/pH relationship is usually very complicated and a factor to correct Eh readings of the natural samples for a certain PH cannot be given. An average change of about -0.061 V for.each unit increase in PH, in the PH range of 3 to 10, has been reported by ZoBell (1946). However, redox measurements are not usually corrected for PH but reported together with- the PH value. Theoreticallyv the effect of temperature is expressed by equation (3).- But in the complex natural systems this effect is as complicated to calculate as that of PH. Temperature affects solubilities, reaction rates, PH, etc. It has been shown in four different marine mud samples that a change in temperature from 00 to 200C caused a change in Eh by between 0.01 to 0.02 Volts (ZoBell, 1946) The redox capacity of a system is responsible for the reproducibility of the Eh measurements. Natural systems often have a low redox capacity. Therefore Eh measurements can seldom be reproduced to an accuracy greater than :LO.01 Volts. A sediment, though appearing homogeneous, contains microenvironments where the redox potential is affected by microbiological activity* Therefore, parallel readings may differ by as much as �0.05 Volts or even more. 3o REFERENCES Hallberg, R.O.0 Some factors of significance in the formation of sedimentary metal 1968 sulphides. Stockh.Contrib.Geol., 15(4)339-66 Manheim, F.T., In situ measurements of PH and Eh in natural waters and sediments. Stockh. 1961 Cont7ri-b@@.ol., 8:27-36 Mortimer, C.H., The exchange of dissolved substances between mud and water in lakes. 1941 J-Ecol-, 30:280-329 Stumm, W., Redox potential as an environmental parameter; conceptual significance and 1967 operational limitation. Ady.Water Pollut.Res., 3(1):283-308 Whitfield, M., A compact potentiometric sensor of novel design. In situ determinations 1971 of PH, pS2- and Eh. LimnoloOc6ahogr., 16:829-37 15 - Whitfield, M. Ion selective eleotrodes for the analysis of natural" waters. 1971 a Sci.Assoc.)Handb.,(2) ZoBell CE Studies on redox potential of marine sediments. Bull.Am.Assoc.Petrol.Geol,, 1946 30:477-513 16 INIMLABORATORY STUDY OF METHODS PCR CHMUCAL ANALYSIS OF WATERV by G. EMAHL and B. RftELL National Swedish Environment Protection Board Research Laborator7l Drottningholml Sweden Contents EWe 1. Introduction 16 2. Preparation of samples and reporting of results 17 3. Treatment of data and interpretation of results 17 4, Criteria for acceptability of results 18 5. Discussion 18 6. References 26 1. INTRODUCTION In accordance with the Environment Protection Act of 196% standardization of analytical methods and analytical quality control have become subjects of prime importance in Sweden. Thelcentral supervising authority for water pollution control is the Environment Protection Board. A Nordic standardization programme started in 19709 and the aim is to reach the same national standard methods in all the Nordic countries (Funder-Sohmidt et al.t 1972). Interealibration trials or interlaboratory studies have been performed by the Nordforsk (Scandinavian Council for Applied Research) Working Group on Water Analysis and also in the IHD (International ELydrologioal Decade) programme (Kar.1gren and Ekedahlt 1971; Henrikeeng 1971; Ahlo 1970). But t;hese interoalibrations have been restricted to relatively few laboratories. In 1971 a national Swedish intercalibiation programme started with about 60 participating laboratories. This programme has two major goals, the evaluation of analytical methods and the control of work quality in meter laboratories. This involves the checking of newly proposed methods, and oomparing the results of already existing methods in different laboratories in order to ascertain their levels of aocura0yo In the first interealibration, June 19719 analyses for specific oonduotancet calcium, magnesium, sodium, potassiumt alkalinityp chloridet sulphate and hardinese were requested. The same intercalibration was repeated during Hawember-Deoember 1971- This report summarizes and evaluates the results obtained by the participating laboratories in these two inter- calibrations, A more complete Swedish version of the report is,available from the Research Laboratory,, National Swedish Environment Protection Board, The interoalibration programme has continued during 1972 with interoalibrations of methods for nutrient analyses and methods for determination of metals (Fe, Mnr All Cu and Zn). Experiences from the Nordforsk programmey for example, have shown that duplicate analyses of a single sample in one laboratory give essentially identical results. Other. experiences show that the between-laboratory error is the determining factor in evaluating a method (Youdeng 1969). For these reasons the interoalibratioins were planned according to "the two-ample technique" of Youden (Youdent 1959, Greenberg et al., 1969). Two different but similar samples are prepared and analysed once only.. The te`c1:n1q6 of Youden 0 possible a more effective and even more understandable statistical evaluation. This tech- nique concentrates on increasing the number of participating laboratories and reducing the number of determinations per laboratory. i/ Reprinted from Vatten 4/73 with kind permission from Irdreningen f8r Vattenhygien 17 2. PREPARATION OF SAMPLES AND REPORTING OF RESULTS Homogeneous and stable water samples are indispensable in intercalibration trials. The samples were taken from natural sources, In the first intercalibration the two samples were waters from lake Mclaren. In the second interoalibratioin,sample A was a natural water and ample B the same water with known additions of the constituents to be analysed. The sample waters were left to equilibrate in plastic containers for about a week then transferred to one-litre plastic bottles and distributed to the participating laboratories as soon as possible. No preservative reagents were added. Laboratories were requested to carry out analyses for calcium, magnesium hardness, sodium, potassium alkalinity chloride and sulphate, and to report She results in meg/1 with two decimal places. Specific conductance was to be measured at 20 C and reported as uS cm -1(equal to umho cm -1) with three figures. Analytical methods were neither suggested nor specified, The laboratories were asked to treat the interoalibration samples as ordinary routine samples A report form accompanied the samples and details of methods used were requested. In the first interoalibration June 1971, 48 laboratories reported results and 22 made all eight requested determinations. In the second interoalibration December 19719 the corresponding numbers were 55 and 24. Nost'participating laboratories were selected from water laboratories approved by the National Swedish Board of Health and Welfare. 3, TREATMENT OF DATA AND INTERPRETATION OF RESULTS Each parameter.is reported in a two-sample chart's The procedure-of Youden was used to evaluate the results statistically, with some modification* For each costituent the concen- tration reported for sample A is the abscissa and that reported for sample B is the ordinate. The pair of results reported for one parameter from one laboratory are used to plot a point, indicated by the coded identifying number of that laboratory. This method can be used to interpret the results with the identity ofthe laboratories remaining anonymous. The mean value (X) and standard deviation (a) were calculated for each sample and con- stituent. Then, all values outside x+-4s Us were rejected (outliers) and the final mean value standard deviation Were re-calculated. The two mean-lines are drawn perpendicular to the corresponding axis, dividing the two-sample chart into four quadrants. The intersection point of the two mean-lines is considered to be the "true value", A line is drawn at an angle of 45 through this point giving the mean difference between the two samples. The same scale has to be used for both "is, Results lying close to the intersection point represent a high degree of accuracy. The figures clearly show that the points tend to concentrate.in the + + or upper right quadrant or in the - - or lower left quadrant, This tells.us that the laboratories tend to get high results for both samples or low results for both samples. The points often fall on a line running from the lower left to upper right (the 45 0 line) and if there are many points the -pattern is that of an ellipse along this line, These patterns demonstrate the dominant role played by-the systematic errors, The systematic errors often indicate poor instrument cali- bration inaccurate standard solutions or improper working technique. The magnitude of the systematic error can be estimated from the ohart Results in the upper left or lower right quadrants often indicate random errors, Such errors are often due to miscalculations, errors in recording, typing diluting the samples or some other simple blunder, The overall precision of the results can also be determined, For each pair of results the difference (DO between the, two results (sample:A and sample B) is calculated, The estimate of the precision can be given by-. S EDi2 - (EDi)2/n r - 2(n-1) 18 For each laboratory and constituent analysed the systematic error is the same in anal@y- sing sample A as sample B. When the difference is taken the systematic error is eliminated. The Di- quantities are free from systematic errors and contain only random errors. 4. CRITERIA FOR ACCEPTABILITY OF RESULTS The estimatel Sr, of the overall preoision can be used to judge the acceptability of any laboratory's results, The following statistical criteria have been arbitrarily estab- lished by Greenberg et al, (1969) and Youden (195@), Points falling between the mean value and � 1,552 @@_ar:e_ acceptable, those between - 1,552 Sr and � 2.448 Sr are questionable and results falling outside these boundaries are unacceptable. These limits can be visua- lized in the figures by the construction of concentric circles, centred in the intersection point of the mean-lines, the '#accepted true value", The two circles are expected to include 70 percent and 95 percent of the results respectively, if systematic errors could be elimi- nated, It must be stressed that these relative acceptability criteria are arbitrarily chosen and founded on pure statistical basis, These criteria can be highly deviating for very similar constituentel compare, for example, values for chloride and oulpbate in Figs, 7 and8. A determination with high 'precision results in narrow circles. If there are normal systematic errors, a rather large number of results will be classified as =acceptable (see Fig. 7), On the other handt determination with bad precision leads to a greater spread of results and to wider circles. Strangely enough, this means a greater number of acceptable ,results (see Fig. 8), For this reason, the limits of acceptability in many cases must be more specifically definedl e.g.9 in prescribed values given in meq/1 or mg/l. A summary of the results obtained in the two intercalibrations is shown in the Tables I and II. Figs, 1-9 show the two-sample ohartal one for each constituent, obtained in the. second interoalibration. 5- DISCUSSION Data on analytical methods used by the laboratories are summarized in Tables I and II. These data can be used to compare different methods for the same constituent. For almost all constituents one method was the predominant choice among the participating laboratories. This fact makes a true comparison rather difficult, since many of the results are not rep- resented by a sufficient number of determinations. For'caloium and magnesium (and total hardness) the most common choice was EDTA- titration or determination by atomic absorption, These two methods were equally precise and accurate. Determination by flame photometry yielded more variable results. For sodium and potaosiump flame photometry dominated the picturet obviously being as-good as atomic absorption. Alkalinity was determined by titration with a mixed indicator. This titration yielded very good results. An acceptability limit of about � 0.05 meq/1 and 68-5 percent acceptable results indicate both high precision and aoovraoy in this determination. Only practical reasons can explain the use of the potentiometric titration technique. Chloride was predominantly determined by Mohr titration. The narrow circles in Fig- 7 demonstrate the high precision in this titration, On the other hand, only 21.1 percent acceptable results indicate frequent systematic errors* In Fig, 7 the results are mainly divided-into two clusters of pointap one in the + + quadrant and one in the - - quadrant. In this titration of chloride With silver nitratev potassium chromate is used as an indicator. At the end point the slightly soluble red silver chromate is formed and a certain amount must be formed before it is visible. In the titration of dilute solutions the indicator blank should be determined separately and subtracted from the amount of standard solution used. Presumablyt some laboratories in the + + quadrant have not done this blank correction and have obtained results that are a little too high. 19 Tab le, I Summary of methods used in analysing the samples by the laboratories (interoalibratio'n 1) Laboratories Mean Stand. Coeff. of Constituent and method using method deviation (meq/1) variation NumberlPer cent 1 2 1 2 Calcium Titr. EDTA .............. 17 51.0 1.16�0.06 1.39 0.04 5.5 3.1 Atomic absorption ........ 9 27.3 1.18 �0.04 1.44 0.04 3.7 3.1 Flame ................... 4 12.1 1.19�0.13 1.30 0.62 1.1.2 47.8 Titr. KMn04 ............ 2 6.1 Unknown ................ 1 3.0 Magnesium Titr. hDTA /M(Ca+Mg) -Ca/ ................. 15 44.1 0.39 �0.11 0.51� 0.15 28.7 29.7 Atomic absorption ........ 10 29,4 0.39 0.02 0.46 � 0.03 6.1 5.9 Titr. EDTA .............. 4 11.8 0.39 0.09 0.51 � 0.07 23.1 14.4 Flame .................... 3 8.8 - - Colorimetric (titan or brilliant yellow) ......... 2 5.9 - - Total Hardness/M(Ca+Mg)/ Titr. EDTA .............. 25 64.1 1.54 �0.13 1.84�0.16 8.7 8.5 Atomic absorption (Ca+Mg) 9 23.1 1.58 �0.05 1.90 0.04 3.0 2.3 Combination (Ca+Mg) .... 5 12.8 - - - Sodium Flame ..................... 22 81.5 0.43 0.05 0.73 0.06 11.2 7.8 Atomic absorption ........ 5 18.5 0.45 0.05 0.71 0.06, 10.4 8.1 Potassium Flame ..... 23 82.1 0.08 �0.04 0.11� 0.04 41.7 36.9 Atomic absorption ......... 5 17.9 0.08 � 0.01 0.11� 0.01 13.0 10.9 Alkalinity (HCO-3) Titr. Mixed indicator ...... 44 93.6 0.89 �0.04 1.06 0.06 4.6 5.6 Potentiometric ............ 2 4.3 - - - Unknown ................ 1 2.1 - - Chloride Titr. Mohr ............... 41 89.1 0.38 0.06 0.69 0.06 15.2 8.0 Potentiornetric ............ 2 4.3 - - Gravimetric 1 2.2 Colorimetric 1 2.2 Eel Chloridemeter ........ 1 2.2 7- Sulphate Gravimetric .............. 20 47.6 0.74 �0.11 0.83 � 0.16 15.2 10.2 Titrimetric EDTA ........ 7 16.7 0.71 � 0.14 0.81� 0.09 19.6 11.2 Turbidimetric ............. 4 9.5 0.63 � 0.08 0.76 � 0.05 13.2 6.7 Titrimetric (Thorin) ....... 4 9.5 0.76 � 0.04 0.87� 0.10 5.6 11.1 Ion-Exchange /1:(S04+Cl)/ 3 7.1 - - - - Colorimetric (BaCr04) ...... 2 4.8 - Titrimetric /Pb(NO3)2/ .... 1 2.4 - Conductometric ........... 1 2.4 - 20 Table II Summary of methods used in analysing the samples by the laboratories (Intercalibration 2) Laboratories Mean � Stand. Coeff of Constituent and method using method deviation (meq/1) variation NumberlPer cenJ A B A I B Calcium Titr. EDTA .............. 23 52.3 1.14�0.05 1.30 � 0.07 4.4 5.4 Atomic absorption ........ 15 34.1 1.18 � 0.08 1.33 �0.11 6.8 8.3 Flame ................... 5 11.4 1.18 � 0.12 1.35 �0.14 10.2 10.4 Titr. KMn04 ............ 1 2.3 Magnesium Titr. EDTA /Z(Ca+Mg) -Ca/ ..... "*''** ... - 22 57.9 0.44 0.09 0.54 � 0.10 20.5 18.5 Atomic absorption ........ 14 36.8 0.40 0.03 0.49 � 0.03 7.5 6.1 Flame ...... 1 2.6 Colorimetric .............. 1 2.6 Total Hardness /!(Ca+Mg)/ Titr. EDTA .............. 36 69.2 1.58 0.05 1.83 �0.05 3.2 2.7 Atomic absorption (Ca+Mg) 14 26.9 1.55 0.07 1.78 �0.10 4.5 5.6 Combination ............. 2 3.9 - - - Sodium Flame ....... **''**"*''" 25 83.3 0.43 0.07 0.74 � 0.09 16.3 12.2 Atomic absorption ........ 5 16.7 0.42 0.07 10.69� 0.11 16.7 15.9 Potassium Flame .................... 25 83.3 0.08 0.01 0.18 � 0.01 12.5 5.6 Atomic absorption ........ 5 16.7 0.08 0.01 '0. 18 � 0.02 12.5 11.1 Alkalinity (HCO-,j) Titr. Mixed indicator ...... 50 96.1 0.88 0.04 1.07 � 0.04 4.6 3.7 Potentiometric ............ 2 3.9 0.88 1.07' - - Chloride Titr. Mohr ............... 43 84.3 0.39 0.05 0.58 � 0.05 12.8 8.6 Potentiometric .............. 3 5.9 0.34 0.53 - - Colorinietric /Hg(SCN)2Fe/ 2 3.9 0.35 0.54 Titr. Hg(N03)2 .......... 2 3.9 0.34 0.55 Coulomctric .............. 1 2.0 - - Sulphate Gravimetric .............. 24 48.0 0.77 0.05 1.02 �0.07 6.5 6.9 Titrimetric EDTA ........ 11 22.0 0.74 0.06 0.96 � 0.09 8.1 9.4 Titrimetric (Thorin) ...... 6 12.0 0.78 0.05 1.01 �0.07 6.4 6.9 Turbidimetric ............. 4 8.0 0.61 0.20 0.81 � 0.20 32.8 24.7 Ion-Exchange /I(SO4+Cl)/ 2 4.0 0.87 1.12 - - Colorimetric (BaCr04) .... 2 4.0 0.75 1.03 Conductometric .......... 1 2.0 - - Table III Laboratory identification numbers listed by number Of COnstituents analysed and number of Unacceptable results reported4 (Intercalibration 2) Numbers of Number of Unacceptable Results Reported Constituents __T _6 Analyzed 0 1 2 3 4 5 ............. 60, 2............. 3.............. 14, 4............. 6, - 53, 5............. - 2, 37, 41, 8, 52, 54, 42, 6............. - 44, 10, 13, 33, - - - 7............. 23, 25, 39, - 11, 12, 21, 9, 30, 15,50, 63, 27, 62,28,57, 8............. - - 22, - - - 9............. is, 1, 5, 20, 4,31, 45, 17, 24, 35, 55, 322, 40, 19, 34, 29, 51, 49, 56, 46, 164, 64, 383 43 1TCa+Mg are in some cases calculated 211 results reported. 314 results reported. 412 results reported. @PD Pl ou Ma, A OP @41 rb w rl) Lj J1 -(D cy) cj ro eb (n r) eD fD rl) ro Ln Ln 10 c c :3 eD aj eb Ln ell D -D 3 P cD LA rm co C3 ro wo wo o Im Lj -o o o Li ep 0 W Z, t:r rb Im Pj Ln 0 m Pj w lu rr a. m Li m lu eb cu cL Ij A@ cr _(D rb Mean A oV Ln C@ fu 3 C) C> C:.- C> CD W C) C) Mean A aq 0 cr Li Li rl) 5@ c 0 6, RL Ln > c rD eD eD '@L - w 0 eb or D '@!@ D w lu 2r tr @ 3 -C> 00 it 0 Ly, 0 Mean A C> ..... ..... ......... .... Ln cn D c =1 m Di Ln n n M eD ;u W o -j eD 3 3. @ p p p p p C> zt Mean A MCI M 5 Z- w Ln 0 r_ D ED n el) m fb Ln 1@ w 3 'a c, C) 3 LA [u @ 3 m C@ C) OD Ul w :7 Di > 0 th n a C) rb eb In 74 = a) ro m iT OD Me2n A Ln C, w OOD .4- 3 2 'a LO 0 W ,P 3-C) 3 T 25 Sample j 285- 15- 1 52@ 280- e34 9 27 5- 660 019 e63 @-54 270- 04 0 91 46 265- @13 *17 031 0: 564 *1 55 :1837 *23 e 3 260 45 -@51. - e38 Oe 045 6;64 .22 '21 255- 07 4o *56 050 12 2507 *30 011 '42 .2 Per@entual diOribution .*53 245- 44,6::Unacceptable 240- 16,1 Questionable e33 39,3 Accept 235- .57 able 230 0 u t I i e r s: 3 2 ;27 z 175 180 185 190. 195 260 205 210 215 CM'1 Sample aps Fig., 9 Labwatory Am4sis for Speoifio'Conduotanoe. 29, 4 @)6 1 0 506 .17 55 2 -3 3 .18 -5 .38b r05 26 Some manuals in current use do not stress this blank correction. Finally, this leads to the conclusion that the considered true value, the intersection of the mean-lines, is a little too high. For sulphate the gravimetric procedure was the most common method, but a rather high number of principally very different methods also have been used. In both intercalibrations the turbidimetric procedure gave lower results than other methods used. For specific conductance there is a general improvement in the results between the two intercalibrations, Presumably, this improvement was due to a better calibration of the instruments. In all eases there is a good agreement between the found "mean-difference" and the "calculated difference". There also appears to have been a general improvement in the analytical work between these two interoalibrations. For calcium, potassium, chloride and sulphate the results indicate an improved overall precision, calculated on differences as described above. This leads to narrower limits of acceptability but to a decreased percentage of acceptable results. For magnesium, hardness, sodium, alkalinity and sulphate there is an increased percentage of acceptable results, the limits of acceptability being the same. This type of improvement points to a correction of systematic errors. Laboratories making complete mineral analysis of natural waters can profitably control the analytical results by calculating the anion-cation balance (Greenberg and Navone 1958). This result can also be compared with the specific conductance value. The sum of the anions must theoretically be equal to the sum of the cations, This calculation is the main reason why meq/1 is used instead of the more common unit mg/l. The difference between the sums of the anions and cations may not exceed a given percentage - at lower levels five percent of the sum of the anions is frequently used, In the second interealibration the average sum of the cations (A 2.08 meq/1 B 2.75 meg/1) compares fairly well with the anion sum (A 2.00 meg/1 and B 2.63 meg/1). This possibility for controlling the overall acceptability of the results was obviously neglected in many laboratories. In the first intercalibration 22 results from a total of 289 (i.e., eight percent) deviated considerably., TableIII shows the number of unacceptable results reported by each laboratory as a function of the number of constituents analysed (Greenberg et al., 1969; Greenberg 1961). 6, REFERENCES Ahl, T., IHD-Interkalibrering: Metoder for storre Konstituenter - sommaren 1970- Vannet 1970 i Norden IHD-nytt,1 Funder-Sohmidt, B. et al, Standardization of methods for chemical analysis of water, 1972 Vatten 28:330-2 Greenberg A., Use of reference samples in evaluating water laboratorieso Public Health 1961 Rep., 76:783-87 Greenberg A. and R Navone Use of the control chart in checking anion-cation balances in 1958 watero J Am,Water Works Assoc 6qm 50:1365-70 Greerberg A. et al Chemical reference samples in water laboratories, JoAm Water Works 1969 Assoc 61:599-602 Henrikeen A., Interoalibration methods for chemical analysis of water 2. Results from 1971 interoalibration of methods for determining orthophosphate and total phosphorus. Vatten 27:44-50 Karlgren L. and G. Ekedahl, Intercalibration of methods for chemical analysis of water, 1971 1. Permanganate methods for determining chemical oxygen demand. Vatten 27:32-43 27- Youden, W.J. Graphical diagnosis of interlaboratory tests results. Ind.Qual.Controlq 1959 @01):24-8 Statistical techniques for collaborative tests, Association of Official 1969 Analytical Chemists, Inc. 28 OIL AND OIL DISPERSANTS by S,R, CARLBERG Hydrograpby Department Institute of Marine Research Gothenburgv Sweden Contents E!je 1. Introduction 28 2. Oil pollution from ships 28 3. Effects amd fate of oil pollution-at sea. 31 3.1 "The likely course of events" 31 3.2 Dispersion solution and microbial degradation 31 4. Possible ways of dealing with released oil 33 5. Biological aspects of oil pollution 34 6. References 35 10 INTRODUCTION During recent decades, and especially the last few years, more and more attention has been paid by the authorities and the public in many countries to the increasing problem of pollution by oil. It is evident that the areas of great oil-consumption are not always the areas of high oil production (Fig. 1). While U.S.A. and U.S.S,R, are dependent to a minor degree on importog Europe and Japan have to rely on imports for about 80 percent of their' consumption. The main oil transport by sea thus takes place from the Middle East along two routes to Europe and Japan. In 1971t 1 355 million tons of crude oil were transported by sea, This means that 60 percent of all sailed distances were oil transports. 2. OIL POLLUTION FROM SHIPS Together with the total increaset the character of the transports has changed during recent years. The crude oil used to be refined in the production area and the different products were transported to the consumers in tankers$ which nowadays are regarded as very small units (e.g. 16 000 dwt). Today the crude oil is transported to the consumption area for refining. The size of the ships has increased enormously; the VLCG (Very Large Crude Carrier) varies from 100 000-200 000 dwto or even larger. The use of larger ships means that comparatively fewer ships have to undertake the transportation. This leads, in one respeotg to less risk of collisions and accidents, but on the other hand, these large ships are more difficult to manoeu4re, requiring greater water depths along the routes and in harbours. Especially in areas of intense marine traffiop this increases the risk of collisions and creates a great demand for highly skilled merchant officers and crews, as well as reliable technical equipment aboard the ships. After a tanker has unloaded its cargo to a refineryl certain amounts of oil remain in the cargo spaces. On her voyage back to the crude oil production area, the ship must take in large quantities of water in the cargo spaces in order to acquire sufficient stability* When discharging this water the oil residues will be discharged too. Fit m c C4 30 The first major step tomard Preventing deliberate oil pollution was taken in 1954 when 42 oountriesp including all the maritime powers# drew up the International Convention for the Prevention of Pollution of the Sea by Oil, which came into force in July 1958. This Convention was deposited with the Government of the United Kingdom and was later transferred to the Inter-Goverranental Maritime Consultative Organization (INGO) when it was established in 1959. This Convention (and its amendments of 1962) allows ships to'release -water oon- taining up to 100 ppm oil in international water within 100 nautical miles off land. Outside this zone the oil content in the water is not restricted. In 1969 more amendments were proposed# but so far only a few countries have ratified them. According to the proposals the following rules were to be enforced. (a) Limitation of the total quantity of oil which a tanker may discharge in any ballast voyage to 1/15 000 of the total cargo caxrying capacity of the vessel# (b) Limitation of the rate at which oil may be discharged to a maximum of 60 litre's per nautical mile travelled by the ship, (c) Prohibition of discharge of any oil whatsoever from the cargo spaces of a tanker within 50 nautioa2 miles of the nearest land. Here it may be appropriate to point out that so far no international legislation or agreements exist concerning the economic responsibility for the consequences of an oil 6iatastrophe. While waiting for this problem to be solved, the companies dealing with oil and oil transportationg as well as the oil owners, have themselves reached two agreements and regulations; TOVALOP (Tank Owners' Voluntary Agreement Concerning Liability for Oil Pollution) and CRISTAL (Contract Regarding an Interim Supplement to Tanker Liability for Oil Pollution). Via these two agreements governmental organizations in countries suffering from oil catas- trophies can receive payment for damages and oleaning-up operations. On the technical side of oil spill preventiont the so-called load-on-tdp system (LOT) 'is widely in operation, About 95 percent of the crude oil tonnage is now equipped with LoT. According to the Loll system one or two cargo tanks are cleaned and filled with ballast water. The oily water resulting from the cleaning is not discharged but led to a slop tank. Then the other tanks are cleaned and may also be filled with ballast water if required. All the cleaning water is collected in the slop tankl where the oil separates from the water during the ballast voyage. When the ballast water is discharged it is virtually free from oil. The water in the slop tank is carefully discharged. When a minor part of the water remains in the tank the amount of out-coming oil increases. The discharge is stopped and the now oil is loaded on top of the oil and oily water in the slop tank. Howevert even if the LoT system is a useful.toolt other methods must also be used in order to minimize oil spillages. Some harbours are equipped with facilities to receive and treat residues of used and unused oil as well as oily ballast water. The oil can be separated from this water and all the oil residues can be refined and used againg but separation and other treatments are expensivel and thus many harboure are reluctant to offer these facilities. According to calculations by,Kluse (1968), the LoT system every year prevents 1.6 million tons of oil from reaching the sea. It is of course very difficult to estimate how much oil is actually released into the marine environment. Some calculations suggest an amount of 0.7 million tons from the crude-oil carriers. Howeverp the product carriers are not always equipped with LoT, and all shipst whether oil tankers or nott release oil to the marine envirornent in the form of contaminated engine cooling watert leakages at the propeller shaftet 31 etc. In addition to thiel oil drilling at sea and natural seepages from oil wells make their contributions. It is'very difficult to estimate how much oil annually reaches the marine environment. Figures published range from 2.1 million tons (SCEP.1970) to 5 or 10 million tons (Blumer et al,, 1971); with several estimates in between. 