THE USE OF ENCLOSED PLANKTON COMMUNITIES IN AQUATIC ECOTOXICOLOGY

Fate and effects of mercury, cadmium and selected aromatic organics in a marine model ecosystem

0000 Promotor :dr . J.H. Koeman hoogleraar in de toxicologie Co-promotor :dr . C. den Hartog buitengewoon hoogleraar in de hydrobiologie /Wôï*x>i,irt

JAN KUIPER

THE USE OF ENCLOSED PLANKTON COMMUNITIES IN AQUATIC ECOTOXICOLOGY

Fate and effects of mercury, cadmium and selected aromatic organics in a marine model ecosystem

PROEFSCHRIFT

ter verkrijging van de graad van doctor in de landbouwwetenschappen, op gezag van de rector magnificus, dr. C. C. Oosterlee, hoogleraar in de veeteeltwetenschap, in het openbaar te verdedigen op dinsdag 23 november 1982 des namiddags te vier uur in de aula van de Landbouwhogeschool te Wageningen

kripsrepro meppe l

\S\h \>\H\*&q -ol HOLLANDI I

ieder moment - midden overdag de telefoon gaat kan het bestaan omslaan in absurditeit

koos koster

-4 - nn v2o\.

STELLINGEN

1.Organisch e stoffendi e in laboratorium biodegradatietoetsen gemakkelijk afbreekbaar blijken,kunne n ind e praktijk in zeemoeilij k afbreekbaar zijn.

dit proefschrift

2. Modelecosystemen dienen ingetrapt e toetssystemenvoo r debeoordelin g van milieurisico van chemicaliën reeds na de eerste screeningsfase teworde n ingeschakeld. Hierbij dient niet noodzakelijkerwijs gestreefd teworde n naar standaardisering.

J. van de Harst,F . Schmidt-Bleek,J.F . de Kreuk enK.O .Günther . Ecotox.Environm . Safety 5, 377-381 (1981). dit proefschrift

3.Doorgaan sword t inmarien e model plankton systemen een successie van het fytoplanktonwaargenome n ind e richting vanmicro-flagellaten .Dez eword t eerder veroorzaakt door selectieve vraat door copepoden dan door verminderde turbulentie ind e modelecosystemen.

G.D. Grice,M.R . Reeve,P . Koeller enD.W . Menzel. Helgoländer wiss. Meeresunters. 30, 118-133 (1977). J.H. Steele enJ.C . Gamble. In:G.D . Grice enM.R . Reeve (eds.) Marine Mesocosms,Springe r Verlag,Ne wYork . pp. 227-237 (1981). dit proefschrift

4. Dewijz ewaaro p Menzel en Steele tot eenoptimaa l volume van eenmode l plankton systeem komen, isonvoldoend e gebaseerd opmetinge n of argumenten.

D.W. Menzel enJ.H . Steele. Rapp. P.V. Réun. Cons. int.Explor .Me r 173,7-1 2 (1978).

5.Binne nd eOEC D end e EEG zijnrecen t afspraken gemaakt,waarbi j kortdurende toxiciteitstoetsenworde nvoorgeschreven ,voorda t nieuwe chemicaliën op de markt toegelatenworden .Dergelijk e toxiciteitstoetsen zijn onmisbaar voor een eerste screening,maa r zijnongeschik t voor debeoordelin g van het milieurisico van chemicaliën inaquatisch e ecosystemen.

OECD Guidelines for Testing of Chemicals,OECD , Paris.1981. 6thAmendmen t EEC.Officia l Journal of theEuropea n Communities L259, 15oktobe r 1979. 6. Stimulering vanbiologisch e processen als gevolg van stress door verhoogde concentraties chemicaliën dient niet zondermee r als een positief effect te worden beoordeeld.

A.R.D. Stebbing.Aquat .Toxicol . 1,227-23 8 (1981)

7. Voor debepalin gva n de primaire produktie door fytoplanktonka n de gebruikelijke C—14method e ind emeest e gevallenvervange nworde n door een opmetin g vand e zuurstof concentratie verandering gebaseerdemethode .

P.J.LeB.Williams ,R.C.T .Rain e enJ.R . Bryan. Oceanologica Acta 2,411-41 6 (1979). S.B. Tijssen.Proc . Ill Internationales hydromikrobiologisches Symposium Smolenice (CSSR), 1980.

8.Momentee l zijnbiologisch emonitorin g technieken onmisbaar om de biologischbeschikbar e fractieva nverontreinigend e stoffen inwate r en sediment tekwantificeren .

E.D. Goldberg (ed.), The international Mussel Watch. National Academy of Sciences,Washingto n D.C., 1980.