3. EFFECTS AND PATE OF OIL FOLLUTICK AT SEA Once oil is released into the sea, certain damage takes place regardless of which co-unter-qneasures (if any) are applied. Before discussing different methods for oil combatment at sea it is of interest to got an idea of what will happen to the oil if it is left undisturbed. A very useful review on this subject is published by Parker et (1971) from which the following is quoted: 3.1 "The likely course of events" "The possible changes that can occur when crude oil is deposited on the sea are sum- marized in Fig. 2. At firstp the oil will spread rapidly to form a thin homogeneous slick. At the same time, evaporation will take place (25-30 percent in 2-3 days@ so that the oil remaining becomes increasingly richer in the leas volatile components and so more viscous. The rates Of spreading and evaporation therefore decrease. Depending upon the type of oil and the roughness of the seat some emulsification may take place due to natural emulsifiers present in the sea and in the oil. Initially, other effects such as dissolution and chemical and biological actions are expected to be small butp as the slick spreads-out these effects will be correspondingly scoeleratedl although their absolute rates may still be slow. Subsequent changes will depend on the proportions of oil converted to water-in-oil and oil-irr-water emulsions. The former are produced when water becomes entrained with viscous oil by wave action. The water content of such an emulsion may increase to 70-& percentp giving a gel-like mass (the so-called "chocolate mousse"), With large spillages in which thick layers of oil persist for some timet large aggregates of chocolate mousse can be produced. Chemical and biological reactions in such material are likely to be slow because the surface area open to attack is relatively smallp although some degradation could take place if suitable anaerobic bacteria were occluded. On reaching the shore# the "chocolate mousse" will pick up sand and debris and the water will evaporate to leave compact tarry lumps in which further degradation will be very slow indeed, True oil-in-water emulsions may'be formed because of the presence of natural emulsifiers or by the application of detergents. In the absence of artificial treatment, it is likely that the bulk of the oil (in a dispersed form) will exist as relatively large droplets formed by the agitation of the waves. In still water these large droplets would coalesce to reform a slioky but in disturbed water the droplets will be dispersed through a large volume of water, Separation will then be slow and the large surface/volume ratio of the oildroplets will permit other forms of attack. Sufficient solar radiation may penetrate the upper layers of the sea to make photochemical reactions possible. The oil may undergo chemical degrada- tion as a result of bacterial attack, it may be ingested by filter-feeding organisms present in plankton, or it may simply stick to plankton# other marine lifeg or debris and become distributed even more widely." 3.2 Dispersion solution and microbial degradation The lightest fraction of the n-paraffins in the oils has the greatest solubility. These substances are also regarded as the most harmful together with the aromatic compounds, Boylan and Tripp (1971) have shown that these substances may be dissolved in sea water to a greater extent (about 1-5 mg/1) than the higher paraffins. Also the formation of fine' dispersions needs careful consideration as a dispersion containing I ppm (1 mg1l) of oil distributed over one km2 to a depth of one metre amounts to one ton of oil. Discharge Evaporation Evaporation Thick layer T.in layer on-volatile of oil of oil components Dissolution of e. Other light fractions effects Dissolution slow of light Chemical fractions reactions Distribution as fine suspension lighter than sea water Water-in-oil Oil-in-water Bacterial including including Distribution degradation chocolate emulsions and consumption mousse by plankton eto. J Other Bacterial processes degradation slow Chemical Tarry reaction lVmps Dissolution Settling on Degradation heavy particles very slow Fig. 2 Processes leading to the consumption of crude oil at sea (Parker et al., 1971) 33 The microbial degradation of oil is frequently studied and reported in the literature. Yet it is difficult to deduce from these investigations to what extent this degradation plays an important role in the natural decontamination, 4, POSSIBLE WATS OF DEALING WITH RELEASED OIL In the open ooean, or where there is a considerable tidal movement, consideration should, be given to taking no action on .the oil at all. If the number of diving sea birds is small and if the oil is unlikely to be blown by the wind onto a recreation beach, there is little point in wasting effort on mechanical or chemical treatment. The best way may be to let nature solve the problem as none of the presently available counter-measures can completely eliminate the biological damage of oil spills. However, it should be remembered that it is' the coastal waters, the most productive areas of th6loceany that receive the heaviest influx of oil. Suction.and Skimming - A good way :of dealing with an oil spill is of course to collect the oil and remo've it from the water. Obviously this method can only be applied when the amount of oil is reasonable.- In order to prevent the oil slick from spreading it must be confined by floating barriers or booms. Once inside this oonfinemenig the oil can be sucked or pumped up if the layer is thick enough. Another mechanical way is to skim the oil off the water. The skimmer consists of a belt forming a loop between two cylinders. The belt (and one oy linder) extends below the.oii surface and the,oil adheres to -the belt. As it rotates, the oil is carried away from.the water and at the upper cylinder is scraped off the belt. Con- siderhble amounts of oil (several cubic metres per hour) can be collected with this method. However, at present no booms are capable of confining oil when the speed of the tidal move- ment or other currents perpendicular to the boom exceed one knot (0-5 m/Bee'), or when the, waves are higher than 0,5-1.m@, Thus the method-iB generally limited to rather sheltered waters or good weather conditions. Absorbing -and Gel - SeveralL absorbing materials such as strawq sawdustv wood chippingol, eto.p have been used to absorb oil. Gelling compounds are certain chemical formulations whichl after addition to.the oil, transform it into a jelly, The disadvantage with all these mechanical methods is that often great volumes of.material have to be handled at sea during the operation and ashore after the collection, Burni - Several attempts have been made to burn the oil when it rides on,the.water surface. The problem is, howeverl that the released oil.spreade out so quickly that the thickness of the oil layer decreases sufficiently for the cooling-effect of the water beneath to prevent ignition. Attempts have been made to overcome this by using different materials to absorb the oil, that then act as wioksp but these trials have not been very successful. Oil 3 One of the methods suggested for removing oil from the surface of the water is to sink it to the bottom,. However, before adopting this methodl,consideration must be given tot (a) the effect of the oil on the bottom flors, and favamp .(b) long-term effects caused by the bacterial degradation of the oilt (9) movement of the sunken oil on the bottom and the effect of such oil on the marine life. 'This immediately suggests that care should be taken not to sink oil down to the sea-bed in areas which have biological or commercial values, such.as shellfish-beds or spawning grounds for fishes, The repeated sinking of-oil in any one area of the sea should also be avoided, Sinking of oil is achieved by means of an agent-such as a sand slurry, siliconized ashl chalk powder, sulphurt oement.t calcium st4arate, etc.9 that is.sprayed over the oil slick, Gravity forces the sprayed particles to sink down to the sea bottom. 34 Dispersi - Dispersants are surface active agents just like ordinary soaps and washing powders. By means of these agents the oil slick is split up into small droplets. The BUZ- face active molecules are hydrophilic (water attracting) at one end and hydrophobic (water repelling) at the other. This second end can also be called oleophilic (oil attracting). Thusthis part of the molecule is attracted by the oil, In this way the oil is divided into droplets, each one surrounded by the dispersant molecules with the hydrophobic part "attached" to the droplets and the hydrophilic end oriented toward the water. This makes it possible for the dispersant to carry large amounts of oil in the waterl even if the oil is not dissolved in the water. The dispersion of oil is effected by spraying the active substancesp dissolved in a suitable solventl over the oil slick. After thatp sufficient agitation of the sprayed area must be ascertained. This is usually done by the propeller of the spraying ship or towing bare or other equipment through the sprayed area. The dispersing and splitting-up of the oil sliok implies that the size of the surface area increases enormously as the number of droplets increases, This increase of surface area is a commonly cited motivation for the dispersion of oil, as the microbial attack is expected to increasel thus accelerating the decomposition of the oil; this, howeverl has not been sufficiently proved. The concentration of the dis- persant in the product varies from a few percent to about 30 percent. In some cases the solvent used is water, but the most effective substances are not water-soluble, and then a petroleum fraction is used. This often contains aromatic compounds. How much of the die- persing agent is required to treat a certain amount of oil varies depending on the conditions. However, it is not uncommon that one has to use half as much# or the same amount as the oil to be treated. Thusl in many cases great amounts of petroleum fractions (with aromatic compounds) are discharged into the marine environment in order to "remove" a pollution 'caused by mineral oil. The authorities in different countries have very different views regarding the appli- cation of the dispersing agents. For examplep the Canadian authorities have banned their :use and the U,S. Coast Guard restricts their application. They are not allowed to be Used in sheltered waters and in shallow areas in the U,S,A, In the TORREY CANYON catastrophe (March 1967) off the south coast of England, 119 000 tons of crude oil were released* In 'the oleaning-up operations of the water and coastal areas, some 2-5 million gallons (or about 11 million litres) of formulations with dispersing agents were used. 5. BIOLOGICAL ASPECTS OF OIL POLUMCN It has been shown by Blumer at al. (1971),Ehrhardt (1972) and otherel that the sub- stances in mineral oil are taken Tp Voertain marine organisms. However, as the organisms are not capable of metabolizing the components, they mav be enriched to a certain degree depending on the rate of excretion. For instanoeg shellfish polluted by fuel oil still oon- tained considerable amounts of oil six montho*after they had been transferred to unpolluted water. Thusy as these substances enter the food chain they may be regarded as risks for the higher trophio levels. Blumer (1969) has also suggested that the long-term effects'of low concentrations of oil may be even more dangerous and long-lasting than the more evident short-term oonsequencese MwV predatory fishes find their prey with the aid of their olfactory senses, while others get away from their hunters using the same sense. Migrating fishes find their way home with a very sensitive analysis of the smell of a certain area, With the olfactory sensep extremely small amounts of different substances may be localized in-watert oil and its aro- matic compounds may completely hide the natural smell in the water. Furthermore there are risks that fishes may be misled by erroneous clues, "Thien, Blumer saysp "May have a detri- mental effect on the survival ability of any species in the sea# thus concerning marW other species whichp+hrough the food-ohair4are dependent on the damaged one". It is often claimed that modern dispqrsing agents are almost non-toxic with LC50 values as high as 3 000-4 000 mg/l. These values are generally the results from tests with adult 35 .fishes. However KUhnhold (1972) has shown that chemically dispersed oil at sub-lethal ooncen- trations considerably prevented the hatching of herring larvae. The hatched larvae were to a great part deformed and mostly died within one day. Furthermore the chemoreoeptors of the larvae seemed to be blocked quickly. The larvae did not avoid the contaminated water when they were given the opportunity to do so, One other aspect of oil pollution is the ability of oil to concentrate DDT and other non-polar hydrocarbons (whether they are chlorinated or not) which are insoluble or almost insoluble in-water. In a laboratory study Hartung and Klingler (1970) found that mineral oil in a sediment could concentrate pp-DDT from water by a factor of 1.08 *100. Analyses of sediments from the Colorado River confirmed this, yielding a concentration factor of 1.45*10 6, REFERENCES Blumer, Mel Oil pollution of the ocean. Paper presented to the Symposium "Man's chemical 1969 invasion of the ocean",La Jolla February 1969 (mimeo) Blumer, M. et al Asmall oil spill. Environments 13(2):2-12 1971 Boylan DB,and B.W. Tripp Determination of hydrocarbons in sea water extracts of crude 1971 oil and crude oil fractions. Nature Lond 230:44-7 British Petroleum Company B.P. Statistical Review of the World Oil Industry. London 1971 British Petroleum Company Ehrhardt# Meg Hydrocarbons in marine organisms, Paper presented to the 8th Conference of 1972 the Baltic Oceanographers, Copenhagen 1972 Pap,25 Hartung# R. and G.W. Klingler Concentration of DDT by sedimented polluting oils. Environ. 1970 Sci.TechnoL, 4(5):407-10 Kluss WM Prevention of sea pollution in normal tanker operations, In Protocol from the 1968 British Institute of Petroleums Summer Conference in Brighton 19689 edited by P. Hepple, London, Institute of Petroleum KUhnhold WW The influence of crude oils on fish fry In Marine pollution and sea life, 1972- edited by Me Ruivot West Byfleet, Surrey,Fishing News (Books) Ltd.,pp.315-8 Parker CA M Freegarde and CG Hatohard The effect of some chemical -and biological 1971 factors on the degradation of crude oil at sea. I 'Water pollution by oil edited by P. Hepple. London, Institute of Petroleum pp,237-44. SCEP, Man's impact on the global environment Cambridge, Mass,MIT,Press 319p. 1970 36 CASE STUDY: CHEMICAL ANALYSES OF A SEA AREA POLLUTED By MINERAL OIL by SR CARIBERG Hydrographic Department Royal Fishery Board Gothenburg, Sweden Contents page 1. Introduction 36 2. Analyses 36 1. INTRODUCTION In January 1973 a field of drifting polluting oil was detected outside the town of Varberg, on the want coast of Sweden. The Coast Guard Service informed the Fishery Board about the situation and two staff members of the hydrographical department joined the coast guards on a patrol ship in order to make observations and collect samples* The oil field consisted of large as well as small lumps of thick oil# separated by clear water* Samples .of water taken between the lumps and outside the field revealed that the concentrations of non-polar hydrocarbons were below 0,05 mg/1. That only small amounts of hydrocarbons were found even In the water samples within the field can be explained by the fact that the oil, was thick and viscous due to the low water temperature. 2.ANALYSES Fig* 1 shows an IR spectrum of the oil. In the first spectral region the sample was run at two concentrations, thus giving two curves. The three peaks which are Marked with crosses, are present in all IR spectra of mineral oils. The GLC analysis is presented in Fig* 2. The oil contained substances with a wide range of boiling points* This can be seen from the elution temperatures which range from 70-3500C. There axe also two pronounced maxima around 1500C and 300C- To got an idea of how different oils look in these analyses a low boiling fuel oil (of the type fuel oil No, 1) is shown in Figs* 3 and 4* In the IR spectrum the three characteristic peaks am found again* The GLC shows that the boiling range of this oil in such narrower, with elution temperatures not exceeding 235C and with a maximum at 14000. All peaks are an the same scale, It must be emphasized that this fuel oil No. I was not prevent in the oil spill; it in just shown here for comparison* The oil was emulsified to sow extent, using a petroleum-based emulsifier, This opera-- tion was later interrupted however when It was decided that the oil was no longer a threat to the coast. The IR spectrum of the emulsifier run at two concentrations Is shown In Fig* 5 where it is clearly seen that the product In petroleum based (the three peaks marked with crosses) The upper curve reveals a number of smaller peaks (arrows, discussed below) which are seen better in the lower curve,, obtained by running the sample with a higher concentration. The GLC curve In Fig. 6 tolls us that the petroleum base is of a low boiling type (in fact It lies in the c10-c14 range). Four peaks dominate the chromatograms, which has no components above an elution temperature of 144C.The emulsifying agent cannot be seen by GLC with the column used In this case. 37 After the emulsifying operation was carried out,some water samples were collected. The IR spectrum in Figs 7 shows the extracted substances (this spectrum was run after evapora- tion of the extract to dryness on a diso of sodium chloride). In the quantitative analysis the extract had to be diluted 40 times in order to be readable in the photometer, The oon- centration of non-polar hydrocarbons thus found was about 40 mg/1. The three oil peaks are marked with crosses. In Figs* 5 and 7 we can compare the peaks marked with arrows indicating the emulsifier (surface active agent). At 1620-1640 cm-1 a small band is seen which points- to the presence of alkyl phenols. At 1100-1120 cm-1 a strong other band in seen; this is characteristic for surface active agents of the non-ionic types Finally, at 950 and 830- 850 cm-1 a peak and a band from the ethylenoxide chain in the. emulsifier, are found. Thus there is no-doubt that the used emulsifiers present in the extract from the waters Prom the IR spectrum in Fig, 7 we cannot say if the oil bands are from the polluting oil or from the solvent of the emulsifier The GLC curve In Fig* 8, however, tells us that there are no hydrocarbons boilling above 1500C* If this GLC in compared with that-In Fig* 6. we can see that even if some now peaks appear the positions of four peaks - and the size dis- tribition of these are quite. similar to that of the solvent for the emulsifier. On the other hand no traces of peaks from the polluting oil can be seem this not imply that the emulsifier is unable to "dissolve" the oil into the water in the forms of an oil-in-water emulsions In this ease a very small amount of emulsifier was used and therefore the effect was weak. What has been said and shown above should of course be compared with the natural comi- position of the organic compounds which can be extracted from sea water of the areas In Pig. 9 a GLC analysis of such an extract in shown, This sample was obtained from outside the drifting oil fields The total concentration according to the IR measurement was below 0,05 mg/l Thus it was not possible to. run a complete IR, spectrum of the sample. When .looking at the GLC is should be remembered that the sensitivity of the instrument-and-con sequently the sizes of the peaks - axe often changed during an analysis* If the peaks in Fig. 9 were to be given on the same scale say that of region B the peaks in region A and peak C.would be five times greater and peak D would be ten time greater than in the figures Thus the GLC is dominated by three peaks:C,D and one in region A. The total appearance of this chromatogram tells us that it represents products excreted by plankton, probably phytoplankton. All IR spectra were run in a precision grating instrument, either in solution (carbon tetrachloride) or evaporated to,dryness an sodium chloride disco. The GLC analyses were carried out in an anallytical gas chromatograph with a flame ionization detector* The column used was 2.1 m long(all glass) filled with three percent .of OV-1 (a silicone oil) impregnated on Chromosorb W 60/180 mesh. Theoven was programmed at a temperature rate of 10C)minute from 50-70C UP to 350C- The samples were injected (1-10 ul) as solutions In chloroform or carbon tetrachloride* . ..... . .. .. L7 - - - - - - - - - - - - - - - - - Programme ending i 0 300 PC 4+ 110 20 Ov UO 70 40 lQO IL G+ C 150 .0 0 tt Programme start 10 PC/minute 4ing-11 --- fF- 6C sl@ `4l 14 OD 0 0 M141MMIMIM lillillitilitillift Kid I Itmailifliff1pi I 11111111111111141 111 liliNiThIllili HUN d1HWIN111 fiT 11111111hill I .1 MIN111M I 1111IN lillill H Will lil, jli 11111111111 H Ill ill! With H I!!! !1111!!i fthl t 1 1, 1 W Ill it 1; t1jill jP'k I I 11111H1111111 111 d Hil 1111IT! I I I IN 111liddl hidil 1T ill i! I 1111H111111111111 1H kl!llll ill 11 H j I h lid, dl Will 11 N I !I I I ill hilik ill TT ll! ill 1111111! [Wilillil-ill] I I HH! @ I I . A''! Jh:il Hill i 1 i l;li; i 1111111111 H1 1 11 111111,1111ilk lhlli;@ill i; R AMIlMi hill i I M ill@ Wilillkllh ii III I MII@Ihlii'!, it 1, 111@ I;, it I 1111IM1111, ikillik :1111111 111111ilijilil IIIIII i!l 1111 N!" Ili Hilhili. !!:iill j;, j I j;j 1 il :1 1WildTli]II :!i !I 11111 fidlidi ihilkliT llWiTlilli 11111111il T1.1111H I Hill d1jili Hill ilk [T 1111111 kh!l 111111111, iiiijilij Ilp Ild ij;! il jUlilll I HiII111111111 -I I! ill! 15,j :1@ it 1 !111111 Hill 1111 li;j !!it I !I il I 1111 It 01 HVII I li I ih Hill H! :111111;11W, I it it: Irl ii'll, 1 101 IT! t;: NI X U ill Hill k; It Nil i 1 +1 tjf@j q" Til ki! Till: it 21!dil i _.J. 2 11 11111111 it 1111111th 1111 Ill Mlil 11111 Il 1kill ZE, it OF 235 -C 0 ...... .. ... 0 E0 10 0 0 70@ 140 OC Mir -7 -7n Programme start 10 OC/ininute 23 III Wrl lit WRI-1111 [I 111111flaff IIIIIIIIii TIN1111 11111111111 11@111INIII NW111IR INT11 I MIT IAfllk@' Olf! 1111111111 11111111dil WIT111 111!jIllillill !JIM 1; 1111111111 THH11111 It I I ; 111 il T11 1@i wli 111 i1:1 llilflil 11111iiiii IT I!H I 111HI'lil .111 !'t k ill I ILL ill -. l. , I I i t It III @@ 1: 1 II;:. !;@l IN I 4LIU T 4'11 11 I'l lit @Ih !;.t i@i ;11 it i; 1111 ii1dii ild Hit tl@: I: I 'i 111101!@ 11@1 ti li'iP@i Ill: !Ii@ ill Hliiih i lit 11 J1111,111 I 1 11 d iki Wilt 11W11!I! 'I :t Ili Mild 11 IiH ilil IlIlli"i! lH[il111Tid1lIP;i it it "IN11:1111w: 1''. 41 ill Hit Mil 111:11111 W id Hit -11111111 Hil @i 11 11 ill@ ii!l di! 11111 tit. III ihp II I f!@!liiiT 'N 1 11 Hill it 111 it 11 "i I I I I I iI 1 11.1. H I! I I! i I ITdii ! @! @ I! i @ !: L! 4tq 11 it ''1 41,1 lit 14 ri, Hfl HH I;i i;lt +j1 @ I :':I it lit A _77 I tw" I I -! !!! , - l 0 lit it fllI1l:l1l!1Ii Ill!! .7 11111H 1111 IT lhi 1111k, it I" z 1711 V .:I: lit; 10 10 10 it0 iD SD to i to 0 -4 0 150 C 1 100 70 60 40 0 Programme start 10 IC/minute ililil@T;:Ji All I w: 1! :7 f 71-1 Ttij :1:1 IT: IL J L-- ;j I dh ;-.i- T fvll@ 71 1; 41 it All" lidl i4 till Ll IL-Ill 0 7- ?