9. Eenvergunnin g voor het lozenva n afvalwater op oppervlaktewateren dient mede gebaseerd te zijno p de resultatenva n veldonderzoek naar de mogelijke ecologische gevolgenva nd ebetreffend e lozing.Bi j dit onderzoek is het onjuist omd e eventueel ter plaatse reeds verontreinigde toestand als "natuurlijke"referenti e tegebruiken .

10. Ecosystemen bestaan.

J. Engelberg enL.L .Boyarsky .Am . Nat. 114,317-324 (1979). C.F. Jordan.Am . Nat. 118,284-28 7 (1981).

11.Overorganisati e enburokratiserin g vanwetenschappelij k onderzoek doodt motivatie enkreativiteit .

12.D e zakkenhebbe n de toekomst.

Jan Kuiper Theus eo f enclosed plankton communities inaquati c ecotoxicology Wageningen, 23novembe r 1982 WOORD VOORAF

Het doen van wetenschappelijk onderzoek is vaak teamwork en het onderhavige werk vormt daarop geen uitzondering. Al die jaren dat ik mij voor het Laboratorium voor Toegepast Marien Onderzoek MT-TNO met model ecosystemen heb bezig gehouden, werd ik gestimuleerd en gesteund door een grote groep mensen. Het was niet de vooropgezette bedoeling dat dit onderzoek zou leiden tot een proefschrift.N u dit ervan gekomen is,wi l ik graag diegenenbedanke n die een bijdrage geleverd hebben. Allereerst wil ik prof. dr. J.H. Koeman en prof. dr. C. denHarto gbe ­ danken voor hun ondersteuning van het initiatief tot het schrijven van een proefschrift. De direktie van deHoofdgroe p Maatschappelijke Technologie TNO, dr.H.J . Hueck en drs.J.S.A .Langerwerf , ex-hoofd, respectievelijk hoofd van de afdeling biologie van MT-TNO bedank ikvoo r hunbijdrage . De grote steunva n Gerard "Sieme" Hoornsman,Pi m deKoek,Joo pMarquenie , Piet Roele, Ben Schrieken, Marijke van der Meer en Henk van het Groenewoud, allen collega's in Den Helder, wil ik hier graag memoreren. Zonder Henk en Sieme als harde kern en Pim als bedachtzame vragensteller was er niets van het onderzoek terecht gekomen. Ook wil ik graag de vele collega's van het MT laboratorium te Delft bedanken voor hun bijdragen. In de eerste plaats de mensen van de werkgroep biodegradatie die meehielpen met de experimenten met de organische stoffen: R. van Aardenne, W.R. de Jong, J.F. de Kreuk, A.O. Hanstveit, M. Pullens, K. Schonk en H. Visscher. Veel analyses van de toegevoegde stoffen werden verricht door Trudy Putman, Jaap Spijk en J.H.L. Zwiers van de afdeling Analytische Chemie.D e mathematische steunva n Janva n deEikhof ,Cee s van de Togt enLe oWittebroo d werd zeer gewaardeerd. Een speciaal woord van waardering voor de groep tekstverwerking rond Gé van Velzen is hier op zijn plaats. Zij wisten uit "kilo's"manuscrip t concept- publikaties te toveren. Verder dank ik: - H.G. Fransz, W.W.C. Gieskes, R. Leewis, G. Kraay, G.C. Cadee, J. Hegeman en W. van Arkel, allen van hetNederland s Instituut voor Onderzoek der Zee voor hun hulp en adviesbi j allerlei problemen. - Jan Jaap Formsma en Ton Grolle, die als studenten eenwaardevoll e bijdrage leverden aanhe t onderzoek.

- 5 Uwe Broekman en Einar Dahl voor de inspirerende samenwerking gedurende POSER, alsmede het Ministerie van Buitenlandse Zaken en de Noorse autori­ teitenvoo r hun medewerking bij POSER. Riékelt de Vries, Dorus Pompert en Sjok Visser voor hun hulp bij de start van de experimenten. de EEG en Rijkswaterstaat-direktie Noordzee voor hun toestemming het deels in opdracht verrichte onderzoek tepubliceren . Inter Research, Springer Verlag, de American Society forTestin g andMate ­ rials, deBiologisch e Anstalt Helgoland en deAcademi ePress ,voo r hun toe­ stemming een aantal artikelen indi tproefschrif t te reproduceren, de Koninklijke Marine voor het ter beschikking stellen van een experimen­ tele lokatie in dehave nva nDe nHelder . Peter Davis voor de engelse correctie van hetproefschrift . CONTENTS: page

woordvoora f 5 1.INTRODUCTIO N 8 2.MATERIA LAN DMETHOD S 12 2.1.Desig nenclosure s 12 2.2.Fillin gth ebag s 13 2.3.Additio no fcontaminant s 14 2.4.Samplin g 15 2.5.Experiment s 16 2.6.Analytica lmethod s 18 3.REPLICATIO NO FPLANKTO NDEVELOPMEN TI NIDENTICALL YTREATE D MODELECOSYSTEM S 20 4.FAT EAN DEFFECT SO FSELECTE DCONTAMINANT SO NENCLOSE D MARINEPLANKTO NCOMMUNITIE S 22 4.1.Mercur y 22 4.1.1.Fat eo fth eadde dmercur y 22 4.1.2.Effect so fth eadde dmercur y 22 4.2.Cadmiu m 26 4.2.1.Fat eo fth eadde dcadmiu m 26 4.2.2.Effect so fth eadde dcadmiu m 26 4.3.Organi ccompound s 29 4.3.1.Fat eo fth eadde dorgani ccompound s 29 4.3.2.Effect so fth eadde dorgani ccompound s 30 5.REPRODUCIBILIT YO FECOTOXICOLOGICA LRESULT S 33 6.POSSIBILITIE SFO REXTRAPOLATIO NO FRESULT ST O THEFIEL DSITUATIO N 34 7.OPTIMA LEXPERIMENTA LSET-U P 36 8.SUMMAR YAN DCONCLUSION S 38 SAMENVATTING 42 9.CITE DLITERATUR E 46 APPENDIX 53 CURRICULUMVITA E PUBLICATIONS 1. INTRODUCTION

Modern industrialized societies produce large amounts of chemical waste. Accidents during production and transport and deliberate disposal of wastes in the sea may cause severe pollution problems. An important aspect ofeco - toxicology is the development of test procedures which can be used to pro­ duce information about the fate and effects of pollutants in ecosystems. This information should help the autorities to set standards limiting damage to the ecosystem inquestio n (Hueck and Hueck-Van der Plas 1976).

Ecotoxicology,however ,i s still in its infancy (Hueck-Van der Plas and Hueck 1979) and the test procedures being considered for national or international use must be submitted to continuous review. In particular the large gapbe ­ tween the laboratory and the field needs tob e bridged. In ecotoxicology eco- logists and toxicologists must cooperate to gather information on thetoxici ­ ty of chemicals in ecosystems. No test can give conclusive evidence that a pollutant is biologically harmless, but the tests used should provide infor­ mation needed to judge environmental risks,a s required by,fo r instance,th e conventions ofOsl o orLondon .

Because pollution problems occur in the field, it seems logical to assess the influence of pollutants in field experiments. In most cases, however, it is impossible to experiment with the natural marine system,althoug h oil spills and other dumping practices might be considered as large scale experiments (e.g. Boucher 1980,O'Sulliva n 1978).

Most studies in the field of aquatic toxicology, biodégradation and bio- accumulation are performed in the laboratory. Although these experiments yield useful information, extrapolation of the results to field conditions is difficult. The deficiencies of small scale laboratory experiments have been recognized by many authors (Gray 1974,Menze l and Case 1977,Zeitsche l 1978, Perkins 1979,D eKoe k and Kuiper 1981). The shortcomings of laboratory tests aremos t apparent in short-time (acute) tests.

The dose applied in such standard laboratory tests can differ from that reaching thebiot a under natural field conditions.Difference s arise from: a. the chemical form of a pollutant. Innatur e itma y be chelated by organic ligands,complexed , adsorbed, transformed,etc .

- 8- b. the duration of the stress applied. Persistent pollutants in nature generally act on a much larger time scale. This difference is especially important with organisms with long generationtime .

c. the concentrations of thepollutants . Innatur e sublethal effects may have significant consequences for the survival of populations at much lower concentrations of the pollutant than those in laboratory tests aiming at establishing LC50.

d. interactions with other chemicals and other environmental characteristics (temperature,etc. )typica l for a certain field situation.

Apart from the dose, other factors in experimental design are important. In simple laboratory tests ecological effects cannot be detected. This results from a number of factors: for example,th e choice of anappropriat e inoculum in biodégradation tests, or the choice of test organisms in toxicity and bioaccumulation studies. Normally only a limited number of marine species, which can be maintained easily in the laboratory, is available; many open sea species cannot be maintained or cultured in the laboratory. The diffi­ culties with the extrapolation from laboratory to natural circumstances is also clear from the monospecies character ofmos t laboratory tests,neglect ­ ing the multi-species interactions typical for the marine foodwebs. These interactions between different species are of fundamental ecological impor­ tance in nature. Appropriate marine foodwebs ofmor e than two trophic levels (primary producers and herbivores) are, however, not easily sustainable in normal laboratory facilities, because of the spatial and temporal dimensions needed and the costs involved (Menzel and Steele 1978).