+ co cl- 0 150 OC to 10 0 1 0 Fj 0 0 --- - - ----- Programme start 10 'C.minute 1 .1 4, _@_ - @llr L150 OC M ------ Programme ending 350 Oc 0 :6- 0 0 4D OD OD 140 4+ 180 OC @pp 0 cr 0 <." Programme start 10 OC/minute 13 Programme, 9tr 47 ECOLOGIGAL X"ECTS OF MARINE POLLUTANTS by A*D* XoINTM Department of Agriculture and Fisheries for Scotland Marine Laboratory Aberdeen, U.K. Contents Page 1. Introduction 47 2. Entry of pollutants into biota 48 .3. Effects of Oollutants 49 3*1 Field studies 49 3.2 Experimental studies 53 4. References 55 1- INMDUCTION A pollutant may for our present purposes be defined an a substance present in the sco- system, as & result of sea's activitiewl in quantities sufficient to produce undesirable effects. As to effectag lot us, in the present context,, exclude the possible results in man' of consumption of contaminated foodg and consider only the changes caused within an ecomystes IW the pollutant in questione InitisllylLit is fteoeasary toLdeteot theLpremenoe of the pollutant and to measure it quantitatively9 and the first problem is 'to decide if any given level measured is in fact ."pollution" an defined above. Artificial substances such as some peatioides and PCBs are made exclusively by man and their presence in nature can always be attributed to human activities. However, mazW other substamem which can be toxic if present in sufficient quantity (nutrient malts, hydrooarbonol, heavy metals) are always widespread, and it is impertant,to distinguish between the natural levels on the one hand and azV input by man ,on the other, if a realistio approach im to.be made to the problems of pollution. A good knowledge of "basalineN levels is therefore an essential requisite for recognition of conditions which night cause concern. Table I(,&) shows levels of naturally occurring metals in sea water and organimas from open water situations and Table I(b) gives values from specific coastal areas where industrial effluents are numerous. Table II shows similar data for sase artifteiial organic substances indicating thatLlevels of organochlorines are greater in an industrial estu=7 than in the open sea. Tabl. e I (a) Concentrations of'some metals in ppm V - Yb Cd Cu. Water 0.0003 0.0003 000001 0.003 0.01 Plants (dry weight) 0.03 8.4 0-4 10.0 150 Animals (dZ7 weight) 0-01-0-5 0-5-10 0.1*-3.0 4-5 6-1 500 48 (b) Concentrations of some metals in organisms from coastal waters in ppm dry weight 99 Pb Cd Cu. an seaweed Paous,vesioulo - 33-0 Zboplankton Sacitta ep'.2/ - 405 1.1 5-0 126.0 folluso Xytilus jdais.3j 2.1 - - - - ?ish Surattus MraltusA/ 7.4 0.2 7-0 140-0 Preston (1973) Severn Estuaz7, England G. Toppl g (personal oommmi;'ation) Clyde Estuary, Scotland Jones et als (1972) Tay Batuary,,Sootland Z/ Anderai-en =et-al. (1973) Inner Oslo fjord Table II Organochlorines in ppm wet weight (A*V. Holdent personal oommunication) Tbtal PCBs IMT. (Open-sea), 0-003 - 0.016 0-04 - 0-06 Plankton Olyde) OeOl - 0-07 0-05 - 1-1 (open se@) 0-04 - 0-09 0.01 Herring (Clyde) -0.2 - 0-5 0-8 - 1-5 J Although we can detect elevated levels of some chemical either in the water or in the sediments, before we are justified in affirming that it is a Pollutant with a biological effect,, we must show first that it can enter the organisms and second that its presence in the organisms is in some way adverse. 2. WIRT OF POLLUUM 33FM BIOU Regarding entry to the biota, we can consider three types or substances. Pirst, those which are normally present at low concentrations in the sea and may be taken up by marine organism at transfer rates proportional to the environmental concentrations, without any threshold. Second, those substances such as nitrogen, phosphorous and silicon whose oanoen- trationo vary greatly from month to month in the sea so that a trophic level such as the - primary producers will have evolved a range of species adapted to the range of oonoentratimaq end a threshold response might be expected. Finallyq there are those artificially produced substances like ]MT. not naturally present in the sea and for which only experimental tests can show what is the nature, of the trenefer'from enviromment to organism. 49 While entry into the biosphere may take place at any trophio level, transfer from water to organism is probably most efficient in the mioro-organisms (e.g phytoplankton and bacteria) because their surface area to biomass ratio is relatively large and transfers across membranes and into cells are likely to proceed at a more rapid rate or at lower sea water concentrations, than for large organisms. Having-been incorporated into cells of,, for,example, phhytoplankton, a pollutant may be transferred to the bodies of the animals grazing on phytoplankton and later to those which prey on the grazers. The pollutant, sometimes in an altered form, may thus be transferred up a food chain. 'There is, therefore, a tendency for persistent substances to accumulate and be concentrated in the higher trophio levels. 3. EFFECTS OF POLLUTANTS HaviNG shown that the pollutant does in fact enter the organisms we must next find if it has an adverse effect. There are two ways in which pollutant effects can be demonstrated. The first is by detailed field observations which must usually involve long time series of data and if possible comparisons of several similar areas, some of which are free of_pollu- tion so that changes in the populations can-be detected. The second is by experimental .studies on single speci es, or better interacting populations over long period%. 3.1 Field studies 3.1.1 Gross effects Considering first the-field aspeot,in oases of roses pollution there is no problem of detection. A totally azoio zone or a zone of substantially reduced fauna in the vicinity of an effluent pipe or other point of discharge is-sufficient evidence. Surprisingly this seems often to result not so much from highly toxic substances, an from situations where there is a large and rapid input to the bottom of material which, even if completely non-toxic would have a smothering effect and could destroy the community,on which it lies. A well documented 'case is that of the aluminium works near Narseilles (Bourcier, 1969). where sludge, consisting. mainly of oxides of aluminium, iron and silicon are discharged from a pipe at a rate of 85m3/hr at a depth of 350 m into a submarine oanyon. After about two years of this discharge, azoic zone was produced, about 2 km wide and, stretching down the oanyon into deep water for about 6 km, where the sludge covers the bottom to a depth of at least 12 cm. For 2 km or so an each side of the azoic zone and stretching down the slope for several km'more to a depth of about 1 800m there was evidence of the deposit.but at much reduced thickness and rather.patchy. Here the fauna was normal and some of the animals had been ingesting the effluent and maldanid Worms Were using it to build tubes, suggesting it was not particularly toxic. Similar exmples of blanketing effects oould be quoted from other arms, not just from red muds produced in the manufacture of aluminim but also from other inert materials such as. fly ash and china, clays. The important point is perhaps that they result in azoic zones only as long as the material is rapidly accumulating, but that given time to stabilize and to weather, a new and perhaps different fauna may colonize the ground. Another type of pollutant which produces effects evident at long distances from the source is oil. This causes obvious lose of amenity on beaches and it certainly kills many sea birds. It mar also produce some mortality of other marine.biota, particularly on the shore by smothering attached organisms, but in general apart from.the possible effects of low conoen- trations of soluble components which will be discussed later, the effects of oil on marine ecosystems may be less serious than the visual appearance would lead one to expect. Often more damage is done.by the methods used to disperse Oil. (such as certain detergents) than by the oil itself, and the general advice at present is to try to prevent oil coming ashore, and even if-it does to leave it as far as possible to natural prooqmses which will eventually break it down and assimilate it. 50 3.1.2 Subtle effects Apart from such gross and obvious examples of the presence of a pollutant,, we are usually faced in the field with detecting effects which may not be at all clear out. The problem is that environmental conditions are continuously varying on a substantial range of time scales, and causing fluctuations in the populations of the organisms which make up ecosystems. It is thus difficult to distinguish these natural fluctuations from any changes which may have been caused IV pollutants. A good example of this difficulty in shown tr a long series of obser- Ivations made on the bottom fauna of a sandy bay in a sea looh on the went coast of Scotland. Fig. I shows the changes in population dowity.of the dominantorgazism of the intertidal benthic populations - a small bivalve mollusc Telling tenuip In 1965 the population was high but the numbers. declined yew by year and there was no successful recruitment. It is of ooiwse well known that in many species recruitment is not an annual feature, and indeed that one good brood may dominate a population for @evaral years. But the gradual decline of a population over eight years is shown here and if at any period after 1965 a toxic effluent ,had been introduced into the bay or if there had been an oil spill, there is little doubt that this would have been blamed for the population decline. Detailed study of this population and of many of the ecological factors involved has shown that in fact natural conditions such as predation and adverse weather can probably explain the decline. Problems of this kind are particularly relevant in determining the effects of so-oalled "thermal pollution". With the increasing world demands for power, electric generating stations are increasingly being built and these are frequently sited on the coast where supplies of cooling water are readily available. This may result in heated effluent discharging into the surrounding water often at temperatures considerably above ambient, so that a region of increased temperature is produced. 'The consequence tauds to be an increase in the metabolism of organisms exposed to this, and again it is difficult to detect effecto.which can clearly be separated from natural fluctuations. This type of situation has been examined in detail by Barnett (19.72) at a power station in the Firth of Clyde. We are not here concerned with effects on organisms which pass through the cooling system - Barnett suggests that this will not harm bivalve larvae, and Fox and Moyer (1973) deduce that dead micro-organisms from this source may have a fertilising effect on outside waters. We are rather concerned 4ith biological effects.in the vicinity of the outfall. In the Firth of Clyde, Barnett and his colleagues showed that several of the dominant species living in the sand near the power-station were influeneedi- For example, a population of the gastropod Hassarius reticulatus living at and just below the low water mark was studied in the area of the power station efflnentg and compared with a population on a beech several km away* It was shown that the animals near the outfall had significantly thinner shells and that their breeding cycle was considerably modified. Spawning started in late Januar7 and early February, some three months earlier than the normal population. One,oonsequenoe of this is that the larvae would appear in the water before the pbytoplankton on which thqy food had developed their spring bloom (controlled by light),, so that a substantial mortality of the larvae by starvation could be e3W4oted. Other animal groups - bivalve molluscs and amphipod and oopepod crustaceans were also studied, and in all cases changes attributable to the increased temperature were detected. These changes, however, could be rooognised only because the region had been carefully studied for a number of years before the power station came into operationg and because nearby sandy beaches were available for comparison an controls. By these means it was possible to distinguish effects due to the elevated temperatures from natural fluctuations. Looking at this question of natural fluctuations over a longer time series and & wider range of species, work of the Oceanographic Laboratory at Minburgh provides information on the plankton of the Worth Sea over more than 20 years, and indicates a significant decreasing trend in the numbers of several important species (Colebrook, 1972). However, there is a 'similar trend in the open north Atlantic and it is not possible to provide & causal link between this plankton change and any not of environmental facts for which mossuremeats are available. It was shown, however, that at least tan years of observations would be required before a trend could be statistically demonstrated showing the time scale involved in this type of work. Numbers per m2 4- CD CD C:) C) C) C) C) C) C) CD CP Ln (D 04 OD (D M 13 C3 U) 52 It is relevant that while this decreasing trend in plankton in the North Sea has been shown, in the same period there has been an increase in the yield of commercial fish from that area. Some of this is due to the greater lendings of pelagic fish thanks to the introduction of the purse seine net,4ut the demersal. landings also rose and this was due to exoeptionally good broods of several species during 1962 and 1967 in particular. Thulk although we can measure increases in the level of pollution in the North Sea and can detect a deorease in the plankton, any corresponding adverse effect on fish must have been masked by some undetermined beneficial fac'tore. Thus,a recent meeting of a working party of scientists concluded that there was no evidence of an effect of pollution on our North Sea resources (Goldberg, 1973). This did not mean that there was no effect but rather that it was not possible to deteot one in the face of actual trends. 3.1.3 Eutrophication Field observations thus seem to provide unequivocal evidence of pollution effects only when these effects are obvious and serious, but there is one aspect on which our knowledge is becoming more precise - eutrophication. Nutrient enrichment can be seen as increasing primary prcduction with some increase in other trophio levelog and is likely to be beneficial as long as the rate of enrichment is not too high. When the rate of enrichment becomes too high however, damaging effects on the environment may occur. As.a, result of numerous studies we have some idea of permissible rates of nutrient addition for certain types of environment (Ketchum, 1971),and it is sometimes possible to predict the changes when they occur. At higher trophic levels also we have some idea of the effects of organic enrichment. Much information comes from studies of grounds where sewage sludge has been dumped at sea and the dumping ground in the Firth of Clyde is a good example. Sewage sludge from the city of Glasgow has been dumped on one particular ground (Garroch Head) for almost half a century and the current rate is more than one million tons per year. Studies of this ground and a comparison of its characteristics and fauna with clean grounds nearby show that the dumping has indeed produced an effect. First the mud is black and fibrous and has a foul emell and the organic content (3-8 percent) is considerably increased, while sediment on nearby ground 1.5 km or so away is grey, of fine texture, odourless, and has a considerably lower organic content (0.3-2.0 percent). Even more significant is the difference in.fauna. On the Olean ground a good mixed fauna of many major groups of molluscs, crustaceans,, 9ohinoderms and worms is found, numbering about 1 300 individuals per square metre and with a biomass of about 40 g dry weight. However, as the dumping ground was approached, the species diversity decreased and the number and biomass of a few species increased until on the middle of the ground only two or three species of worms were found, but the-numbere of individuals reached almost 50 000/m2 and the biomass almost 130 g wet weight. However, this marked effect was, confined to the centre of the dumping groundq an area of some square kilometers, so that the effect of &-long period of dumping was considerably restricted, and it could be suggested that the environmental loss is small compared with the economic convenience. It is perhaps useful to speculate on what increased load this dumping ground would stand without becoming totally azoic. Experience in New York Bight in relevant here (Piercog 1972). Sewage sludge from the New Tork metropolitan area has been dumped there for some 40 years, a%d recently in quantities almost ton times those in the Clyde. This has resulted in a sediment which in places is three or four times rioher in.organic matter than in the Clyde and, as would be expected, in a very substantial modification of the fauna. However, there are no large stretches completely devoid of lifev and again the total area affected, some 36 km2. is relatively small. It may be of interest to consider the different sorts of situation available for sludge dumping,, and the areas chosen by Glasgow and London perhaps represent the two extremes. As described above, Glasgow sludge in dumped in water of 90 a depth in a region of little water movement, so that the material is concentrated in a relatively small area which is thus significantly altered. London sludge on the other hand, is dumped in the outer Thames in qUiteLshallow water (15-20 m) where strong tidal currents are found. It has not been possible to deteci-any substantial effect of this dumping. Surv4W@ (Shelton# 1970), indicate nothing 53 more than a slight increase in n@mbers and diversity of the polyohaete worms. The choice between deepg, still water, where the sludge can be detected and monitored, or shallow, flowing water,- where the material is widely dispersed, must be made on the basis of local conditions and fishery considerations, and also in the light of the precise composition of the sludgeg which may ormay not.contain toxic substances (trade and industrial waste) as well as sewage. ,3.2 Xxverimental studies 3.2.1 Short-term toxicity experiments The simplest of these tests examines the direct transfer of pollutants from water to organisms. This is done by short-term toxicity tests when selected species are kept in a range of concentrations of the pollutant for iseveral days and the concentration for 50 percent mortality is determined. Such tests p rovide valuable quantitative information on effects, but it is often difficult to extrapolate the result to field conditions and to greater timo soalea. The aajqr difficulty is that in the sea we are usuallyLdealing with low levele.of a wide mixture of pollutants acting.over long periods and also that in the sea there is the question of transfer through whole food chains. 3.20'2 Experiments with natural water on individual species One approach to the problem in to experiment with the effecto.of.the water taken directly from a polluted area. The recent work of Steele ot al. ( 1973) on herring is aL good example. Eggs from spawning. herring on a small bamk in tkie firth of Clyde were collected and fertilized artificially from-male herring caught in the same area. The eggs were then hatched and the resulting larvae grown in water from four different locations: first in water from the spawning ground (Ballantrae Bank), second in water from the central deep section of the Firth, then from water collected inshore in a polluted coastal region (Irvine Bay) near the spawning bank, andflinally, as a control, in clean water from Loch fte on the north-west coast. The results show that the time taken for the eggs to hatch, and the mortality after hatching, was signi- ficantly different in the four water types. The order was the some in each case, with a short time to hatching related to a highmortality. The mortality rates after hatching were highest in the inshore areas of the Clyde; 72 percent for the polluted ooastal-region and 33 percent for the spawning bankt with a very low mortality-, 6 percent in the clean water from Loch Eve. 2heso effects may be ascribed to some factor or.faotors in the different waters# and it is particularly significant that the inshore water had much higher nitrates and had &-greater BOD than the others, indicating contamination. It in also known that plankton and herring from the Clyde have higher levels of DDT and PCB*.than those from outer waters off the north-went of Scotland (Williams and Holden, 1973) 3.2.3 Food chain experiments While results of this kind demonstrate that polluted water in the field does have effects .on single species at sublethal levels, there is still the question of how whole food chains may 1@e influenced, and this required not short-term tests on single species, but long-term tests an at least simplified food cha ins. Work of this sort has been done on the ecosystem of sandy,beachas (Saward et-al., 1972). These beaches are'often nursery grounds for flatfish, and in some such communities a favourite food of the fish is the siphons of the small bivalve, Teiiiiii, which are cropped off by the predators, the bivalves themselves feeding mainly on terial filtered off from the water. We have therefore a food chain, phytoplankton-to bivalves to fish, incorporating the three main trophic levels, primary productim, herbivores and carnivoresp and after some trials we were able to set up this food chain in large experi- mental tanks indoors. Using copper as a pollutnat atthree different concentrations with a control we-studied over a period of ten days the effects of these three trophic levels, having first determined the 96 hour LD50 of copperm 1211JQA and plaice as 1 000 and 750 micrograms per litre respectively. The level in.sea. water in me of our industrialized areas in the Clyde is about 3 micrograms per litre and the concentrations used in the experiments were 10, 30 and 100 micrograms per litreo 54 These experiments showed that the copper added to the wiater became distributed through- out the entire experimental ecosystem. In the water it was mainly in soluble form and only a small proportion was found in suspended particulate material. The bottoms of the tanks were covered by a layer of sand, and copper was taken up by the layer in amounts reflecting both the level of copper in the water and the duration of exposure - in no case was a plateau concentration reached during the course of the experiments. Considering next the biota, shell and soft tissues of Tellina were analysed separately, and although copper was taken up by both, the accumulation in soft tissues was much greater. Levels increased with both dose concentration and period of exposure. After 100 days the Tellina soft tissue concentrations of copper per unit dry weight were 1 100 PPmt 470 ppm and 270 ppm for dosage regimes of 100, 30 and 1OAg Gu/1 respectively. The normal background level of animals collected from the beach was less than 50 ppm. For the fish, there was little change in copper content of body tissue during the experiment, but the levels in the viscera increased considerably. The copper content per unit dry weight of plaice viscera exposed to 100/ug Cu/1 for 100 days was 570 ppm, compared with a normal background of 30 ppm. Corresponding with the evidence of accx=ulation in the food chain, there was evidence of metabolic effects. Both the chlorophyll a and total pigment levels of particulate matter in the tank water decreased with increased c-opper dose and the total c14 fixalis for treated tanks was lower than the control. For Tellina, condition indices (dry soft tissue weight/shell length) and carbohydrate reserve levels were F@wer in all treated tanks, and mortality was greater at dosage levels of 100 and 30,.ug CU/1. Considering the fish, growth of plaice in all polluted tanks was lower than in the control and no threshold was indicated. It was of interest that fish biomass in the 100/ug Cu/1 tank was higher than in the 30,gg/l tank and this can partly be explained, first in that an algal mat which forms on the Band surface and interferes with fish feeding was much reduced at high dose levels, and second in that withdrawal of Tellins, siphons -is inhibited again at high dose levels, so that food availability for plaice was increased in these tanks. The conclusion from these experiments is that copper affects the Telli plaice food chain at all levels, not only directly but also in a complex manner due to interaction of ecosystem components. One important point is that while the viscera of the fish showed some increase in copper levels, there was no increase in the flesh so that we have an effect without the usual chemical evidence. We have thus been able to show .that at concentrations not much above those found in the natural environment we have significant effects on the ecosystem and a further justifi- cation for extrapolating this to the field is that the levels of copper in our experimental Tellimi after 100 days was similar to that found in Tellina in nature near the outfall of one of the main effluent pipes in the industrialized areae 3.2-4 Parent-offspring experiments We have shown that pollutants can have ecological effects on being transferred to organisms from the water and also through the food chain. One final consideration is effects by transfer from one generation to another of a given species. Relevant work has been described by Menzel 2t @Ll. (1970). They reared in brackish water oalanoid oopepods in various DDT concentrations from juvenile to adult, and the nauplii produced were transferred to clean sea water and their growth studied. It was shown that offspring whose parents were reared in concentrations of at and above 0.01 micrograms per litre could grow to the pro-adult stage but could not meta@_ morphose into adults. There has thus been some effect transferred from one generation to another at &,MT concentration that could be found in polluted areas@ 55- 5.REFERENCES Andersen, A. T., A. Dormmes and J.H. Heathagen Some heavy metals in sprat (Sprattus 1973 sprattus and herring (Clupea harsums) from the inner Oslo fjord.Aquaculture, 2:17-22 Barnett, P.R.O. Effects of warm water effluents from power station on marine life. 1972 Proo.R.Soo.Lond.(B.) 180:497-509 Bouroier,M. Ecoulement des"bouesrouges" dans le Canyon de la Cassidaigne (decembre 1968). 1969 Tethys. 1: 779-82 Bowen, H.J.M Trace elements in biochemistry. London Academic Press 1966 Colebrook, JAL, Changes in the distribution and abundance of zooplankton in the North Sea, 1972 1948-1969. SYmp.Zool.Soo.Lond.,29:203-12 Fox, J.L. and M.S.Moyer, Some effects of a power plant on marine biota. Chesapeake Soi, 1973 14(1):1-10 Goldberg, E.D., North-Sea Science. Cambridge, Mass. MIT Press- 500 P. 1973 Jones, A.M.,Y. Jones and W.D.P. Stewart, Mercury in marine organisms the Tay region. 1972 Nature, Lond., 238:164-5 Ketchum, B.E, Pollution of estuaries and coastal waters. In Man's impact on terrestrial 1971 and oceanic ecosystems, ad. Matthew, Smith and Goldberg. Cambridge, Mass. MIT Press- 540 P. Menzel, D.W., J. Andersen and A. Randtke, Marine phytoplankton vary in their response to 1970 chlorinated bydrocarbons. Science, Wash. D.C. 167:1724-6 Preston A., Cadmium in the marine environment of the United Kingdom. Mar.Pollut.Bull., 1973 4(7):105-7 Pierce, J.B., The effects of solid waste disposal on benthio communities in the New York 1972 Bight. . I& Marine pollution and sea life, edited by M Ruivo. West Byfleet, Surray, Fishing News (Books) Ltd., pp. 404-11. Saward, Do, F.A Stirling and G. Topping, Experimental studies on the effects of copper on a 1972 marine food chain. ICES ON 1972 Fisheries Improvement Cttes Pap.E:24 Shelton, RAJ, Me effects of the dumping of sewage.sludge on the fauna of the outer Thames 1970 estuary. ICES ON 19709 Pisheries'Improvement Cttee Pap.ES8 Steele, J.H. et al.. Pollution Study in the Clyde Sea area. Mar.Poll.Bull..4(10):153-7 1973 Williams, R. and A.V. Holden, organochlorine residues from plankton* Mar.Pollut.Bull, 1973 4(7):lO9-11 56@ - INVESTIGATION OF ACUTE POLLUTION PROBLEMS AFFECTING FISHERIES IN ESTUARIES AND COASTAL WATERS by J.S. ALABASTER Water Pollution Research Laboratory Stevenage, U.K. Contents Page 1. Introduction 56 20 Lethal effects 56 2.1 Effects of mixtures of poisons 57 2.2 Relation between long-and short-term effects 58 of poisons 2.3 Fluctuations in water quality 58 30 Sub-lethal responses 59 4. Response of artificial ecosystems 60 5. Field observations 60 5.1 Empirical relationships between water quality 61 and fisheries 62 6. Summary'and conclusions 62 7. References 62 le IETRODUCTION Since adverse effects ofwater pollution on fisheries have been studied extensively in the freshwater environment, approaches that have been made to fishery problems there will be discussed for their relevance in identifying and investigating some of the most acute and urgent problems affecting fisheries in estuaries and coastal waters. Questions relating to general ecology and to the open sea will not be raised. The main pollution problem for freshwater fisheries in the United Kingdom stems from the combined effect on rivers of sewage and industrial wastes in lowering the concentration of dissolved oxygen and raising that of poisons, principally metals (copper and zinc), cyanides, ammonia and phenols. It will be shown that the short-term lethal effects of these conditions on trout can be reasonably well defined from laboratory studies. The question remains of predicting the long-term effects on survival and on other responses under the fluctuating conditions characteristic of polluted streams and of determining their ecological importance not only for trout but also for a variety of coarse fisheries, so that realistic water quality criteria for fisheries can be established. Such criteria can then be used in the management of polluted waters. The logical long-term approach, alongside appropriate field studies, would be to develop more elaborate laboratory experiments culminating in work an simulated polluted rivers and partially-controlled natural systems and indeed some work of this kind has already been done, but an alternative interim approach is to seek empirical relations between short-term esti- mates of toxicity to fish and the presence or absence of fisheries in polluted rivers. 2* LETHAL EFFECTS To define maximum permissible levels of pollutants implies that the causes of pollution should have been identified and that their adverse effects on relevant components of the ecosystem should have been assessedo Clearly, any practical approach to the problem must be limited, at least initially, not only to those factors that have the greatest impact on 57 fisheries but also to those effects that are most relevant and most easily measured both in the field and in the laboratory. In the United Kingdom the approa'ch during the last decade or so has been to concentrate much of the work on fish themselves, mainly through laboratory studies to assess under various environmental conditions the direct toxicity of the most commonly occurring poisons* The assessment is generally made in terms of concentrations at which times of survival are expected to be relatively long (several months) and at which only a negligibly small proportion of a population of fish would be killed after such a long period of continuous exposure. 2.1 Effect of fixtures of poisons 2.1.1 Pure substances Much of the published information on the'toxicity.oE substances to fish relates to materials tested singly, whereas generally in polluted rivers, and frequently in effluents that are discharged to them,-mixtures of poisons are present. A few early workers, for example Southgate (1932), showed that it was possible to determine the eEEect-of a mixture of similar poisons in water by summation.oE their individual toxic fractions, and others have suggested that synergism may be important, e.g. with heavy metals (DoudoroEE, 1952), but in the last decade a considerable amount, of more detailed and longer-term work has been carried out. In general it has been shown that.the c6mbined action .of several poisons is approximately additive; if, for example, the concentrations of two poisons that kill 50 percent of the fish in a few days (LC50) are At and Zt respectively, then the concentration of each (As and Z@ re uired in a mixture to given the same -toxicity is such that the sum of the fractions A@At + Z@Zt equals unity. This relation has been found to hold with rainbow trout (Salmo gaird,neri for (i) mixtures of zinc and copper in both hard and soft water (Lloyd, 19TT7, (1`1@ -ammonia and phenol (Herbert, 1962); (iii) ammonia and zinc in both hard and soft water, and also in the presence of reduced dissolved oxyg 'en (Herbert and Shurben, 1964); (iv) copper and ammonia, and zinc and phenol, though there is a tendency to over-pr6dict the toxicity of these two mixtures the lower the percentage response chosen (Herbert and Vandyke ! 1964); (v) copper and.phenol (Brown and DaltQn, 1970); and (vi) phenol and cyanide,in saline waters. at up to 20 percent sea water, though at 50 to 70 percent sea water the observed 48-h LC50 was 1.7 times, the predicted value,(Ministry of Technology, 1969). 14ixtures of three substances,have been tested together, including (i) phenol, zinc and ammonia, in which there was a tendency to over-predict the toxicity where zinc predominated (Brown et al., 1969), and equitoxicconcentrAtions of (ii) phenol, zinc and copper, and (iii) rirc_l@_e_l, zinc and copper (Brown and Dalton, 1970), with which there was good agreement .between predicted and observed results. While the@bulk of this work indicates that satisfactory prediction of the toxicity of simple mixtures of the commonly-occurring poisons can be made, there' is no theoretical basis for this and no reason to suppose it would be true for other substances.or other fish, or for effects other than.survival. On the contrary,,circumstances can be envisaged when it would not, for example when complexes are Eormed-between metals and cyanide; certainly there are examples of unpredictable toxicity 92 complex mixtures, particularly pesticide formulations (Alabaster,, 1.969). 2.1.2 Sewage effluent The predictive method has been used to estimate the relative importance of the commonly- occurring poisons (ammonia, monohydric phenols, zinc, copper and free cyanide) in sewage eSkluents in industrial areas in the United Kingdom, samples being taken'from a variety of disposal works covering a rangq.of treatment-processes and receiving various types of indu- strial wastes (Lloyd and Jordan, 1963, 1.964);-the toxicity of the effluents to rainbow trout 'was measured under controlled conditions and predicted on the assumption that the toxicity of the mixture of all given poisons could be calculated from chemical,analyses in a manner similar to @,that used in laboratory studies. In .13 out of 18. toxic. effluents, 'the predicted 58 toxicities were within 30 percent of the observed values, and 6 effluents were correctly predicted to be non-toxic; only 2 effluents were more toxic than predicted, probably because of the presence of other unidentified poisons. 2.103 Field data The same empirical approach has been applied in field studies of fishless rivers in the Iddlands (Herbert, jordan and Lloyd, 1965), taking into account the effect of environmental factors on the toxicity of individual poisons in the manner recently summarized by Drown @1 960) , and in general there was reaso.-iable agreem ent between predicted and observed toxicity. More recent worh has been summarized by Brown et al. (1970), in which it was shown that there __Tt@y_by'about 35 percent in polluted freshwaters was ax, overall tendency to under-predict toxici and by about 65 percent in polluted saline waters. Thus, all the evidence available suggests that much of the short-term toxicity of the commonly-occurring poisons can be satisfactorily predicted, at least for rainbow trout in freshwaters. 2.2 Relation between long- and short-term effects of poisons the five common poisons, ammonia, phenol, cyanide, copper and zinc, and mixtures of them, the relation between period of survival of trout and concentration of poison is such that a good indication of the median threshold concentration is obtained within a few days. other metals, however, (cadmium, chromum, and nickel), are much more toxic to trout after long periods of exposure than the results of short-term tests might suggest; for these it would not be appropriate to use 48-h LC50 values alongside those for copper and zinc when seeking correlation of toxicity with long-term effects on fish populations but to use the best estimate of the median threshold concentration. 2.3 Fluctuations in water qu,@lity Most laboratory studies have been carried out with continuous exposure of the fish to constant environmental factors, whereas conditions in rivers are continually changing. This makes it virtually inpossible to predict the survival of fish under field conditions. With some poisons, e.g. cyanide, exposure of the Ash for a short time to a concentration that eventually would have proved lethal, has no apparent adverse effect after the fish are transferred to clean water, and also has no effect on their resistance to subsequent exposure to the poison. Such recovery of the fish is not found with all substances. With DDT and an organo-phosphorus pesticide (Abram, 1967), and an effluent containing pesticide residues (Alabaster and Abram, 1965), under conditions in which short (6-h) periods of exposure to the poison alternated with similar periods in clean water, exposure time is approximately additive, the survival time of the fish being approximately equal to the time they would have survived in the mean co-.acentration had it been applied continuously. Further work with ammonia, zinc and equitoxic mixtures of the two (Brown, Jordan and Tiller, 1969) has also shown that under conditions in which the concentrations alternated between 1.5 and 0.5 times the 46-h LC50 at intervals of a few hours, survival of trout was. close to that of fish kept continuously at the 48-h LC50- More recent tests (ministry of Technology, 1971) have been made with two poisons, phenol and cyanide, to which trout were exposed for alternate 4-h periods, under continuous-flow conditions, one poison gradually replacing the other, and then over the next period being itself replacedp each having a final concentration equivalent to the 48-h LC50. The results showed that the mixtures were only slightly less toxic than predicted, assuming no recovery of the fish from cyanide, and that the toxicity of the system was reasonably well described by the average conditions. 59--- With othe@ poisons, e.g. the herbicide Reglone kAlabaster and Abram, 1965),and cadmium (Ministry of Technology, 1969), fish which had been exposed to potentially lethal concen- trations for times far shorter than those at which they would have been killed -under conditions of continuous exposure subsequently died after being returned to clean water; for these substances attempts to predict survival from average concentrations would, therefore, tend to under-estir-Late toxicity. Long-terra exposure of fish to sub-lethal conce 'ntrations of some substances has also been shown to affect their subsequent resistance to poisoning. Trout exposed ' to ammonia for several months became more resistant to lethal concentrations of ammonia and- less resistant to lethal concentrat ions 'of zinc (ministry of Technology, 1968). Somewhat similar results have been reported for, rudd exposed to sub-lethal concentrations of zinc; they developed aaa increased resistance to this poison (11inistry of Technology, 1966). On the other hand, trout, after exposure to sub-lethal concentrations of zinc were of similar sensitivity to a detergent (an alkyl benzene sulphonate) as control fish that had not been exposed to zinc, but were slightly more sensitive than the control to a mixture of detergent and zinc (Brown,* Mi@rovic and Stark, 1968). These 'examples serve to illustrate some of the possible effects on direct toxicity of i'luctuations in chemical conditions-in polluted rivers. More information is required on both the nature of variations in quality that occur in rivers and of their effects on the survival of fish of '11 kinds before reliable prediction of toxicity can be made. 3. SUB-LETHAL RESPONSES Generally, experiments with whole animals are mbre likely to yield ecologically signi- ficant results than those involving only organs or tissues, though the latter may be more useful in the.long term in providing a basis for 'understanding the effects of pollutants in ecosystems* It is important, however, that, when in vitro tests are conducted, concentrations of pollutant should be used.that are similar to tkiose likely-to be found in the whole animal when exposed to the pollutant under -field conditions. An example of a potentially useful test' relates to the blocking of cheilloreception in crustacea which could reduce the foraging success and also possibly interfere with the ability of ria16 lobsters to detect the female pheramones and.thus find a mate (IDOE, 1972). Another example of changes in behaviour induced by environmental stress and appearing to have ecological significance is provided by the reaction of the shrimp (Crangon vulgaris) to reduced dissolved oxygen (DO) (Huddart and Arthur, 1971). Normally a proportion of a population swims by pleopod action, rising from the bottom at a shallow angle and then, swimming actively downwards, return to the bottom to burrow where it is sandy. In the presence of lower DO a much larger proportion is Eree-swirrming, rising from the bottom at a steep angle and then sinking passively; with further reduction in Do swift upward escape reactions occur, involving rapid abdominal flexure with the uropods extended, followed by passive sinking. Standing individuals appear to be "on tiptoe" and are unresponsive to stimuli until some time after the restoration of hilgh levels of Do. The consequence of this behaviour in an estuary in which a zone of partly deoxygenated water oscillates with the tide would be for the shrimp to be carried by the water current away from hypoxic conditions. This suggestion is reinforced by the results of a mathematical simulation of the Thames Estuary embracing background swimming, stimulated activity in the presence of reduced DO and exhaustion after 30 minutes of increased activity; these results simulate the distribution and the large rapid reductions in number of shrimp which have been observed@in part of the estuarys On the other hand, many experiments, such as those .(Deubler and Posner, 1963) with post- larvae of the American flounder (Paralichthys lethostigma), while they demonstrate the ability of the fish to detect and avoid water in which the DO has fallen to about 3 mg/1 in a few minutes, give practically no indication of how they would react to the more gradual temporal and spatial changes likely to be encountered in the wild. 60 Other reactions caused by reducing DO and possibly having survival value for the species would include schooling behaviour which, for the northern anchovy (Engraulis mordax)v is altered when DO is slowly lowered to less than 2 mg/l (Moss and McFarland , 1970) In all these cases there is a need to conduct experiments on as large a scale as poss ible, to test conclusions using mathematical simulation and to confirmresults in the Eield whenever possible. 4. RESPOUSE OF P-nZIFICIAL ECOSYSTEMS One of the most relevant parameters to study is growth of the species because, .in conjunction with the biomass it determines production; this, in turn, is related to yieldt which is particularly important for commercial species. Growth of a species may be affected by environmental variables directly or indirectly (through effects on its food) or both, phenomena which have lent the.,.riselves to elegant elucidation in artificial ecosystems (warrem and Doudoroff, 1971; @,.Tarren and Davis, 1971), comprising riffle sections of fresh- water streams containing salmonids and sculpin (Cottus perplexus) in the presence of kraft mill effluent, dieldrin, or low DO. Empirical relationships have been established between the growth rate of fish and their biomass and therefore between fish production and their biomass, between fish growth rate and food density, and between food density and fish biomass; i.e. fish production is shown to be depedent upon the biomass of both fish and their prey. Such relationships are parti cularly valuable in making possible a detailed examination of the functions of parts of an ecosystem and yet circumventing the need to rely upon a deep understanding of all the mechanisr,,-,sm involved,, which, although an ultimate goal,is a very distant one even for the riffle system. Another type of laboratory estuarine microecosystem, (Copeland et al., 19i2) designed. to the hydrological characteristics of Trinity Bay, Texas, and con-CainIng representative samples of @he natural sediment showed how reduction in freshwater flow reduced net primary production (ratio of photosynthesis to respiration) in all but the seawa@ segment, and whereas under normal flow numbers of zooplankton were not related to photosynthesis because they were utilizing imported organic nutrients, under drought conditions they were entirely dependent upon photosynthesis. Thus, a large import of polluted freshwater bringing nutrients might offset to some extent, the disadvantages of its'toxicity. At the same time, by virtue of its low salinity and its contents of poisonst it would have an impact on the survival of larval fish and oysters and their predators (Walne, 1972). It has been stressed in the context of cycling.of. metals in estuaries Cdilfe and Ricej 1972) that meaningful evaluation. of the ecological stresses imposed by wastes would entail detailed knowledge of the inter- actions involved, which would require considerable research to quantify the reservoirs and the routes and rates of flux; in the meantime, however, it might be more rewarding to concentrate attention on identifying empirical interrelationships, such at Warren and Doudoroff (1971) and co-workers have done for production in their systems. It is of interest to note that, as migpt be expected from the experiments of Warren and Doudoroff'(1971), Jensen and Snekvik (1973) have described an increase in growth of trout populations followed by a decrease in abundance and eventual extinction in streams in 1..7hich the pH has fallen steadily over a number of years. _5.@ FIELD OBSERVATIONS Laboratory tests, even those using artificial ecosystemst although of help in under- standing some of the complex effects of pollution under natural conditions, are no substitute for observations and experiments in the field. An example of a potentially valuable field experiment is the sub-lethal dosage, with DDT, of tagged salmon smolts and their subsequent release alongside untreated tagged fish; untreated fish showed a markedly better return from the sea as grilse the following year than the treated individuals (Fisheries Research Board 61 of Canada, 1971). This result is much more meaningful than those of laboratory tests on the effect of DDT on learning ability and behaviour of the fish in temperature gradients. 5.1 Empirical relationships between water 5uality and fisheries 5 .1 .1 Freshwater 'Intensive studies of catchment areas have been initiated (Alabaster et al.,'@9/2) in close collaboration with the River Authorities which have provided data on the chemical quality of the water at a large number of sampling stations and also made an assessment of the status of the fisheries at each'. For each sample the concentration of each of the poisons ammonia, monohydric phenols, cyanide, copper, zinc and the-other toxic metals (nickel, chromium and cadmium) - were expressed as the fractional 48-h LC50 to rainbow trout, taking into account the effects on.toxicity of environmental factors, incl uding temperature, dissolved-oxygen concentration, pH value and water alkalinity. These fractions were summed for each sample taken and the probability distribution of the summed proportions.of the 48-h LC50 were calculated for each sampling station. When this was first done in 1968 for stations on the River Trent system, the distributions fell into -two fairlydistinct groups, one associ- ated with the presence of fish and the other with their absence; the coordinates of the approximate boundary di stributi on of 48-h LC50 between fishless and fish-supporting waters are given in Table I. Table I Coordinates of approximate boundary distribution of 48-h LC50 betweenfishless and-fi.sh"suppgrting waters Per cent@probability 1 5 10 25 50 75 -60 9@ 99 Sum o,f proportions 0.07 0.10 .'0.13 0.18 1.28 0.41 0.60 0.73 1.1 of 48-h LC50 Somewhat similar results have been-f6uhd subsequently for several other river systems. By calculating for each station separate distributions'for toxicity attributable to one or more poisons, the contribution of each to the totalcan be easily demonstrated. It underlines.the importance of both dissolved oxygen and of metals in determining toxicity aver the range where conditions appear. to be marginal for fisheries (around 0'.28 of the 48-h LC50) and the added importance of cyanide and. other poisons where they are worst. Whether or not the empirical relation found in these studies have any validity else7...rhere is not known. It is quite possible that, in attempting to sum small fractions of the 48-h LC50 of several poisons, the toxicity of some is underestimated, for example, because o2 un- known long-term effects, while that of other is overestimated, for example, because of adaptation by the fish. The resulting errors might tend to-cancel each other and the same could happen with the relationship derived between toxicity to trout and the presence or absence of another species where the relat 'ive sensitivity of two species may be reversed for two different poisons. For this reason alone it would be surprising if the relationships already found were universally applicable. Clearly, it would be better to *make direct measurements'of toxicity instead of relying on predictions and to use species appropriate to the.fishery of interest instead of trout; much more intensive population studies would also be desirable. 62 5.2.2 Saline water Doubts have bee n expressed about the possibility of detecting the effects of pollution on marine -fish stocks by examining commercial catches in view of the large fluctuations brought about by climatic factors and fishing pressure. There is a reasonably adequate theoretical framework for studying, particularly, the exploitation of stocks of some commercial fish, there being good estimates available of recruitment, growth and natural mortality, but estimates are needed for many more species (Gulland, 1971). In particular, more should'be discovered about events during the first few weeks or months of life, particularly as the young stages may be the most vulnerable to pollution. It might then be possible to carry out meaningful mathematical simulations of the effects of mortality, perhaps on the lines"already attempted by Waller et al. (1971) for the fathead minnow (Pimephales promelas) 1evertheless, problems are perhaps best identified, as well as being best studied, "Luad6r field conditions, particularly in estuaries, coastal waters and seas having compara- tively little exchan .ge with the open ocean. Special effort ought to be paid to these regions, in view of their importance for shell-fisheries and as nursery and feeding grounds for fish, and also the need to understand the most acute problems first. Fruh et al. (1972) showed that the diversity of benthic organisms, zooplankton, and fi s were related to e@-timated toy:icity in San Francisco Day and Galveston Bay, but further wor is required to identify the main pollutants concerned. The progress already made in that direction in the freshwater environment has depended upon the assumption that the effect of' sub-lethal concentrations of five common.poisons (ammonia, phenols, cyanide, copper and zinc) is additive when they are present together in a river; this is not necessa- rily valid for all poisons in all circumstances, especially when very low concentrations are present. For the mummichog (Fund@ulus heteroclitus), for example, there appears to be s-yi-iergism between copper and zinc and between these and cadmium (Eisler and Gardner, 1973) and for polychaetes, antagonism between phosphate, nitrate and silicate (Reish, 1970). 6e SUDITARY AND CONCLUSIONS Freshwater environment studies of river systems have revealed empirical relations ,,between water quality and the presence or absence of fish., By arbitrarily expressing water quality as the fractional 48-h LC50 of known poisons it is possible to identify the comparative importance in a given situation of different poisons and environmental factors. When a single water quality characteristic predominates, an empirical relation between its temporal distribution and the status of a fish population can be used in formulating water quality criteria, though this also requires data on lethal and sub-lethal ecologically significant adverse effects on fish and other aquatic organisms, including reduced growth and fecundityt avoidance of low concentrations of pollutants and other alterations in normal behaviour and physiology. These aspects have so far received relatively little attention, partly because of the difficulty in demonstrating their relevance and partly because they would involve the use of considerable resources that might otherwise be deployed on existing lines of enquiry. 7e RSFERENCES Abram, F.S.H., The definition and measurement of fish toxicity thresholds. Adv.Water 1967 Pollut.Res., 3(l)-.75-95 Alabaster, J.S., Survival of fish in 164 herbicides, insecticides, fungicides, wetting 1969 agents and miscellaneous substances. Int.Pest Control, Mar./Apr. issue:29-35 63 Alabaster, J.S. and F.S.H. Abram, Development and use of a direct method of evaluating 1965 toxicity to fish. Adv.Water Pollut.Res., 2(l):41-60 Alabaster, J.S. et al., -An approach to the problem of pollution and fisheries.SYMP.Zool. 1972 Soc.Lond., (29) :87-114 Brown, V.I.-, The calculation of the acute toxicity of mixtures of poisons to rainbow trout. 1968 Water Res.,4(2) :723-33 Brown, V.I-'. and R.A. Dalton, The acute lethal toxicity to rainbow trout of mixtures of 1970 copper, phenol, zinc and nickel. J.Fish Biol.,2:211-6 Brown, V.M., D.H.M. Jordan and B.A. Tiller, The acute toxicity to rainbow tout of 1969 fluctuating concentrations and mixtures of ammonia, phenol and zinc. J.Fish Biol., 1:1-9 Brown, V.I. V.V. Mitrovic and G.T.C. Stark, Effects of chronic exposure to zinc on toxicity 1966 of a mixture of detergent and zinc.. Water Res.4(2):255-63 Brown, D.G. Shurben and D. Shaw, Studies on water quality and the absence of fish 1970 from some polluted English streams. Water Res, 6(4):363 p Copeland, B.J., H.T. Odum and D.C. Cooper, Water quantity for preservation, of estuarine 1972 ecology. In Conflicts in water resource planning, I-later Resource Symposium NO. 5, edited by E.F. Gloyna and W.S. Butcher. Austin, Texas. University of Texas, Centre for Research in Water Resources Deubler, E.E. and G.S. Posner, Response of post-larval flounders (Paralichthys lethostigma) 1964 to water of low oxygen concentrations. Copeia, 1963(2):312 Doudorft, P., Some recent developments in the stvdy of toxic industrial wastes. Proc. 1952 Conf.Ind.Waste Pacif.NW, 4:21 Eisler, H. G.R. Gardner, Acute toxicity to an estuarine teleost of mixtures of 1973 cadmium, copper. and zinc salts. J.Fish Biol., 5(2):131-42 Fisherit Search Board of Canada., Review of the Fisheries Research Board of Canada 1969- 1971 1970. Ottawa, Canada, Information Canada, 217 P. Fruh, .E. Armstrong and B.J. Copeland, Water quality for estuarine ecological 1972 stability. In Conflicts in water resource planning, Water Resources Symposium edited by E.F. Gloyna and W.S. Butcher. Austin, Texas. University of. Texas, Centre for Research in Water Resources Gulland, J.A., Ecological aspects of fishery research. Adv.Ecol Res., 7:115-76 1971 Herbert, D.W.M., The toxicity to rainbow trout of spent stillliquors from the distillation 1962 of coal. Ann.Appl.Biol., 50:755-77 Herbert, D.W.M. and D.S. Shurben, The toxicity to fish of mixtures of poisons. 1. Salts 1964 of ammonia and zinc. Ann.Appl.Biol., 53:33-41 Herbert, D.W.M. and J.M. Vandyke, The toxicity of fish of mixtures of poisons. 2. Copper- 1964 ammonia and zinc-phenol mixtures. Ann.Appl.Biol. 53:415-21 Herbert, D.W.M., D.H.M.,Jordan and R. Lloyd, A study of some fishless rivers in the 1965 industrial Midlands. J.Proc.Inst.Sewage Purif., 1965:569-82 64- Huddart, R. and D.R. Arthur, Shrimp in relation to oxygen depletion and its ecological 1971 significance in a polluted estuary. Environ.Pollut., (2):13-5 IDOE. Baqeline studies of pollutants in the marine environment and research recommendations. 1972 International Decade of Ocean Exploration (IDOE) Baseline Conference, May 24-26, 1972. La Jolla, California, Scripps Institution, 54 p. Jensen, K.W. and E. Snekvik, Low pH levels wipe out*salmon and trout populations in southern- 1973 most Horway. Ambio, 1(6);223-5 Lloyd, R., The toxicity of mixtures of zinc and copper sulphates to rainbow trout (LaL]Lmo 1961 gairdneri Richard,son). Ann.Appl-Biol., 49:535-8 Llo@ffi,.R.@ and D.H.M. Jordan, Predicted and observed toxicities of several sewage effluents 1963 to rainbow trout. J.Proc.Inst.Sewage Purif., 1963:167-73 Predicted and observed toxicities of several sewage effluents to rainbow 1964 trout: a further study. J.Proc.Inst.Sewage Purif., 1964:3-6 Ministry of Technology, Water pollution research, 1965. London, H*M, Stationery Office 1966 Water pollution research, 1967. London, H.M. Stationery Office 1968 V Water pollution research, 1968. London, H.M. Stationery Office is-69 Water pollution research, 1970. London, H.M. Stationery Office 1971 Moss, S.A. and W.N. McFarland, The influence of@ dissolved oxygen and carbon dioxide on fish 19@O shoaling behaviour. Mar.Biol., 5(2):100-7 Reish, D.J., The effect of varying concentrations of nutrients, chlorinity and dissolved 1970 oxygen on polychaetous annelids. Water Res., 6(4)1721-35 Southgate, B.A., The to)dcity of mixtures of poisons. Q.J.Pharm.Pharmacol., 5:639-48 19,3 2 Waller, W.J. et al., A computer simulation of the effects of superimposed mortality due to 1971 i-oll7ut-ants on populations of fathead minnows (Pimephales promelas). J.Fish.Reso Board Can., 28(8):1107-12 Walne, P.R., The importance of estuaries to comme rcial fisheries. In The estuarine environ- 1972 . -ment, edited by R.S.K. Barnes and J. Green. London, App@Ued Science Publications Ltd, Warren, C.E. and G.E. Davis, Laboratory stream research: objectives, possibilities and 1971 certainties. Ann.Rev.Ecol.System., 2:111-44 Warren, C.E. and P. Doudoroff, Biology and water pollution control. London, W.B. Saunders 1971 Co., Philadelphia, 434 p. Wolfe, D.A. and T.R. Rice, Cycling of elements in estuaries. Fish.Bull.NOAA/NWS, 70(3): 1972 959-72 TOXICITY TESTING AT KRISTINEBERG MARINE BIOLOGY STATION by 14. SWEDMARK GRANMO and S. KOLLBERG Kristineberg Marine Biology Station Fiskebtlckskil, Sweden Contents Page 1. Introduction 65, 20 Methods and equipment 66 3. Choice of test animals 4. Biological functions studied '69 .5. Data presentation 69 6. Some aspects',of bioassay work @73 7. Reference Is 74 1. INTRODUCTION Increasing coastal water pollution caused,by the establishment of industry and in6reas- to exercise ing population at the coasts has made it essential for@the authorities concerned control of the mari -ne'environment and to decide on limits for permissible quantities of .pollutants in the sea. Such measures demand knowledge about the reactions.of marine organi- to different pollutants.and. the ecological consequences of their presence in the marine environment. For these reasons bioassay work has been done at the Kristineberg..Marine Biology Station,-situated at the Gullmar Fjord on the west coast of Sweden, for several years with marine animals as test organisms. Polluting components in coa:stal waters are studied, and their effects on the animals determined. The studies are particularly concerned with the toxicity of surface '-.active agents, frequently present in polluted.coastal waters as they are active ingredients in detergents and oil dispersants. Recently, the effects of heavy metals have been,included in theinvestigations. The work is supported financially by the National Swedish Environment Protection Board and the Royal Swedish Academy of Sciences to which the station belongso' The toxicology studied at.Kristineberg can be characterized as an ecological toxicology because it is.aimed not only at the necessary determination of the acutetoxicity of different contaminants but. also at a prediction of possible ecological consequences of their presence in the marine environmento For these purposes, a great variety of marine animals areused, belonging to different taxonomic groups, such as fish, crustaceans and bivalves, each represented by two or more species from different habitats and with different modes of life. .Both adult animals and developmental stages are studied as sensitivity to toxicants varies considerably during the life cycle, the early phases generally being the most sensitive. Short@-term tests are carried out over 96 hours-.to det6rinine 'acute toxicity and lethal effects, and long-term tests, running for several months, are used to observe chronic effects of sublethal concentrations,' corresponding to those which,may occur in coastal waters. During the tests, mortality and survival timis are-recorded continuously. A particular interest is taken in the observations of the effects.of various biological functions which may be of ecological importance to the animals, such as behaviouri activity patterns, 66 defence reaction, response to food, etc. These functions serve as early indication-. of effect, but may differ greatly in sensitivity. Even small changes in such reactions may hav@ fatal consequences for the animals when part of the ecosystems. Effects on growth and development are also studied in the early phases of life. Attempts are made to understand the general principles for the toxicity of the naterials tested, by studying their action on physiological functions such as reS7,iration and osmoregulation as well as their accumulation in tissues and organs. 2* METHODS AED EQUIPY211T The biotest technique used at Kri.-tineberg is based on the continuous flow principle. Continuous flow aquaria systems are generally considered to facilitate the maintel--a-rce of more natural conditions in the physico-chemical environment of the test tanks, i.e. of such factors as temperature, salinity, pH,oxygen content, NH3-content, bacterial level, etc. Aquaria systems with running sea water are possible at Kristineberg as the. station has a constant supply of clean sea water, being pumped into the laboratory from a -10 m de@;th of the Gullmar Fjord. At this depth the salinity is fairly constant, betwee-ri 32-3401'o@-.), and the oxygen content is high, 80-90% saturation. The incoming sea water has a terperature variation throughout the year from 4,to 160C but with no rapid fluctuations. In t:-ds -region ,there are no tides. The biotest equipment (Fig. 1) is principally designed as described by Swedmark et al. (1971). Details are given on p.75 It consists of 3 levels of aquaria with a varying 'number (4-13) of test tanks at each level, of which one is a control tank. The mixing of the standard solution of the test material and the water takes place in each vertical section via its respective funnel. To increase the mixing efficiency, the funnels are equipped with glaff beads on coarse nylon mesh. In order to keep the concentrations of the tested material constant, the standard solution is added to the test aquarium by means of precision dosage pumps (electric piston pumps or microperistaltic pumps). The runninc i sea water is distributed by siphons, valves or by a special dosage apparatus (Granmo and Kollberg, 1972). All tubing is of Tygon or Silicon. Rectangular'or cylindrical Perspex plastic aquaria are used. Their volume varies between 10 and 60 1. according to the size of the test animals. Cylindrical-vessels are always used for testing eggs and larvae. @'@encrally the number of adult animals in each tank is 10-20, but sometimes less of practical reasons. The equipment described here permits simultaneous testing of several species in a nu;@iber of different concentrations at uniform environmental conditions and yields a great quantity of data within the test time. It has a simple design and has been shown to be of good reliabi- lity. The standard solution of soluble material (surface@active agents, some oil dispersants, heavy metals) is made by dilution with distilled water. While testing insoluble dispersants and oils, the concentrated, undiluted products are used. These latter materials have proved to be very destructive to the tubing which raises some problems. The emulsions (mixtures) of different oils and oil dispersants are made up by means of a stirrer in the standard solution tank. In a tight emulsion the droplets are smaller and more homogeneous in size than in a coarse dispersion. For light fractions of refined oil, such as marine diesel oil, the proportions of dispersant and oil were 1:9, to give a tight emulsion. For heavier fractions as fuel oil or crude oil, the corresponding proportions were changed to 1:1 to give a similar emulsion. By the mixing methods described, surface active agents, dispersants, as well as oil emulsions, were well mixed with sea water. The dispersions formed, generally kept their homogeneity during the passage through the aquaria, with the execption of crude oil, which caused some practical difficulties as a tight emulsion was not obtained. In the tests,tfie oil floated on the surface of the wateri or appeared as droplets in a coarse dispersion. 67 ea-water inlet ,:,Piphons /PIN Dosing Standard A A pump solution of surfactant 0 0 0 0/ 0 IMixing' tanks Control A C. Test tanks j L 11 j I Outlet Fig.-1. Equipment for toxicity testing.of marine organisms according to the continuous- Blow principle* (Swedmark et al.-, 1971), G@ r U (Reprinted with permission from Springer Verlag) 68 However, this was considered to correspond quite well to the conditions in water of low turbulence and without tides. For this reason, no other mixing method was.-tried. Concentrations from. 1 to 1000 ppm (m.9/1) are used as standard concentrations. As a rule, each product is tested in duplicate series. Substances are tested both pure and in mixtures for the study of joint eEfects. The sea water flow is generally kept at 0.6 1/min and there are only small variations in the flow (less than 1%). Regular control of the oxygen content shows that this flow is enough to give a satisfactory oxygenation (More than 70%) in all tanks. According to our experi.ence, short-term testing is best run at temperatures which do not differ from the normal sea water temperature of the season, as the animals are then exposed to less extra stress. In our region, 60C corresponds to the mean winter temperature at 40 m, 110C to spring and autumn temperature and 160C to the mean summer temperature. However, when necessary, the temperature of 'the test tanks is kept approximately constant with an immersion heater.or a cooler equipment* As the susceptibility of t 'he animals has been shown to vary with temperature and season, probably in connexion with reproduction periods, it is preferable to make comparative short- term testing during the same season. Short-term exposure always lasts 96 hours and is followed by a recovery period of at least 48 hours in clean sea water. The recovery period is particularly important when testing bivalves as they often manifest-a delayed mortality. Temperature, salinity and dosage-rates are checked daily. The oxygen content is checked once per short-term test, or once a week in long-term tests* Mortality and-effects on bio- logical functions are recorded several times a day during short-term testing. 3* CHOICE OF TEST ANIMALS A variety of marine fish, bivalves and decapod crustaceans are used as test-organisms* They all live in the littoral zone (0-40 m.) and are caught in the Gullmar Fjord and nearby waters. Thetest animals.have been chosen according to the following criteria: .(a) They should represent different taxonomic groups and also different biological types with regard to locomotion (mobile or sedentary),and nutrition (Eilter-feeders - carnivorous); (b) they should inclu de species of economic import ance; (c) they should include species representing different trophic leveis; (d) they should be susceptible and the biological fLmctions.studiied should be easy to define and observe. The choice may be limited for practical reasons, such as availability in the region and ease of collection and adaptability to laboratory conditions. Our standard test species among fish are cod (Gadus morrhua), three-spined stickleback (Gasterosteus aculeatus) and flounder (Pleuronectes flesus , two free-swimming species and one bottom-dwelling. A@ong decapod crustaceans, prawn Leander adspersus), brawn shrimp (Crang crangcn) and shore crab (-Carcinus maenas),@i.e. two more mobile and one less mobile species. Sometimes hermit crabs (Eupagurus bg@r_diciirdus) and spider crabs (MMs araneus) have also been used. The following bivalves have been chosent the scallop (Pecten opercuiaris) a swinmdng species, the cockle (Cardi edule) which is a superficially burrowing type, the clam (Mya arenaria , a deep-burrowing Rrid relatively stationary species, and the common mussei @Myt lus edulis) which represents a sessile form. 69 The animals are acclimatized to laboratory conditions for at least one week before testing to retain normal behaviour and physiology. If fish and crustaceans are kept for longer periods, they are fed up to.48 hours prior to, but not during, a short-term test. Even if the animals are fed, they may.lose their condition in the store tanks, and thus ,newly caught animals are generally preferred for-the testing. Exceptions are made'.1for species with vertical seasonal migration, such as prawns and crabs, which are caught in shallow water at the end of the summer and kept in store tanks during the winter. In long- term testing all the speciesincluding bivalves, are regularly fed. Considering the variation of susceptibility during the life cycle, it is preferable to select test animals of the same size in.order to obtain comparable results.. For the study of the effects on early p hases of the life cycle, including fertilization, -eggs and larvae of cod, common mussels and lobsters have been used. The.effects on the sensitive phases of the moulting cycle in decapods are studied on prawns. 4. BIOLOGICAL FUNCTIONS STUDIED To determine mortality different criteria for death., appropriate for each organism, are used. For fish and decapod crustaceans, the-cessation of respiratory and body move- .ments are used and for bivalves the cessation of function of the,adductor muscles and the complete inactivation.of the muscles of the mantle edge. The planktonic eggs of cod and larvae of cod, lobsterand',mus'sels are considered dead when'laying immobile on the bottom of the tank. To determine sublethal and chronic effects of toxic materials, the following biological ftmction@s of,adult animals are studied: locomotory behaviour of mobile species, respiratory movements of fish, response to food by fish and decapods, shell-closure ability of bivalves, burrowing and siphon retraction of cockles and clams and byssal activity of common mussels. The,eEfects on hatching rates and on the development of eggs and larvae are observed. The effects of surface active agents, including oil dispersants, and heavy metals on the functions studied appear according to certain patterns in the following sequence: (a) increasing activity (avoidance 'reactions); (b) decreasing activity (successively impaired reactions);, (c) immobilization and death. 5. DATA PRESENTATION The results-obtained-on acute toxicity and lethal effects of the toxic materials for each species are presented in.diagrams, which express the relationship between concentration and survival time. From these toxicity curves the median lethal concentration or LC50, i.e. the concentration that kills 50% 'of a sample at any exposure time, is interpolated and can be tabulated. The 96 h W.90 values are the standard form used for the comparative ranking .of the toxic substances. Due to the different susceptibility of the test species, a ranking list is established for each species. The procedures of plotting toxicity curves and calculating LC50 values are given in detail on p . Examples of such curves areshown in Figs 2-and 3, illustrating the toxicity of some oil.dispersants to cod and common mussels respectively - one asensitive species, ,the'other a-very resistant one. The,exposure time is 96 h, followed by 48 h in clean sea- water.- Symbols with an arrow indicate that median mortality is not obtained. The distri- bution of the curves in the diagram shows the different toxicity of the tested material. It is evident-from Figs 2 that the.tox.icity.to cod generally decreases rapidly with decreasing concentrations. Ano .ther type of toxicity curve is obtained for common mussels (Fig. 3). 7o - w 100 El 01 D 10 z 4 w 10 100 1000 CONCENTRATION, PPM Fig. 2. Toxicity to cod of some surfactants and oil dispersants 1000 100 M 10 z I I I 11IJ 10 1000 CONCENTRATION, PPM 6@ Fig. 3. Toxicity to common mussels of some surfactants and oil disp@ersants 71 The toxicity decreases much slower with decreasing concentrations and most of the mortality occurs after exposure. This is characteristic of bivalves which stresses the importance of including a period of clean sea water in the test. The 96-h LC50 values for each product are tabulated as shown in Tables I and II, where the LC50 values obtained.after 48 h in clean sea water are also includede Due to the different ability of 'recovery for different species and the often delayed mortality, particularly of bivalves, these values are often lower than the 96-h LC50 values. With prolonged exposure time there is also a decrease in the LC50 values, i.e. the toxicity increases with time. The variation obtained in the survival times and in LC50 values, probably due to temperature, season and individual condition, has been statistically treated and the 95% confidence limits calculated. For variation in survival times, Litchfield's method (1949) has been used, and for variation in LC50 values that of Litchfield and Wilcoxon (1949). Mortality in controls has been, and should be, rare. When'this occurs, the percentage mortalityin the test series has been calculated after taking into consideration the control mortality as described by Tattersfield and Morris (1924). The sublethal effects on the biological functions studied are expressed as the lowest' median concentrations of,the toxic material that affect activity or reaction patterns. They are called effective reaction concentrations for 50% of the sample, i.e. EC50 values, after the proposition,by Sprague (1969). In Table III an example of EC50 values is given for different phases of the activity pattern observed on the swimming activity of cod, and shell7' closure ability and byssus thread formation of common mussels. From the knowledge of such concentrations for different biological functions, important for the survival of the animals from various ecological habitats,,predictions of the possible ecological consequences of pollution in the natural.environments can be made. Table I Median lethal concentrations (LC50) of some sutfactants and oil d.ispersants for cod Tested Material LC50 (ppm) 96 h 96 + 48 h Sur2actants: Linear dodecylbenzenesulphonate (anionic) 1.6 1.0 Nonylphenol ethoxylate 10 (nonionic) 6 5 Tallow alcohol ethoxylate 10 (nonionic) 0*5 0-,5 Oil disp_ersants: Fina-Sol SC 30 30 EP 1100 120 80 Corexit 7664 130 Fina-Sol OSR-2 180 130 BF 1100.X 700 700- Corexit 8666 950 950 72 Table II Median lethal concentrations (LC50) of some surfactants and oil dispersants for common mussels LC50 (PPm) Tested Material 96 h 96 + 48 h Sur.factants: Linear dodecylbenzenesulphonate (anionic) 100 25 Nonylphenol ethoxylate 10 (nonionic) 12 10 Tallow alcohol ethoxylate 1.0 (nonionic) 50 15 Oil dispersants: Fina-Sol SC. 110 90 BF 1100 1000 250 Corexit 7664 1000 Fina-Sol OSR-2 700 700 BP 1100 X 700 700 Corexit 8666 1000 1000 t .Table III Lowest median concentrations of.some surfactants and oil dispersants affecting locomotory behaviour of cod, shell-closure and byssus activity of common Ymssels at 96 h exposure Concentration (ppm.) Cod Common Mussels Tested Material Decreasing Decreasing Increased Decreased shell- byssus activity activity closure activity ability Surfactants: Linear-dodecylbenzene- Oe5 00 sulphanate (anionic) Nonylphenol ethoxylate 2 4 10 5 10 (nonionic) Tallow alcohol Oo5 0*5 ethoxylate 10 (nonionic) Dispersants: Fina-Sol SC 26 100 12 13P 1100 40 100 500 40 Corexit 7664 50 100 1000 500 Fina-Sol OSR-2 140 700 110 BF'1100 X 300 700 140 Corexit 8666 950 950 950 65 73 6. SOME ASPECTS OF BIOASS,AY WORK For the necessary information needed by industry and government bodies as to the relative toxicities,of new products, routine testing must be done, the purpose of which is to produce conparative rankings of,toxic-materials in standard form* A usef)Alstandard form,is the median lethal concentration, theld 50 value. It must be borne in mind that.the LC50 values vary considerablyiaccording to the test species-used, the testing technique '(continuous flow or static aquaria . systems) and,ekposure time (24,148,. 96 h or more).. Temperature, salinity, seasonF'individual@'donditio@-i as well as the'stage of, the life cycle of the animals also influence the values. They should, therefore, be*consi- dered as an estimation of toxicity and, instead of being arranged in a close rank order, the toxic materials are preferably grouped in categories-with respect to their LC50 values,, for example as follows: 96-h LC50: _below J.ppm (mj/1) very toxic 1- 100 ppm toxic 100- 1000 ppm moderately toxic 1000-10000 ppm Slightly-toxic The choice of the test species is of gre@t importance as there is a great variation in susceptibility-b@etween'taSconomic groups and between species related generzilly"to the mode of lifei more active species being more sensitive, as a rule, than less active ones; neither is the susceptibility of a species always the sameto different mat Ierials.. It is, therefore, preferable to usemord.than one species. If the results of short-term tests are translated to the field situation at pollution,. they may correspond to the acute effects of accidental pollution, often of short duration due to,quick countermeasures and rapid di:lutiono- Predictions,of ecological consequences of pollution in aquatic (marine) environments are made possible by givi 'ng the bioassay work a biological design* Thus,' several species should be used representing different modes of life and activity, different habitats and different taxonomic groups. The lethal toxicity is less essential than the sublethal and chronic effects on biological functions which are important for the..survival of the animals in their'natural environment. Such effects should be studied at low concentrations and preferably in long-term tests. The functions examined can be locomotory behaviour, defence reactions or food response of adult animals and development and.'fiatching rates of' eggs and larvaeo The lowest median concentrations which affect activity or reaction, patterns, i.e. median effective concentrations (EC50), should then be determinedo The initial effects an the behaviour of an animal are generally an increased activitye The biological si-gnificance of this reaction is avoidance, ioee flight reactions in mobile species or other protective mech@aAsms in sedentary species such as shell-closure of bivalves* Protective mechanisms seem to be especially well,developed in littoral species, naturally adapted to large variationsof their environment*. These reactions increase the chances of survival of individuals at shorter exposure to pollution and may explain the often mild effects of accidentai,pollution of limited sizeo At a longer.expqsure, the inpairment of biological functions that follows the period of avoidance, implies a danger for the individuals as all reactions locomotion, response to food, defence, shell-@cl6sure, byssal activity, etco - become weakero If this happens in.the natural habitat, where competition for food and space is important and where the *animals are 'part of a prey/predatory system, the*most resistant species are favoured, while the more sensi- tive species find their chances of survival diminished. This exemplifies how a pollutant present in a concentration that is sublethal for an individual isolated in an aquarium, becomes lethal in an ecosystem, and must be borne in.mind in the interpretation of laboratory results when applied to.the field situations. It also points to the importance of the determination of the ecological threshold concentrations of toxicity of the pollutants 74 introduced into the sea. Knowledge of such concentrations is necessary to determine effec- tive limits for permissible concentrations of pollutants in controlled marine environments. The difference in resistance of species and the successive impairment of biological functions at prolonged exposure may explain why a constant pollution, even in low concen- trations, is more serious for the marine fauna and flora than a limited acute pollution. It is well known from many polluted areas (Reish, 1970) that chronic pollution changes the composition of the marine communities and that the biological balance becomes disturbed. An ecological succession of animal and floral communities takes place, which starts with decreasing diversity as the susceptible species disappear and continues with an increasing number of resistant species (indicator species), ending with the replacement of macro- scopic organisms by microbese 7- REFERENCES Granmo, R. and S.O. Kollberg, A new simple water flow system for accute continous flow 1972 tests. Water Res., 6:1597-9 Litchfield, J.T., Jre, A method for rapid graphic solution of time-percent effect curves. 1949 J.Pharmacol.Exp.Ther., 97:399-408 Litchfield, J.T. Jr. and F. Wilcoxon, A simplified method of evaluating dose-efEect 1949 experiments. J.Pharmacol.Exp.Ther,., 96:99-113 Reish, D.J., A critical review of the use of marine invertebrates as indicators of varying 1970 degrees of marine pollution. In Marine pollution and sea life, edited by M. Ruivo. West Byfleet, Surrey, Tishing News (Books) Ltd., pp. 203-7 Sprague, J.D., Measurement of pollution toxicity to fish.. 1. Bioassay methods for acute 1969 toxicity. Water Res., 3:793-821 Swedmark, M., A Granmo and S. Kollberg, Effects of oil dispersants and oil emulsions an 1973 marine animals. I-later Res., 7:1649-72 Swedmark, 11. et al., Biological effects of surface active agents on marine animals. Mar. 1971 Diol., 9:183-201 TattersEield, F. and H@M. Morri 's, An apparatus for testing the toxic values of contact 1924 insecticides under controlled conditions. Bull.Entomol.Res., 14:223-33 75 PROCEDURES FOR TOXICITY TESTING IN A CONTINUOUS FLOW SYSTEM by I. BERG and R- GRANMO Kristineberg Marine Biology Station Fiskebackskil, Sweden Contents Page 1. Equipment 75 2. Preparation procedures 75 3. Testing 77 4. Calculations 78 4.1 Mortality (lethality) 78 4.2 Biological effects '80 5. References 80 EQUIPMENT At the Kristineberg MarineDiology Station the apparatus used for toxicity testing is, based on the continuous flow principles (Fig. 1). Sea water is introduced at the top (A), and distributed by the water flow system (B), described in detail by Granmo and Kollberg (1972). The water is dosed to the Eunnels (C) through nozzles (D) in the water flow system.- The test solution is distributed from special stock aquaria (G), and is dosed with the aid of a pump (H) to the glass funnels via dosage'tubes (1). In the funnels the water and the test material are mixed with the help of glass beads (E). The-mixed solution is then distributed in the aquaria, which are placed at several levels After passage through the equipment, the'waste water passes a flocculation unit beEore,discharge into the recipient. The tubes used must be of A resistant and non-toxic mat erial, as many compounds are corrosive. It is important to ensure that the circulation in the differeent aquaria is as good as possible - since non-homogeneous dispersed compounds will give incorrect results. The volume of the aquaria should be chosen according to the size and number of the animals. General rules cannot be given for the Volumes, but the American Public Health Association biotest recommendation (1965) gives 1 1/g of animals/day as a sufficient now, or a minimum flow rate which equals the volume of water in the tank in 6 hours. Best results are obtained if the test unit is placed where external disturbances are low, as the test animals, especially fish, are easily stressed, which often results in increased sensitivity. Cylindrical vessels are preferred for fish to give them better possibility of swimming. 2o PREPARATION PROCEDURES To attain the desired water flow from the dosage apparatus, choose nozzles with an appropriate diameter. For more careful adjustments, screw the threaded tube downward or upward depending on whether more or less water is wanted (Granmo and Kollberg, 1972). A simple way to.measure the water flow is to use a graduated glass and a stop-watch. For the dosage of the standard solution, two -types of pumps are used, a piston pump and a microperistaltic pump. With the piston pump it is possible to regulate the amount of 76 B B A -fl C D, E G F F Fig. 1. Equipment for toxicityztesting of marine organisms according to the continous-flow system (Swedmark et al., 1973) TR-e@p`ri@ted with permission from Pergamon Press Ltd.) 77 standard solution by adjusting the pump. With the microperistaltic pump there is no possibility to correct the volume dosed. With this type of pump the standard solution is distributed through special tubes by the pressure of a roller. By choosing different diameters of the tubes,. it is possible to obtain the concentration desired. When joining the tubes from the-pump with those fromithe standard solution and the dosage tubes,. connecting glass tubes are used. To protect to tubes against friction damages, grease is rubbed now andAhen onto the roller in the microperistaltic pump. ,The following procedure is used to,calculate the concentration'.of the standard solution wanted: a 9/1 b 1000 or, if the tested material is not 100% of.the compound added a b 10 g/l where a = water flow, mi/Min b = dose amount, ra./min 1 . concentration wanted, ppm, m percent.toxic material When the concentration of.the standard solution is determined, the different concentrations wanted in the test aquaria are easily obtained from the same expression. ..Soluble material is dissolved in distilled water to prevent biodegradation, while- insoluble products are dissolved in,other, preferably low-toxic, solvents or the indiluted material is used. To get anidea of the materials' toxicity.before starting.the .test, and to facilitate the choice of the most suitable concentration range, a pre-test could be made. This may last for 24 hours with only a few concentrations within a wide range. According to the aim of the testing, one species or a variety of species of test animals are chosen. Before testing, the animals must be acclimatized't6 laboratory conditionsq which usually takes.3 to 7 days. Furthermore, it is better to keep the animals in the test tanks for one day before an experiment. From a statistical point of view, it is preferable to use as many animals as possible in the aquaria, but,.as mentioned'above, the number may be limited for practical reasons. The distribution of the test animals in the aquaria ought to be done at random using, for example, a table of random numbers. 3-. TESTING After the start, the dosage rates of standard solution to each aquarium-are checked* For a careful control, 2-5 drops are collected from the dosage*tube and the time interval for the drbps to fall is noted. The drops-are then weighed and from this infdrmation the concentration can be calculated. B 6 D A 100 C pp' 'where A . number of seconds between each drop B = the weight of one dropAn mg C. water flow in mi/Min D . the concentration of the standard solution, in ppm, using a pure material 78 Dosage, as well as temperature and salinity, are checked daily, oxygen content and pH weekly. Observations on the animals. It is important to use parameters which are easy to observe and define. For example, in our laboratory, the following parameters besides death are used: Fish: Breathing (opercular movement per minute), coughing frequency, swimming behaviour as phases in activity pattern (increased or impaired reactions), or equilibrium disturbances, i.e. side to side rolling and swimming in a vertical position (nose up, tail daown), shoaling, etc. Bivalvesj clams: Shell closure ability and mantle-edge activity and, for common mussels, also byssal activity. Crustaceans: Swimming activity and locomotory behaviour, food response. After death, according to the aim of the study, lengths and weights of the animals may be measured, tissues and organs examined and samples taken and fixed for histological examination. It is important that the animals are observed often, i.e. several times a day in short- term testing. Fixed observation times are preferable, as it is easier afterward to compare the different test results. Observations and calculations are more easily made if data .charts, ready to fill up, are prepared before the test. After 96 h exposure the pump is stopped and the test continues another 48 h in clean sea water. 4. CALCULATIONS 4.1 Mortality (lethality) Note the times from the start of the experiment will death occurs for the individuals 'in each'tank. Then, a percentage/time diagram is established as in Fig. 2. The time scale .on the abscissa-axis can as well be a linear one, but in most cases it is preferable to use a logarithmic scale. in the diagram the time to death (survival time) is plotted in a cumulative way for each test concentration. When drawing the curves the points form a straight line (Fig. 2a) or a sigmoid one'(Fig. 2b). Mortality may occur in the control tanks, and is then taken into account by using, for example, the expression of Tatterfield and Morris (1924): X (Y - Z) 100 100 z where X . calculated lethality in the concentration, in percent Y = real lethality in the concentration, %in percent Z = lethality in the control, in percent When the mortality curves for all the concentrations are established, the time (for each concentration) at which 50 percent of the animals are dead (LT50) can be read off* Knowing this, the time/concentration relationship, i.e. the toxicity curve, can be drawn as in Fig. 3. Plot on a double loga:iithmic diagram your LT50 values for each concentration and draw the toxicity curvee To"obtain the 96-h IC50 value, check 6n the curve the concen- tration for 96-h exposure. Statistical treatment of the LC50 values obtained is preferable. The methods of Litchfield and Wilcoxom (1949) or Finney (1952) can be used. 79 100 80 60 0 40. 20 a SURVIVAL TIME,h Fig. 2. Curves showing the cumulative mortality in test tanks Uj Soo z Uj 10 1 2 .5 10 CONCENTRATION, PPM Fig. 3. Example of toxicity curve in a double logarithmic diagram. (Symbol.with an arrow indicates.that no median mortality was obtained) 80 4.2 Biological effects Fish: Make tables of diagrams for coughing frequency, opercular movements, swimming activity, etc. Bivalves; Byssal activity and shell closure ability can be noted as in Table I. Crustaceans: Make tables or diagram for swimming activity, etc. Table I Number of common mussels with the ability to form new byssus threads after 96-h exposure and subsequent immersion in clean sea water (10 test animals) Exposure time Concent .ration (ppm) Control in hours 450 320 174 115 60 22 3/10 4/10 9110@ 9/10 8/19 9/10 43 0/10 5/10 6/10 8/10 6/10 9/10 65 0/10 2/10 2/10 6/10 8/10. 9/10 93 0/10 1/10 .3/10 5/10 7/10 10/10 117 4/10 6/10 10/10 10/10 10/10 10/10 137 6/9 9110 10/10 10/10 9/10 10/10 5e . REFERENCES American Public Health Association, Standard methods for the examination of water and 1965 waste water including bottom sediments and sludges. New York, American Public Health Association, 769 p. Finney, D.J., Probit analysis: a statistical.treatment of the sigmoid response curve. 1952 London, Cambridge University Press, 318 p. Granmo, A. and S.O. Kollberg, A new simple water flow.system for accurate continuous flow 1972 tests. Water Res., 6:1597-9 Litchfield, J.T. and F. Wilcoxon, A simplified method of evaluating dose-effect experiments. 1949 J*PharMaColeEXp*Ther*, 96:99-113 Tatterfield, F. and H.M. Morris, An 'apparatus for testing the toxic values of contact 1924 insecticides under controlled conditions. Bull.ZntomolqRes., 14:223-33 81 BIOASSAY METHODS USED BY THE RESEARCH LABORATORY OF THE NATIONAL SWEDISH ENVIRONMENT PROTECTION BOARD by ,T.B. @HASSELROT.- National Environment Protection Board Research Laboratory Drottningholm,,Sweden Contents Page 14 Laboratory investigations. 81 2. Field' investigations .81 ..2.1, Equipment 82 Test animals 82 2.3 P rocedure 82 2.4 Results- @82, -.3. Investigations at industries., and.. 89 waste water'treatment plants 3.1. Industry 89' 3.2 Waste water:treatment plants 90, 4.. References 91. 1. -LABORATORY INVESTIGATIONS The purp9se. of these investigations is to test@the toxic effect.of new chemical compounds and,.to a certain extent also, to check different.types of polluted water from industri'es and.communities. The technique-for acute toxicity studies is rather simple, using glass aquaria and fish, fingerlings of Salmo salar-,or Salmo gairdneri. The investi- gationsare usually carried out as short-terr@ -testC, -aver five da@s. The result is expressed as the median lethal concentration, LC 50 values. The physico-chemical qualities of the water are registered continuously or controlled with repeated sampling (temperature,,pH, specific conductivity, oxygen content, etc.).', The-effect of toxic compounds on thehatching of fish eggs and the early development Of young fish is studied oncichlids, mostlyof the species.Cichlasoma biocellatum or -Aeguidens-latifrons (Hasselrot, 1956). These species-,have been shown to be "normallyll sensitive Fo -toxicants and they Produce 500-1000 transparent eggs at each spawning. The eggs are.deposited on plastic plates which are then transferred to small glass aquaria containing 100 ml of a test solution. The larvae hatch,after 3- .4 days at a water tempera- turi of 25-260C. The effect of the tested compound can be observed at the different stages of development from egg to free-swimming fish.: The avoidance reaction of fish, mostly young salmon, in the presence of a water pollutant is.studied in a "fluviarium" (H15glund, 1961). In this apparatus a fish is able to choose between different concentrations of a:,tested compound the fish is sensitive to 2* FIELD INVESTIGATIONS To study the effect of water Pollution on fish, caged specimens have.been stationed in lakes,vater courses and coastal waters (Hasselrot, 1964, 1965).. The purpose has been to follow the effect of a known discharge or to trace an.unknown one at fixed positions in a recipient over a certain time. 82 2.1 Equipment Two types of round cages have been used. Both are covered with nylon netting without knots (to avoid -injury to'the fish by friction) and fitted with rings, mostly of stainless steel. One type of cage has rings 35 cm in diameter, the other 45 cm diameter. The cages are kept in position by means of sinkers, lines and plastic floats. 2.2 Test animals Fish. For studies of the toxicological effects and the direct intake of compounds by the fish from the surTounding waters, one year old,salmon fry are used in fresh waters. In .coastal waters, from the area of bresund, young rainbow trout have been used instead, because of the higher salinity. To study smell and taste-impairing substances, and also intake of metals, older specimens of both species have been used. Salmon and rainbow trout are used generally because of the convenience in having a uniform fish material from piscicultures all year round. Other test organisms. Since 1972 bivalves have also been used for uptake studies in cages: Anodonta cygnea in fresh waters and Mytilus edulis and Macoma baltica in coastal waters. To study the concentration of compounds in epiphytes, the laboratory uses a wooden construction consisting of a strip with plates in horizontal and vertical position, connected with a cord to the surface cage at each station. 2.3 Procedure First, the distribution in the recipient of the effluent from a known source of pollution. is determined. The caged fish are then placed in positions estimated to be most representa- tive for the effect,of the effluent on the recipient. In each cage of the small type, 30-40 one-year old fry are placed, and, in the lar4er type, 10-15 two-year old fish. The duration of the investigation var-ies from five days (for studies on toxicity, taste and smell impair- ing effects) to one month (for studies on the accumulation of mercury and other heavy metals). The condition of the fish during exposition time is observed and, if possible, various tests are made to check any physico-chemical alteration's in the environment of the cages. At the end of the investigation time the fishes are frozen and taken to the laboratory for analysis of accumulated substances, or to the National Swedish Institute of Public Health for an odour and taste test (organoleptic test). The scale for this test ranges from 1 to 5, where 1 means a fish with a very bad smell and taste, and 5 is an unaffected fish. Bivalves and samples of epiphytes are also sent to the laboratory for analysis of accumulated substances. 2.4 Results The reports which follow are given to provide a more concrete presentation of the applicability of the cage investigations as a means of following known sources of water pollution or of tracing an unknown one. 2.4.1 Known sources Kraft pulp mills. -Frtsvifors Bruk, FrOvis This mill had an annual production of 55 000 t of 90 percent pulp, which was used exclusively for paper making. Steps taken by the'mill to prevent water pollution included a well developed fibre recovery system and a scrubber tower for collecting the noxious and evil-smelling sulphuric substances in the digester and evapo- rator condensates. Before the outlet the waste water passes a big dam. The factory is 83 0 0 situated at a river, Barsan, with a discharge of 3-30 m3/s. The river ends in a rather shallow lake, Varingen, which.is about 7 km in length. An investigation of the odour.of the water before the exposure of fish in net cages in March'196@ showed that the sulphate waste water was concentrated in the part of the lake between the mouth of the receiving river and the outlet river of the lake (Fig. 1). The waste water was also more concentrated in the deeper'layers of the lake. There was a-total lack of oxygen'below a water depth of 4 m. Only in the upper water layer was the oxygen content sufficient for fish life.. The caged fishes were therefore placed only at 1 m levels. The mortality of the exposed salmon@fry was total by the end of the exposure time, from Station 2 to Station 5, 6 km downstream from the outlet of the factory (Fig. 2). Also the organoleptical test (made on perch) gave-very bad values for the same row of stations and the values were low also at Stations 11 and 12 in the outlet river at a distance of about 10 km from the factory.* In the other part of the lake, outside the current of the polluting river (Stations 6-10)no@effect was shown on the fish. Ai canbe seen in Fig. 2 the concentration of .sulphate waste water in the northern part can also be traced in the raised values in the physico-chemical analyses of the water.(Co'lour, permanganate consumption, conductivity, contents of sulphate and sodium). The mortality among the salmon fry at Stations 2-5 may especially be explained by the presence of recin acids in the water. 0 Lbvholmens bruk, Pitea. The annual production of this mill in 1965 was 150 000 tons 90 percent pulp. Steps taken by the mill to prevent water pollution were similar to those taken above. The waste water was led into a bay of the Baltic by a tube 1.6 km in length. This investigation was also carried out during late wintert when the influence of the waste water on the surrounding water is greatest because for a long time the water has been isolated by ice and snow from the oxidation effect of the air. An investigation before the exposure of the fish cages showed that the water in the Baltic had an odour from the waste water even at a distance of 9 km from the mouth of the outlet tube (Station 9b) (Fig. 3). The waste water spread mainly to the south and the odour of the water was usually.,strongest at a depth of 3-4 m. In the sound between the two islands opposite to the tube, the waste water was concentrated to the surface layers. i The test fish was one-year old (2-4 g) to two-year old (30-40 g) salmon. Mortality (100%) occurred only at Station 2 at a depth of 4 m. The organoleptical test on the fish showed low, bad values not only at Stations 2 and 3 near the,tube mouth, but also at Stations 4 and 9 and'as Bar south as Station 10, at a distance of 12 km from the tube mouth. The lowest values in both odour and taste were registered for Rish from cages at a depth of 3 m. The values obtained from physico-chemical analyses of the environment around the cages can neither be placed in relationship'to the effects produced on the fish, nor used to detect the presence of,sulphate waste water. It can be established that the use of direct cage tests in recipients can give a reliable tonception of the effects on the fish of discharging waste water from a certain sulphate mill. The cage tests can also serve as a very sensitive method to indicate the presence of sulphate waste water. In this respect the test results from sea investigations suggest these methods are far superior to the'physico-chomical examination techniques. in current use. Mercury-discharging industries. Outlets of waste water containing organic or inorganic mercury compounds have caused significant increases of the mercury content in the exposed fish. This has been the case also when no increase of mercury has been observed either in water or in bottom sediment. An example of the latter are the cage investigations in the 84 12 )c x x x X x x x 1. 2 3 5 km Fig* lo Lake VUringen with sampling stations downstream from Frbvifors bruk 85 Mortality of salmon 100 in Taste and odor-of 5 0 odor perch 4 taste 3 A FW A r]L,,-.,, 2 LM LA L LJ Temperature C Oxygen mg/ I 12" 6- PH 4 Color mg Pt /I 120- 60- KMn04 mg/l 120- 60- Sutphide mg H 2S 0,2, Suiphate mg /I 40- Recin acids mg /I Ei. conductiviiy 'PS 60- Sodium mg Stations I T 1 2 3 4 S 6 7 8 9 10 11 12 Fig. 2. Results of cage investigations downstream f.rom.Frtvifors bruk in March 1962* 86 river Gbta 31v downstream from the wood pulp mill in Gbta (Fig. 4) In the autumn of 1967, the waste water from this plant caused an increase of mercury in Q river fish to 830 n9/9 or 8.8 times the basic value in the hatchery after the month of exposure. The mercury content in water and in sediment of the same station showed no significant increase. The investigations with cages in the river Gbta X1v are a good example of how,*by using this 3 sensitive technique in a water course with a very great discharge (mean discharge = 500 m /s), one can observe, the presence and the strength of a discharge of mercury, check the effects of taken measures, etc. The series of analysed values from the nearest station upstream, the chlorine alkali plant EKA in Bohus, down'to the station of Alelyckan (the intake of water for the city of Gothenburg) is a revealing picture of the extension of mercury emission to the water from. this industry. As,a result of measures taken at,the plant,'the concentration in the caged fish decreased progressively from 1967 to 1970- Exposure of fish in cages has been done every year as a routine check, and that is how a sudden, increased mercury emission in 1971 and 1972 was observed (Fig. 4). New efforts'have been made to reduce this registered discharge of mercury. The highest increase of mercury in liver of one-year salmon exposed for one month, 56 times the basic value in the fish hatchery, is noted downstream from a bi wood-pulp mill situated at the river Ume Ilv, with a mean water discharge of 200-300 mg /S. 2.4.2 Unknown sources In the beginning of February 1962, an abnormally high mortality in young salmon was observed in a fish hatchery supplied with water from the.river Dalalven. At this spot the river has a mean water discharge of 215 m3/s during winter and 350 m3/s for the whole year. It was established that the mortality was caused by a marked anaemia of the fish. Hemato- crit values of 5-10 percent (normally about 40 percent) were observed in these fish. Investigations on fish caught in the riv'er-itself from March to May showed that the anaemia was limited to the river downstream from Avesta, a highly industrialized town 26 km upstream from this fish hatchery. Even in similar piscicultural establishments at-the mouth of the river, 100 km downstream from Avesta, fish taken within a month's interval demonstrated very low hematocrit values. The anaemia gradually ceased at the end of April and the beginning of May. The primary cause of the fish anaemia could not be disclosed before the water temper- ature again dropped in late autumn. Net cages with young salmon were then put into the river at many places to ascertain the origin of the anaemia causing agent, which proved to be waste water from a yellow phosphorous producing factory in Avesta. Aquaria tests with waste water from a settling tank in this factory, containing about O.f mg P/1, gave within two days a descent in hematocrit values in young salmon'from 50 to 11 percent. Still lower concentrations gave low hematocrit valueswithin two weeks. The water temperature was only 1-20C. When the temperature was raised to 1ODC, the hematocrit values increased again. Other aquaria tests with yellow phosphorous in colloidal form added to water, partly from the river DalMlven upstream from Avesta, partly from the Lake Mlaren, gave the same decreasing effect on the hematocrit values of the test fish. In such a way it has finally been proved that no other agent than yellow phosphorous in the water from the river Dalalven was responsible for the fish anaemia. In the river DalMlven the effective concentration of yellow phosphorous at low water temperatures must.have been very small, probably far below ppb. @We can thus concludethat the use of caged fish was the only way to solve the problem* The physico-chemical methods in curTent use for water analyses were not specific and sensitive enough to be of any use. 87 Q@O OD 0 1 2 km sewage tube ED G) CD ra 0 G Fig. 3. Sampling stations in the.archipeiago o@xtside L?3vholmens bruk 'D 88 mg Hq kq STAI 11 F1 015 LAKE V4NERN ST Al.U . . . STAT. 13 TROLLHATTAN Ito GOTA (wood-pulp mill) :C) STAT. 14 13 BOHUS 015 14 (chlorine alkali plant) I ALELYCKAN ST AT. 15 F GOTHENBURG SKAGERACX %5 STAT. 16 66 67 68 69 70 71 72 73 Fig. 4. Investigation area in C;Ota a1v since 1966. Diagrams showing concentration of mercury in test fish t indicates that test fish died) '89 After these surveys the factory in question started to treat the waste water more effectively, e.g. secondary precipitation with milk of lime. Since then'no abnormal hematocrit values have been found either in salmon in the fish hatcheries or in caged salmon in the river itself. For safety, there are still (1973) some control fish cages situated in the river just downstream from.the outlet of the factory* In' 1962 this dangerous effect of yollow,.phosphorous was quite unknown. The result of the investigation was first given in 1.963 inan official,report to the authorities .concemed, then in a short report to. 'an OECD'Conference in Paris in 1 .964...In 1970, the same effect of yellow phosphorous onmarine fish and crustaceans was reported from Placentia'Bay, Nevifoundland (Jangaard, 1970, Zitko et al 1970,-Anon, 1970). .2.4.3 Complementary investigations To check how-far out in the sea the,smell- and taste-impairing effect of,,for instance, sulphate waste water on free-!swimming fish could poSsib3 ly reach, some complementary invest- igations were done. In one'case about 5 000 specimens of one very important local commer- cial species, vendace, Coregonus albula Le, were tagged apd put into the,Baltic.near the mouth of the outlet tube from 'tW-kralt Pulp.mill,.Lbvholmens,bruk. The intention was to follow how far from the tube the polluted fish had swum in a fortnight, which was 30-40 km. A fortnight is a minimum time for a polluted fish to,expel.the bad smell and, taste from its flesh. 0 0 Kyrksjlon, belonging to the water system of the river Delangersan, is a lake in Sweden containing the most mercury contaminated fish. Here there has also been tagging of fish as a complement to the cage investigations in the lake. on recapture in 1:970, in 17 out of 400 tagged pike which had been moved from Dellensj1dn (an unpolluted-upstream lake) to Kyrksjbn the year before, very much increased mercury content'could be-analysed in muscle tissue: 1.3-6.3'mg/kg with a meanvalue of 2.7 mg/kg, compared to a basic value of about b-5 mg/kg in pike fr-om Dellensjbn (i pg/kq in *fish flesh, is the. blacklisting value in Sweden) However, more than a thousand salmon.fry that had been exposed in cages showed no difference in.mercury value between the stations in Kyrks'jbn and Dellensjon. As the cage investigations give a measure of the direct intake by the fish of mercury from the water environment, the food eaten by the wild fish was undoubtedly of decisive importance for the concentration of meicury in the fish (the industrial mercur-y discharge was stopped 'at*the turn of the year 1965-66)- 3. INVESTIGATIONS AT INDUSTRIES AND WASTE WATER.TREATMENT PLANTS 3*1 Industry. Investigations were mad:e in order to find out what types- of waste.water within a. factory caused the greatest effect on the smell and taste of the fish. 3.10 Methodology Containers with 30 test fish each (mostly two- were connected directly year old salmon to the waste water in question using different dilution degrees. The apparatus functioned according to the continuous water flow principle. Certain chemical compounds could be added to the apparatus, instead of the industriail waste water, in-6rder to find out the affect that compound had on the fish. Each container was cylindrical and. made of stainless steel, 70 cm high and 40 cm in diameter. In order to have good water circulation, the horizontal inflow was in the periphery and,the outflow in the cenire.of the.bottom. The waste water was pumped.first to a special container for'mixture andadjustment of the watero The time of the investigations was 5, 10,.15,and 20 days. 90 The organoleptical test on the fish was made by the test group at the National Swedish Institute of Public Health. 3.1.2 Results Lbv,holmens bn;Lk, 1966. To get rid of the smell- and taste-impairing effect on fish, the ipain discharge from the factory after purification had to be diluted about 1 000 times (see 2.4.1). The uptake of the smell- and taste-impairing compounds was greatest during the first five days. The additional accumulation during the following ten days was only slight. The elimination of these compounds from the-fish flesh was also greatest during the first five days. It took, however, a fortnight in clean water for a fish with low organoleptical test values to become satisfactory in this respect. Among the pure chemicals it can be said that methylmercaptan was lethal for the fish in a dilution of 1 : 5 000 000. In a dilution of 1 : 10 000 000 the fish remained alive without any influence on the smell and taste. Dimethyidisulphide (the'threshold dilution about 1 : 7 500 000) gave a much stronger, bad effect than diniethylmonosulphide on the smell and 'taste of the fish flesh. These results have been of great use for the future p@ans concerning the improvements in measures taken for the purposes of water protection at that factory. It can be mentioned that such container tests with fish at another sulphate mill, Gruv*dns Pappersbruk, 1963, have shown that the waste water, after finally Passing through an activated sludge tank, has practically no adverse effect on the smell and taste of the fish. At a cellulose factory, Fiskeby bruk, fish in containers have also been used conti- nuously since 1965 as control organisms against sudden appearances of poisonous compounds in the waste water. A biotest establishment has been in use in Sweden since 1971 at Astra, S75dertalje (Ht3glund, 1972). This industry makes pharmiacologica3 products. There have been inspections several times a day to study the condition of the fish. If this is bad, water samples are taken for analysis with an automatic sampler, which is connected to the same water that passes through the containers. This method has been successful in finding the sources of sudden toxic effects in the waste water. 0 At a mechanical pulp mill, Bowaters Svenska TrRmassefabriks AB, Umea, in 1966, and the chlorine alkali plant EKA AB, Bohus, in 1967, young salmon were also kept in containers (about 30 specimens in each), which for a month were continuously connected to the outlets of waste water of different dilations. The mercury content of +-he fish usually increased in the containers with waste water. This could most clearly be seen in the Bowater case, where a mixture of 1 percent industrial waste water and 99 percent water from the river . caused, in spite of the low winter water temperatures (0.5-0.80C), an increase of mercury in the liver to 460 ng/g, which is 2.4 times the basic-value in the fish hatchery. The result from the container test at EKA was a little more doubtful. The reason for this was the very uneven mercury content in the waste water. Mercury could, for example, suddenly occur as drops of metallic mercury, a form in which mercury cannot be taken up in fish. 3.2 Waste water treatment plants In Gothenburg's waste water treatment plant at Rya, a bioassay establishment for fish has just been set up to control the treated waste water before its release to the receiving body of water. This establishment has been arranged under the management of the Research Laboratory belonging to the National Swedish Environment Protection Board in Cooperation with hydrotechnical experts from the Water Treatment Works of Gothenburg. -91- The establishment consists of two parts.One part is intended for a study of the concentration in fish of heavy metals and biocides. This consists of four cylindrical container of the type described above for the control of outgoing industrial waste water but these containers are wider and lower. one container is filled with dechlorinated tap water, while the other three have different degrees of waste water dilution (1:1, 1:2 and 1:4) and are continuously connected with the main sewer. There are always in each container 31 two-year old 'rainbow trouts, which,after exposure for 14 days, are taken up and frozen for sample preparation. The other part of the establishment is an acute- toxic part where the trail fish's inability to swim in a water stream in the presence of a poisonous substance is registered. The water speed (5-15 cm/s) is adjusted by an electrical motor with a rotary propeller. The establishment is connected with the waste water indilution 1:1. An affected fish is carried backwards by the water flow in the apparatus and is finally registered by photocells at the back of the experiment container. To prevent accidental registration of a healthy fish,there is a. region of strong light which repulses influenced fish. For the same purpose the speed of the water flow is accelerated by a bottom plate, which rises towards the end of the experiment space. There are I also front photocells which produce an election flash when, the fish passes. The acute-toxic establishment has partly been made with models from West Germany, Switzerland and France (Juhnke and Besch, 1971; Zahner-Schneider, personal.communication; Leynaud- Barbier, personal communication). By this method, the presence of a poisonous substance can-be registered much earlier than with the currently used apparatus, which,is based- on the death of fish. as indicator, and action can be taken at an earlier stage to prevent fish death in the receiving body of water. 4. REFERENCES Jangaard, P.M., The role played by the Fisheries Research Board of Canada in the "red" 1970 herring phosphorous pollution crisis in Placentia Bay,Newfoundland. Circ.Off. Atl.Reg.Dir.Res.,(CAR.1):20p. Juhnke, I-and W.K. Besch,Eine neue Testmethode zur FrUherkennung akut toxischer 1971 Inhaltsstoffe im Wasser. Gewass-Abwass., 50/51:107-14 Hasselrot,T.B., En laboratoriefisk for giftforsok.Svensk Fiskeritidskr., 65(5):80-3 1956. . (in Swedish) Investigations with caged fish as an indication of pollution from kraft pulp. 1964 mills. Vattenhygien, 1964(2):72-84 A study of. remaining water pollution from a metal mine with caged fish as 1965 indicators. Vattenhygie ,1965(1):11-6 Hoglund, L.B., The reactions of fish in concentration gradients. Rep.Inst.Freshwat.Res. 1961 Drottningholm (43):1-147 Biotest for giftkontroll av avloppsvatten vid Astra.Astra.Farm.Rev., 1972 71:433-6 (in Swedish) Zitko, V. et al., Toxicity of yellow phosphorous to herring, Clupea harengus, Atlantic _1970 salmon, Salmo salar, lobster,Homarus americanust,and beach flea, Gammarus oceanicus. J.Fish.Res.Board Can., 27(1):21-9 Anon. , A, marine pollution case study.The "red" herrings of Placentia Bay. Fish.News 1970 Int.9(11):28-33 92 MONITORING OF AQUATIC -POLLUTION by G. TOMCZAK Fishery Resources Officer (Oceanography) Fishery Resources and Environment Division PAO, Rome Contents Page 1. Introduction 92 2. Concept of an integrated marine pollution monitoring programme 93 2.1 -The network of stations. 93 2.2 Elements of national programmes 93 2.3 Elements of regional programme 95 2.4 Elements of worldwide programme 96 3. Existing monitoring activities 96 3.1 National monitoring programmes 97 3.2 Regional and worldwide monitoring programmes 100 4. Pollutants 100 5. Outline for a monitoring programme in developing,countries 102 6. References 108 1. INTRODUCTION The conservation of natural resources, maintaining both the quality of the environment and its living resources, depends to a great extent on man's ability to manage them and to set up and enforce regulations in the best long-term interest. Unrestrained exploitation of .the living resources may yield-short-term benefits but in the long run is potentially disastrous This Course deals with pollution as it relates to the living aquatic resources, and with regard to these resources, the above statement has been proven true by.the fact that uncon- trolled fishery activities have led in some regions, to overfishing and to the reduction or even the lose of certain species. Regional fisheries bodies have been established to discuss scientific research and regulations needed to overcome such difficulties*- Similar to unrestrained exploitation, the =controlled release of waste through rivers and outlets from munioipal and industrial plants,. as well as through dumping from ships and aircraft, may lead'to ecological disturbances. In addition, it can be a risk to man's health through the ingestion of freshwater and marine food products. Therefore, monitoring systems have to be established aimed at defining the health of the aquatic environment and the level of contaminants in the various organisms especially, in species of commercial interest. Thor should also determine possible trends in order to develop, if necessary, a warning-system related to pre-test criteria or toleranoe.limit for critical pollutants. From the beginning,it should be pointed out, however that strong efforts are still needed in research to ensure that decisions about monitoring and standards, critera, and tolerance limits needed in this connection are made on a real scientific basis. Until -these data are available judgements, have often, to be mode on the basis of our present knowledge and from results of sometimes inadequate surveys made as investigations for siting of outfalls or dumping into the sea, making at the same time use of a good deal of common sense. 93 Monitoring, in such bases, has to be started with the specific goal to control the fate of the waste in the environment in order to act immediately in case the basic considerations for releasing the waste are proven wrong. 2. CONCEPT OF AN INTERGATED MARINE POLLUTION MONITORING PROGRAMME The concept for an integrated marine pollution monitoring system is represented in Fig.-I. It consists of four main partS: (a) The network of stations (b) Elements of national programmes (c) Elements. of regional programmes (d) Elements of worldwide programmes 2.1 The network of stations The network of stations listed on technology top of Fig.1 forms the basis for the national and regional monitoring programes. A relatively dense network of '*impact stations" has to be established in areas where the introduction of pollutants is considerable, such as estuaries,industrial oentres or concen- trations of population at the coastline with sewage discharge and waste water outlets,and especially in areas where the effects of pollutants are likely to be particularly harmful to living resources, eg., in major fishing areas in and near the coastal zone. On the contrary, in areas distant from the direct introduction of pollutants only a few "reference stations" may be needed to ensure that a rise in the degree of pollution in these areas in not overlooked and to show the trend for growing. pollution in due time so: that measures can be taken avoid further input of pollution. Ocean weather ships,mid-ocean island stations and fixed buoys in oceanic areas at high latitudes may Iserve as reference stations, Extensive and detailed monitoring of either natural or manmade accidental event e.g., . hurricane flood at tiands, big oil-well seepages or tanker. accidents,, should also beLiD organized as part of*national and regional monitoring programmes as they can provide valuable and unique information on.the response of natural systems to the strong impact of pollutants. 2.2 Elements of national programmes Research is one of the most important elements in connexion with monitoring programmes* This has been stressed by many working.parties, workshops,seminars, etc.which have been convened, especially during the last five years,e.g., in the U.S.A. , as part of the Inter- national Decade of Ocean Exploration (IDOE) or panels sponsored by the National Academy of sciences or the National Oceanic and Atmospheric Administration (NOAA)(IDOE, 1972; National. Academy of Scienes.1971; Goldberg, 1972) in the United Kingdom (Cole,1971) and the Federal Republic of Germany (Kinne and Aurich, 1968). On the international- level, the Joint Group of Experts on the Sciontific Aspects of Marine Pollution (GESAMP) and the Food and Agriculture Organization (PAO), supported by some, other international bodies (UNESCO,IAEA,SCOR, WAD) have mainly. promoted the discussion on research needed in connexion with monitoring marine pollution (INCO/FAO/UNESCO/WAD/WHO/IAEA/UN,1971;FAO,1971). The research on a national level has to 'be related to such problem as establishing "maximum permissible levels" (i.e., accepted level of a pollutant in an organism or in a population or resource to be protected from a specific risk)and "derived working levels" (ie,maximum acceptable level of a. pollutant in specified media designed to ensure that under specified circumstances & primary protection standard in not exceeded), and figures ,which would be needed for the definition of water quality criteria valid in the various 0 0 Z > 0-1 L m 0 > Z > m r., 0 Er 0 CA z CA 0 cl) cl) cm CL EL 0 @ 0 > z 000 cn m m > :0 c 00 0 0 0 m > - :) < < M m Z -4 CL ;;: F) 0 M > 0 M Z ;7 -1 > M 0 m rj) 0 E: -< > 0 m cn --j :() > -< 0 1 m 0 -n ID E; CD Co 0 CL i I M 0 C) G) - :v 0 > Z (n > > 0 :;, CD 0 3 = @ 0 m 176 95 coastal zones. However there are also research needs in connexion with the standardization of analytical methods for the main pollutants in fresh water and marine environments of the countries; laboratories should also make strong efforts aimed at monitoring biological para- motors. So far, monitoring in mostly directed to detecting physical and chemical changes in he natural environment with the underlying philosophy that by measuring parameters that are easy to measure one has taken a shortcut to measuring the health of the sea and its living resources and thus, biological changes which may be accompanied with them. In fact, such a presupposition needs more profound and detailed knowledge of the relation between environment and living organisms than we have. ,On the other hand, direct biological monitoring, e.g., through using indicator species, or making ecological surveys,Cannot be used until fundamental ecological research provides a firm basis so that disturbances due to pollution can definitely be distinguished from natural population fluctuations in various marine communities. Finally, services are included in Fig. 1 as an integral part of national programes. They will coordinate the monitoring efforts on a national level, act.as counterparts for other national services within the framework of regional monitoring systems, and provide facilities for storage and retrieval of the data an contaminants in the environment and in the.living aquatic resources. 2.3 Elements of regional Programmes Regional programmes are the second stop within a concept of an integrated marine pollution monitoring programme. In many oases, regional programmes are needed to complete the efforts made on the national level because pollutants introduced at the coastline of one country may be spread through currents,dispersion,and mixing processes into adjoining coastal waters or into the open sea where fishing grounds may be used IV various fishing fleets. Elements of theme programmes are given in Fig. 1. As far as research and services are concerned, the problem are similar to those already dealt with under national programmes. To understand what in meant by intercomparison and regional laboratories, the setting up of a regional programme for the North Sea may be taken as an example. The countries bordering this area undertook within the framework of the International LCouncil for the Exploration of the Bea, (ICES) - a one year baseline study to determine the level of contaminants in the living resources of commercial importance and after that time. they decided to continue to, coordinate their national monitoring programmes through ICES Problem which came up during their cooperation were the intercomparison of results and availability of laboratories. (ICES, 1974). Although all participating countries dispose of very well developed oceanographic and fisheries research institutes which should be able to perform the analyses needed in the monitoring programs, they had to limit the programme to the most important pollutants and to a small umber of samples due to the lack of personnel. Furthermore, results of the intercomparison - using standard samples for the various contaminants - showed that, in the beginning there were significant differences, and efforts had to be made to make the results of analyses comparable's More severe difficulties of this kind may arise if regional monitoring programmes are envisaged in developing countries as the lack of trained personnel and adequate instrumentation will be more grave. For that reason the. establishment of regional laboratories for aquatic pollution research and monitoring will be an important element for the execution of regional monitoring progrmmes Theme laboratories have to be well equipped with the sophisticated instrumentation needed and sufficiently staffed with wall trained scientists and laboratory assistants in-order to take responsibility for all the analyses which have to be made as part of the specific regional programmes. 96 2.4 Elements of worldwide promrammes Monitoring programmes with a worldwide scope will depend - as it is indicated in Fig. I on the development of national and regional programmes. Only with the experience gained in national and regional laboratories and with the'aid of their well educated and skilled person- nel will it bepossible to establish system on a global scale. Programmes of'this kind are envisaged by the Long-term and Expanded Programme of Oceanic Exploration and Research (LEPOR) and, as a major part of LEFOR, in particular by the Global Investigation on Pollution in the Marine Environment (GIPME). Both programmes are being pro- moted by the Intergovernmental Oceanographic Commission (IOC),, which acts as a special joint mechanism of various agenciesl including PAO# cooperating through the Inter-Seoretariat Committee on Scientific Programmes relating to Oceanograpby (ICSPH0). The machinery developed by IOC for oceanographic obsqrvationo*and environmental monitoring is the Integrated Global Ocean Station System (IGOSS)(IOC9 1971) (see Fig- I)- IGOSS has also ,been designated as the suitable frame for'establishing a worldwide marine pollution monitoring system. Plans have already been developed to start a pilot project aimed at monitoring petro- leum hydrocarbons, tar balls and oil slicks in the North Atlantic Ocean and adjacent seas, as well as along the main tanker route from the Persian Gulf through the Indian Ocean. and around the Gape of Good Hope to Europe. This pilot project became operational on 1 January 1975 and will continue for two years as approved by the General Assembly of IOC in Navemiber 1973. Monitoring will be based on observations and sampling made from voluntary observing ships, ocean weather ships, research vessels and suitable offshore platform, as well an on observations from aircraft and'satellites. Working groups will be established to define the methods for sampling and analyses, and to deal with standardized methods for reporting, dissemination, e=hange and archiving.of data resulting from the project. With a view to the quickly growing number of'data on pollutants in the aquatic environment and inorganisms which will result from monitoring programmes at all leveleg the arohiving of data in a suitable worldwide storage and retrieval system has to be organized very carefully. This is the task of the IOG Working Group on International Oc graphic Data Emohange. (IODE) in cooperation with the responsible world and national data ,contras belonging to the International.0ceanographic Data Fmchange System (IOC, 1967)- As part of the system the PAO Fishery Data Centre acts an a specialized permanent data oentre aimed at collecting and maintaining fisheries data, including data of contaminants . in living aquatic resources. An inquiry has already been started in order to get the infor- mation needed on institutes which, on,a routine basis or an part of special research pro- grammes, are analysing oontaminants,in aquatic organisms. As a first step, the Fishery Data Centre already compiled a directory of these institutions in order to facilitate the emohange of results and views between scientists working in this field. An a final atepq it is envisaged to publish an inventory of data on contaminants in living resources,, especially in commercially important fish, available at the relevant institutes in the world (FAO, 1974; 1976). 3. FMSTING NDNIMRUG ACTIVITIES Monitoring programmes have already been developed at all levelsp i.e.9 on a national, regional and worldwide basis* In principle, we have not only to take into consideration such programmes which are dealing directly with pollutants but we hAve to realize that monitoring enviromentalt i.e., pbysioalv chemical and biological parameters in required in order to estimate the rate of input of pollutants to the marine environment an a whole,, and,their rate of emohange between various water masses and ocean basinsp as Well &A between different phasesl such an solutioz4 suspension and biota. Monitoring these parameters mar also become important in relation to obsery d changes in the ecosystem ascribed to pollution but only if, as has already been pointed out earlierg the long-tam natural and cyclic changes in the ocean are sufficiently known. 97 3.1 National monitoring programmes Table I gives a review of national monitoring programmer based on scientific publications, in particular, on summaries published by the Smithsonian Institution (1970)and more recently, in May 1973 by the Intergovernmental Oceanographic Commission as a result of an inquiry to member,states of IOC about.their national monitoring activities (IOC, 1973). Although one can assume that the table contains information about the bulk of activities. in this field the review is most likely incomplete and both corrections and amendments will be needed. The first three columns of TableI inform-about environmental parameters and pollutants monitored by national programmer, the last two column give some details related to sampling stations and analytioal methods used for,marine pollution monitoring. The table contains data from 63 countries, and although there seem to be no nee& for a .specific discussion of its contents, it should be stressed that the number of activities is surprisingly high, also with regard to the analyses of pollutants from samples at sea. In case the scientific community was able to correctly.oompare the information available through these activities-and to interpret them according to our understanding of the pathways by which the contaminants are distributed both before and after they are deposited in the ocean may be facilitated. This could also lead us a stop forward in providing a long-term account of the accumulation of chemical contaminants in the marine environment which, in particular, is envisaged a strategy for a national programme developed recently in the U.S.A.(Goldberg, 1972). 1 should also like to' refer in this connexion to the scientific results obtained within the framework of the already mentioned ICES monitoring activities in the North Sea, e.g. Preston,.,( 1973) ICES (1974). Before we go on to consider existing regional or worldwide monitoring system I should like now to leave these general considerations on existing programmes and to deal with one specific example of a pesticide monitoring programs in order to demonstrate the carefulness which has to be given to suoh activities at all stages, i.e., during the preparatory and operational phase and for, the evaluation of data. The,programme had-been developed by the U.S National Marine Fisheries Service in order to assess the subtle harmful changes in the marine environment and its living resources which result from the accidental or.intentional transport of pesticides into the estuaries along the whole U.S.A coastline (Butler, 1969). Proceeding on the assumption that. pesticide pollution in estuaries would be intermittent depending on the seasonal use of pesticides, and that there would only be very low conoen- trations in the water mass, automated sensing devices could not be used. A suitable bioassay technique as selected on the basis of available studies on the ecology and managemen of the eastern oyster, Crassostrea virginioa. typicallly, a mature oyster is feeding 90 percent of the time and transports about 1 of water an hour through its gill' system to extract the planktonic food. The oyster is physio- logically active the entire year throughout much of its extensive geographical range. Most important, it is sedentary and easily handle& Through suitable experiments it could,be shown that oysters remove and store chlorinated hydocarbon pesticides present in the surrounding water at concentrations as low as 0.1 microgrammes per litre.Oysters continue to build up such residues in their tissues at. uniform rates an long as. the toxicant is present, This biological accumulation may produce DDT residues, for example, 70 000 times-as high an the DDT concentration in-the surrounding water.Further it is in an important fact that the oyster flushes these residues out of its tissues at a uniform rate when the water supply is no longer 'contaminated. By sampling an oyster population regularly, it is possible to determine when the water supply becomes contain- nated and when the contamination stops. 0 0 gr 4 v � N N m " H m N H H N H N H H H N N M N K M H H N Salinity � H H H H H m m N H N m N U'a Hm N N N m H " m H N " Temperatur � m m m m H m m m H H H H H H H H H K N 03Wgen N N m N m m m m m m m m H m m m m m m wutrients, H H m m m m N H Waves N N N m N N K m m m m " " m H H m m N Stratific. m H m N m m H m H N H H N m N )4 H H Currents m M M H M H H H H m Prim. Pro m m N K m m m m H m m H m Plankton m H N H K ca w H m N m N m m H Organic " H m m K m M r.0 Inorganic 14M H H K K M H m0 Pesticides m H H H0 metals H N m m m m N H m m H m N N Ox.Conx. V N m K m 4 a m m m H m m H H N H Don. SeWag a H m N H N H H H H m Heat m m H H m Radioaot. H H H m Organic m m Inorganic m H m m N m H m Pesticides N m K N m K N m metals H m m m m H N 03F.Cons. W H m m H N H m Don. Sevag Heat H. H R&dioact. m m H H m m Oil N Gas chroma H Thin-layer m IR- or UV- Mass spect Ot Spectroflu 0 Atom.abs.s Neutron ac M, Colorimeti7 Gravimetry H Ga=m-spec H m H m Ships of o N H Patrol shi H m m m m H H H m Research s 14 H H @eather of H Manned pla k m Unmanned p Buoys N m N H Coastal st H Aircraft o 0 S 0 0 0 M H. 14M H K M M N N' M H N N H H N M H IN H H M m m Salinjt@ N M M M M M M M K M N M H M M H H H H H H H H M H H N M M Tempera M H M M N M M N M M H M H H H N M M N Oxygen- M H H M H N H N M H H M k M M M M M M M H: H m m Nutrien M H H M M M , M M. H H Waves M H M M M M M H. M H M N M M M M M H H H Stratif M M H H M. H M M H M M_' H M H M M H M M M, Current H M H M H N Prim. P M N M M M M M K X M M 14 N M M K Planktoi M M H M M M M-N M H H M Orga.nic M M M, N H 'M H M Inorgan M M M H -M H H Pistici: M M M Metals a H H H H H H H M M H N M H M H M H M M Ox.CohS H M M M H H H M H H N M H@ H M H M N H H 'Dom. Sei H H H N'H K M M M H 0 H M M H M H M, Heat H K 14 M M Radioac M H H H Organic M H H H N H M Inorgan H H. H N, H M Pestici M H H K@ N m H- Metals M M M H M H Ox.Cons H N M M M H H M Dom. Se M H I Heat 0 N M A M M Mi Mi. M M 'M M Radioac H M M k H M H H H k M M Oil' M CA M H to Gas chr M M M Thin-la JR- br M 1.- .0 1.- M 0. M 0 Kass sp H. M -.gectro 0.0 0 H :1 1 H M it0m-Ab 1+ H M M, 1@eutron C61orim MI M Gravime H Gamma-s M H H ships o M M M M Patrol M H H M M N M _N H M Researc M M H 0' H M Weather M H 1@anned M H M Unmanne M M Buoys H H H M H Coastal H M M M Aircraf. 100 In order to use the oysters for a monitory programs on pesticides careful attention had also to be given to the sampling and analytical procedures. The main problem which had to be tackled in this connexion were the wide spread.of sampling stations from about 500 to 300 northern latitude both at the Pacific and Atlantic coast of the United States and along the coastline in the Gulf of Mexico which meant that assistance and cooperation of many people was required and therefore both sampling and analytical uniformity was difficult to maintain. Besides careful training of sampling personnel it was decided to accomplish all analyses in the same laboratory. This became possible because a technique had been found to prevent spoilage of the samples and degradation of peticide,residues for at least 30 days without refrigeration. 'Thus, samples could be sent by ordinary mail to the analysing centre. Analyses were made by gas liquid chromatography which allowed to quantify 10 of the most commonly used organochloride pesticides at levels above'0.01. ppm. Within nearly four years about 5 000 samples have been analysed from 170 Permanent stations and it was found that DDT and its metabolites are the most commonly present pesticides;dieldrin, endrin and toxaphene being the next in frequency of occurrence. Normally, the amount of-DDT was less than 0.5 ppm and only rarely the residue exceeded 1.0 ppm with a maximam value of 5.4 Ppm following. a single incident In general the MDT-residnes were not of sufficient magnitude to constitute a human health problem. However, their presence indicates the ubiquity of DDT in the estuarine food web. In particular, it was shown that in estuaries receiving significant amounts of agricultural ran- off,, the seasonal residue pattern is proportionally higher. In some oases the monitor data could clearly demonstrate the industrial discharge of pesticide waste whose presence had been unsuspected by the state agencies responsible for clean water programes. Several times the method has been proven sufficient to identify the specific sources of pesticide pollution. 3.2 Regional and worldwide monitoring programmes With regard to regional and worldwide cooperation, information in literature is only sporadic and Table II, in this case. should only be taken as giving some examples. With regard to joint marine pollution monitoring there are-only a few programmes included but it is likely that more joint surveys for pollution investigations are in existence and the-table, therefore, needs completion. 4.POLTANTS The pollutants to be monitored have in fact already been selected by the.present practice of baseline studies on national and regional levels. Your categories, however, seem to constitute the pollutants that appear to present clear and definable threats to the ocean system and theme groups, therefore, should be taken into account for worldwide monitoring on the basis of experience which may be gained with the pilot project envisaged within the framework of MOSS; these categories as defined by an ad hoo Working Group at the 5th Session of the Joint Group of Experts on Scientific Advice on Marine Pollution (GESAMP) in June 1973 are: petroleum, halogenated hydrocarbons heavy metals and transuranics. Patroleum in at present estimated to enter the oceans at an amount of about 2 million tons annually, and with the predicted doubling-of oil production over the next ten yews this amount may still rise. Although recently none groups argue that harmful effects of oil am its products to living resources have been resources have been overestimated, it is a fact that catastrophic oil spills have been.followed by mass mortalities of marine organisms and that fish eggs can be damaged with oil concentrations of about 0.1 ml/1 (e.g., oil spill at Bussard Bar, Massachusetts; experiments in U.S.S.R. with sturgeon eggs).Furthermore the reduction of the amenity of many beaches of the world which are soiled with tar ball and oil slicks is quite obvious. The impact of oil on the marine environment has recently been studied by a working group (IMCO/FAO/Unesco/WMD/IAEA/UN, 1976). 101 Tabl a II Reicional and Worldwide 16onitoring_Programmes A. Regional Coax-4. BQdly Environmental monitoring Pollutants Mumber, of or monitored stations Name Countries particip. 4x 0 rq IM0 0 44 rq 4& 41 40 0 4- Qa&- @ -4 41 -.4 to 0 1k. V) Coop.Siudiee Kuroshio. CSK IOC X, x xx Joint ope- Dozens x x x xx X,X x rational Coop.Studies Caribbeaz4 pollution CICAR @IOC, x x x -X--X X-Xxx smote X'X x progr Coop-Studies West Africa ICES, GINECA IOC, PAO xxxxxxxxx x x x x Coop-+Sttidiea Nediterra- JOC, nean, CIM, FAO,.ICBEK. X,Xxx.xxxxxxx x x x Nediterranean Sea IW1O x Unknown Nediterranean Sea OECD Sewage (6 oountrie4 Bacteria' Ligurian Sea France, *, 0 Ital. nao V Projeot "Ramoga- Some x x X. North Adriatic Sea Italy, Yugoslavia Project "As-aros" 09 x x North Bea, ICES + XXXXXXXXXXXI Festicidei Dozens x x x PCBS Baltic Sea ICES x x x xx xx x X.X x metal a x x x Petroleum Antarctic Research Progr., kRP 12 countries x x x xx@x, x x x x West and North Atlantic Canada Id ,alQ Bilateral coopera ion Unknown S. Vorldvide@ Worli %ather Watoh,'IMW. WIND X, 1, 0008 x x x Global Radiation Network IAyA I I .". . 1. k . 150 x x Kaja and BioaPhere, NAB Unesoo @X, x x 1008 x x Environmental health Monitor WHO X Baoter is, Long-term and expanded- x x PrOgro on ocean-remearch IOd-, LEFOR ICSPHO X X X x x x x x x x x x x Global invest. on main Cate- pollution in the-marine, gories @ of, enviromo, GIPME WMD/Ioc pollutants @IGOSS ICSPHO X,xx Netwo, .rk x x x x Integr- Global ocean IOC7 Petroleum Station SY8t am,IdOSS ICSPHO i.x XXXX X XX XL bydro_ 1.We x x x x carbons, Oil alioks, Tar. bal is 102 Halogenated bydrocarbons belong to the most ubiquitous pollutants in the oceans, as the above example of monitoring pesticides in the U*S.A* has shown. Polychlorinated biphemyls (PCBs) have been proven to produce a chloracne type of disease upon ingestion by man; DDT and its metabolites, although not affecting human health according to our present knowledge, seem to interfere in hormonal production in higher organisms of the food chain and affect the photosynthetic process in some algae. From metals, the damaging effects of mercury contained in fish products is well known. However, there is also an urgent need to monitor other elements which are likely to affect life processes, such as cadmium and lead. Finally, transuranic elements and their compounds which originate from nuclear reactors and devices, are very toxic as regards their chemistry and ra4ioaotivity. As their production is growing and their environmental concentrations are expected to increase, programmes of analysis and monitoring in the marine environment should be initiated. The main characteristics of these pollutants of worldwide importancet as well as those of interest for national monitoring programmes are given in Table III, which summarizes the effects of the various pollutants and discusses possibilities for monitoring them either at the source or in the field. The table further contains information about research which is still needed with regard to monitoring these pollutants and lists most suitable methods of measurement. 5- OUTLINE FOR A MDNITORIM PROMIME IN DEVELOPING GOUNMES General survey of_l*ssible sources. The first stop should be a general survey aimed at finding out the main sources of pollution along the coastline. This sounds simple; however, it requires extensive spade work. At every coastal place industries have to be examined with regard to the composition of their effluents and solid wastes, and analyses have to be made in order to got quantitative estimations for each of the main pollutant8. Figures for the biological o;Wgen demand for the waste coming from municipal and industrial sewage systems should be calculated for each of the big cities and the pollution load of the main rivers has also to be investigated taking into account seasonal variations of the runoff. Resulting from these estimations a map of the coastline may be designed which can a as a basis for further steps. This map will contain the main rivers (with figures for their rwwff and pollution load in BOD), the various types of.industries (indicating the main pollutants), cities and recreation places (giving the number of inhabitants, taking into account seasonal changes of these figures in touristic zones), and the number of treatment plants for domestic and industrial waste (indicating the degree of treatment). Base-line studies in selected areas. Based on the above maps, are" .can be defined in which either the total pollution load is Iextremely high or a specific pollutant is seemed to be predominant. Pollution investigations may concentrate on these areas first. Syste- matio analyses of sea water, sediments and/or organisms should be started in these selected zones for a period of at least one year. Guided by the results of such "base-line" studies plans can be,made for continuous marine pollution monitoring and these activities may then be'extended to the whole coastal zone* Establishment of a suitable laboratory Whilet the baae-line studies as basic investi- gations preceding any monitoring activities can be organized with the assistance of University or industrial research institutest or with the support from foreign laboratoriest the real monitoring programme has to be done by a specialized national laboratory equipped with the necessary sophisticated instrumentation and staffed with well trained scientific and tech- nical personnel, The capacity of these laboratories should be high enov4gh to deal both with all the analyses to be included in the monitoring programme and with basic research needed in this connexiozi, e.g., toxicity tests for chemicals released from industrieng bioassays With suitable organisms. G+ 31C pa & at 9 - .0 4+ 1@ 0 rl C+ 9+ 4+ w 41+ o", rr a 'ok F@ P, 9+ W. pl. 0 M OW e+ 1+ H o I-b 0 0 PA. ts 0 0 0 C+ 0 w m 0 m W. C+ 0 W. la 0 0 m 0 m 3c "a 0 12 Ia 0 tws* cl* tmr 0 Ci. v .01 .v cp- R i @ @, ".. . Voo og+ m Co- 2 e 0 1@ @* @ I P, P C 11 w- a V 1-0 @w 0 4 (D W* 0 @ cr; Cf. m p v F- 4 r. a VI- r 4+ is V W. 1--j 0 D) S2, w 0. We+ A m 0 c4- I Tr P,0 OM 0 I-j +o -04., OC.- m0 (D Fl, m Co- 0 w C4. o (D. W 0 o ce P.'--- m I- po W. - 1-- 9 4 3c 0 0 ti) 0 CR F6 ci- is m C+- p C+ Ic t-s- 0 09 64 1 F p i G+ (D W. + 0 p 9+ CD (), m C+ 0 c4- 42 C+ C+ 0 cr w P. 03 w (D C+ (D 0 so o o H 4.8 mc ce tr rpa W. em 0 C+ m m C+- tr m C+ -1 0 C4. c; 'a' ON P114 4 lp M 1-199, o" 4 ck 0 M 9) - M 0 0 4+ 0 0 10 a 0 Co. (D " @ -wi- 'a a 00(4, c+ P. c+ eq. w (D w 0 + 8. 8. rr 0 0 C+ 4+ m ID 74 0 4 C+ it 4+ w 0 :0 0 4 m 1 0 .0+00 6@ I.A. m P 9+9 e+ w & 0 w 0 ,i4;0 -4 9: 11 0 r a C+ c+- C+0 m m C+- C+. (D 0 tr 4 o 9 1 Flo 4 0 90 4 a a P'. 0 :9 0 .4 9: a 0 0 9 ", *. .1 11 W. 8 & 0 a 0 n 0 w p I- cf- :3 a a + o" C4. 0 t- 0 3904 Cpl- o) 0 0 cl, w-+- cf- I 1@ 0 P, C4. m to P. pw C4. r PO Ci- fag r p e+ CI- 0 COL 9+ 0 0 w 0 tr L-1 r0 C+ JM m AS 0 Ls IL m Ow 0 A T 4+ la 0 C+ 04 C+Oo Kn 0 0 U 0 a Ic 0 t:r a 0 0% 0lb. Ire n 1.4 0 P W. a1; 8 1'c'p W. C+ 4+ 1 e a+ H 0 We &+ Igo w A 00 S.p a 4 4 0 Q 9: W $3. a -I S La R I e M + 00 0-t 40 Not I'" 6@ 0 1.& F, I.& t+ c 0 40+10 0 10, m 0 ts 0 0 ts r 0 4+ tr m P Elf* F3 n- w 0 0 n Co. 0 F@ 0 A PI P+ 1 11 tau, 00 151* 0 i PL+ 0 o' -b ce 0 0 o o o 0 0 1 Id - +0 0 4 ts tr 0 n 0 0 ts Ic I --t @ ts a 0 13, po. n - 4+ a rr q p n 4tr I a 0 Ito 0 0 t 0 0 P 0 0 0 W. C+ 0 im. i+ CS @- @- C+ t+ cf- @J. tr 0 pt m p ie a f; m a 0 a 0 pp M a P 0 C+ pl. P ca lo. & .0 cp F a 4+ . a C+ 3 C+ 00 njeRtal a C+ P 9 0 a" "1 0 C4. 0CS 0 a C4 '0 P. P. 0 9 P C+ C+ > 9 n14 0 0 Do o Jw 9, MO n P. lid 0 P. 17 0 - ts C+ o 0 cor P, 0 0 w C+. to PIP 0 1.4 A9 00 w 0 00 w W. C+ 0 C+ 0 +(m 0' 5 0 0 0 1. 0 94 ts tr 4t 0 03 1.41 0 0 'no m tor @0- ts 0 ;L w, Al �0 tj-@ JA ts 0 IPA P, t7OL EEL 4+ Ia. C4. wo 0 w rr 'a 0 cr w 4 og+ -te 0 0 M tr 0 Poo 8 .1.4 4+ w v 31C 0 4+ -0 ts rr 4 A lk Pr 0 *t 0 a P. 9 FA PC F@ '& 0 0 w O:r PA 1 0 ok M w 0 0 0 "a w 9+ a sod a 0 FA. w 0 to$ O."S " I og* 94. C* Co. VIM 8, '.'+ St o 49 a eq. w Ic 1@ 0 tr 40 P. + to t@ Pm FA 12 0 0 1 0 ts m We p cf- 0 PA 0 i & 0,4 0 0 04- WA 8 !@ R I." R t@ f1b 0 Its a -,Am ts 04- ts C* & .0 rr a a 0 0 at 40 0 PA 0 0 11 9 0. I-J' 0 m 1+ W, C,+ Sr P, pa 0 I:r 0 &0 e 0!@* 1:r 0 M Sao 0 I"t 0 0 0 Ilb @& 1 0 a "o-I I 4 1@ 1 4 -1. IV 44 t* '90 I 1 *1 @ 0@ + C. l.+ 0 9 cf, 0 P. ow 10 0 0 o*r+ C4- (ft a 04 w a 0 "0"1 00 w a v 0 r 1:11, ago 0.1+ p rr 1 0 4 LW. tr a+ 0 M C+ p 0 4 G+ 0 40 0 0. tol 04 t 0 0 I-t 93 Vr I P- (F4 4 C+* po 0 1 0 Pr '4 13 00 o0 p9+ Co. 11 o 0 pa a 1-"44 Di. 0 0 0 ts 10 Q+ Pr to cf- M 0 14. W. --@ 0! i cr. lp 0 W M A I.- p 0 0 or) 0wi- co. Id 0 0 0 jp tr OR p 0 W cf. 91.2 " W* 1.4 Pr 4+0 I M 0'. 00a 0 C+- 0 0 0 64. A @OL Table III (continued) Source Effects on Monitoring Methods of measurement Research still needed 7. Other@ waste from ch 'cal industEj There is an ever growing Living resources and Wherever possible, sub- Gas chromatograpby, Bioassays and toxicity number of chemical sub- man eating.contami- stances well known to mass spectrometry, tests for a high stances from industry nated fish or shell- become accumulated by spectrometry and other variety of substances. which can be claasified fish. species should be con- techniques which have For biodegradable sub- either as biodegradable trolled at the dis- to be determined for stances their time : ubstances or as stable charge outlets from the various chemical dependent toxicity has ubstances, the latter factories - for biode- substances. to be established. being more dangerous. gradable substances monitoring the dilution and quick degradation in, the field way be sufficient.. Stable substances, howe.verg should be carefully monitored in the field with regard to their distribution 0 in sea water, sedi- ments and organisms. 8. Physical wants (suck as solid waste, susp ed matter, heat) Municipal and industrial Biological processes, Automated oontrolli ng Optical instrumen- Laboratory experiments outlets,, dumping of living resources. of temperature and tation, such as aimed at establishing waste, heated effluents . content of suspended scattering meter, models concerning the from power stations, etc. matter at the points fluorimeter. Tempera- dilution of waste of discharge (e.g.9 ture-measurements. under specified hydro- outlet of pipelines) graphic conditions - examination of water relevant field work masses and the sea to prove models by bed in areas of post-<Iisoharge surveys. dumping of sludge or dredging. t 107 Denim of a network of stations.Some few reference stations and more numerous impact stations,as explained at the beginning of thin lecture, have now to be chosen in order to set up the final monitoring programme. Impact stations will mainly be concentrated in areas' found to be most influenced by domestic And industrial waste,And special attention will certainly, be given to rivers and estuaries.A decision. has to be made about the type of sampling Stations (see Fig.1) However, as buoys and unmanned. platforms,need highly sophisti- cated instrumentation and are too expensive and because for satellite observations specially trained personnel in needed, most of the sampling activities will, at least in 'the beginning be done by small vessels which belong to the pollution laboratory,or to which the laboratory has easy access. Study of suitable sampling technique and frequency. Great attention has to be drawn to the suitable sampling technique in order to reach and maintain the highest standard of quality, during collection and analyses of samples.It is often easy to contaminate samples by careless sampling because the contaminants to be assayed occur at, very low level. Water samples for measurement of dissolved hydocarbons, for instance, taken, on board &'ship can easily be conta- minated by the ship's own activities.Metals, the other hand, are sometimes added to water, samples in measurable quantities through contact with metal surfaces in water samplers. Finally, compositional alterations Of water samples have to be taken-into account and require, therefore, the immediate measurement of these chemical perameters,and,if some pollutants (such an pesticides or 'none petroleum compounds) occur at a very low level My, which does not allow direct measurement, bioaccummulation in marine organism has to be used for conoen- trations-of these pollutants. The sampling frequency is limited, on the one hand, by the laboratory and collection capabilities (upper limit) and on the other hand, by the need to obtain statistically signi- ficant data (lower limit). Between these two limits, the sampling frequency has to be chosen so that seasonal changes in the concentration of pollutants arising from changing biological and chemical activity can be defined. As a minimum,four sampling periods-in the various seasons are required. Education and training. Whilst in most countries sufficient scientific and technical personnel are available, in some countries more emphasis should be given-to the education of chemists, biologists and oceanographers,as well as to laboratory technicians familiar with problem of marine pollution. These people should be enabled to reach a high standard in the field of pollution analyses and in the evaluation of the results through close' cooperation with very experienced scientists in foreign laboratories-through training being organized at home an part of bilateral aid or with the assistance of the UN Specialized Agencies. The training should cover both sampling and the use of strandardized methods of analyses of the most important pollutants in the environment and, at least the species of commercial value in the area. Evaluation of the data. The final goal of national pollution monitoriing programmes is the provision of suitable data both for scientific research and for the management of the living resources in fresh water-and the coastal region.Therefore the data' have to be evaluated in relation-to the environmental conditions, such an, temperature and salinity distribution, stratification of water masses, mixed layer.depth, tidal stream and currents, i.qe. that these additional parameters have to .be included in the monitoring activities. They are required to estimate the rate of exchange between different water masses or ocean basins,, to estimate the rate of input to the marine environment as a whole and to investi- gate how they are related to changes in the ecosystem which otherwise. may be ascribed to pollution. Exchange of data. One can envivsage that numerous data of containments in the environment and aquatic organism will be available as a follow-up of base-line surveys and monitoring activities. 108 From the beginning, Pollution monitoring programmes should be combined with a national data centre so that outcoming data can readily be handled according to the rules of the "Manual on Oceanographic Data Exchangell(IOC, 1967). This means that data,' after careful quality control, will be stored and are ready for retrieval at any time for researchers 'or desicion makers. This is of.particular importance for developing countries in the process of industrialization, when establishing derived working levels for discharge water quality, thatis the acceptable pollution level in the waste water which will depend on the already existing pollution load of.the recipient water. 6o REFERENCES Butler, P.A.f Monitoring pesticide pollution. Bio Science, 19:(10):889-91 1969 Cole, H.A., (convener), Discussion on biological effects of pollution in the sea. 1971 Symposium*held at the Royal Society, London, 28-29 April 1970. Proc*R*Soc. Lond.(13), 177-2 .77-468 FAO, Report of the Seminar on Methods of Detection, Measurement and Monitoring of 1971 Pollutants in the Marine Environment, Rome, 4-10 December 1970. Report of Panel 9. Design of World Monitoring System. FAO Fish.Rep., (99) Suppl. 1: 123 @p- Directory of institutions engaged in pollution investigations. Contaminants 1974 in aquatic organisms. FAO Fish.Circ.,(325) Rev. 1, 48 p. Inventory of data on contaminants in aquatic organisms. FAO Fish.Circ.,(338) 1976 Goldberg, E.D., (convener), marine pollution monitoring: strategies for a national 1972 programme. Deliberations of a workshop held at the University of Southern California, 25-28 October 1972, La Jolla, California ICES, Report of the working group for the international study of the pollution of the 1974 North Sea and its effects on living resources and their exploitation. C-oop. Res.Rep.ICES, (39):191 p. IDOE, -Baseline studies of pollutants in the marine environment and research recommendationso 1972 IDOE Baseline Conference, 24-26 May 1972, New York IMOAAO/bnesc0/WM0/WH0/IAEA/UN Joint Group of Experts on the Scientific Aspects of Marine 1971 Pollution (GESAMP), Report of third session. Annex VII: Scientific basis for a monitoring system for marine pollutionp including registration of deliberate or.accidental discharges into the marine environment,. FAO Fish.Rep., (102):pago var. The impact of.oil on the marine environmento Rep.StudoGESAMP, (5) 197@ IOC, Manual on international oceanographic data exchangeo Unesco TechoSer., (4) 1967 IGOSS '(Integrated Global Ocean Station System)o General plan and implement- 1971 ation programme for phase 1* Unesco TechoSer.p (8) Summary of replies to joint IOC/WMO circular letter No. 5 on on-going and 1.973 planned national marine pollution monitoring programmes. Paris, IOC-WMO/ITECH 1/12 109 Kinne, Ok and H. Aurich (Eds.), International Symposium "Biologische und hydrographische 1968 Probleme der Wasserverunminigung in der Nordsee und angrenzenden Gewlssern". Helgoland.Wiss.Meeresunters., 17(1-4):530 p. National Academy of'Sciences, Chlorinated hydroca, rbons in the marine environment. In Report prepared by the Panel on Monitoring Persistent Pesticides in the Marine Environmpnt. Washington, D.C., National Academy. of Sciences Preston, A., Heavy metals in British waters. Nature, Lond., 242(5393):95-7 1973 Smithsonian Institution, National and international environmental monitoring activities. 1970 Washington, D.C., Smithsonian Institution, October 1970, 292 p. MR/H6403/E/1.76/l/600 3 6668 00002 5215