It is clear that methods should be developed which enable the results of ecotoxicological experiments to be extrapolated to natural conditions. To evaluate the fate and effects of pollutants many authors (Ringelberg and Kersting 1978, Saward et al. 1975,Dortlan d 1980,Oviat t 1980)hav e recently used more complex systems that may be regarded as approximating field con­ ditions more closely than do laboratory experiments. In aquatic ecology Stepanek and Zelinka (1961) were one of the first to use large, flexible enclosures, suspended in natural waters, to study the rela­ tion between the plankton and itsbioti c and abiotic environment. McAllister et al. (1961) were probably the first to use this approach in the marine field. Since then many investigators have isolated part of the natural en­ vironment in large plastic enclosures (e.g. Schelske and Stoermer 1972, Menzel and Steele 1978,Davie s and Gamble 1979, Brockmann 1980,Kerriso n et al. 1980, Gächter 1979, Glesy 1980, Grice and Reeve 1982). The general aim of these investigations is to bridge the gap between laboratory and field conditions.

The method has several advantages. Wall effects in large bags are much less significant than in laboratory cultures andman y samples canb e takenwithou t excessive disturbance of the system. The large volume allows experimentation with several trophic levels and with organisms,whic h are often difficult to culture or even to maintain in the laboratory. A closed system has the further advantage thatmineralizatio n processes canb e easily studied. If the bags are made from a suitable material,th e temperature and light regime in­ sidewil l be (almost)th e same as that in the ambientwaters .

The use of large plastic bags has also become popular in studying the fate and effects of pollutants in marine and freshwater ecosystems (Davies and Gamble 1979, Schindler et al. 1980,Marshal l and Mellinger 1980, Grice and Menzel 1978).

In 1974 the Laboratory for Applied Marine Research MT-TNO started this type of research using Dutch coastal plankton communities. The ultimate aim of the research was to investigate the suitability of the method for the eco- toxicological testing of chemicals. In this thesis results of investigations carried out in theperio d 1975-1980wil l be discussed. Details of the various experiments were published in separate papers, which are partly reproduced in theAppendix , In the first experiments the replicability of the plankton development in non polluted systems was investigated, a prerequisite if the method was to be used in toxicological research. Chapter 3 gives information on theplank ­ ton development in identically treated plankton systems.

Further investigations were aimed at the determination of the fate and ef­ fects of pollutants in low concentrations on the enclosed ecosystem. In the first years the heavy metals mercury and cadmium were used, later organic compounds were introduced as contaminants. The rate of disappearance of or­ ganic compounds from the model ecosystems was compared with results of laboratory biodégradation tests carried out with water from the enclosures. Inthi swa y information on thepossibilitie s of extrapolation of results from these laboratory tests to a more natural environment became available. Chapter 4 isdevote d to these experiments.

- 10 Because several experiments were performed with the same chemical, informa­ tion on the reproducibility of results between experiments was obtained, which isgive n in chapter 5.

In spring 1979w e joined German and Norwegian colleagues inth e POSER project (Plankton Observations in Simultaneous Enclosures in Rosfjorden, Brockmann et al. 1982). In this project we studied the influence of the dimensions of the enclosure on the plankton development, important from a practical point of view. A second aim was to investigate if the toxicological results ob­ tained in relatively polluted Dutch coastal waters differed largely from results obtained in relatively clean North Sea water as found in the Rosfjord. Some of the results are discussed in the last chapters of this thesis.

The most important articles inwhic h the different experiments were discussed in detail are reproduced in the Appendix. These articles are referred to as Al, A2,etc . in the text.

- 11 2. MATERIALS AND METHODS

2.1.DESIG N ENCLOSURES

Figure 1 gives a diagram of the construction of the bags; these arebase d on those described by Brockmann et al. (1974). The diameter of the bags is 0.75 m. In most experiments bags with a depth of 3.5 m, containing approxi­ mately 1.5 m3, were used, although bags with depths of up to 20m have also been employed (A9).Th e bags are made from a two-layered foil (manufactured by Alkor-Oerlikon Plastic GmbH, Munich, Federal Republic of Germany). The inner layer is 100 \jm thich polyethylene (biologically inert) and the outer layer is 30 |Jmthic kpolyamid e (resistent tomechanica l damage).

plastic bag

•ea water

Fig. 1.Diagra m of aplasti c bag.

The frames from which thebag s are suspended aremad e from seawater-resistent aluminium; PVC buoys serve as floats. If there is a risk of damage to the bags by floating objects, the former can be surrounded by aprotectin g net. Contamination by rain water or bird droppings (input of nutrients) is pre­ vented by perspex covers. These covers do not adsorb more than 5% of the light in the ecologically important part of the spectrum (400 - 700 nm) (Brockmann et al. 1974). The foil was impermeable for gasses and also for

12 metals (Brockmann 1980). In routine experiments 4-8 enclosures are anchored near a raft in aquie t corner of the harbour of DenHelde r (Figure2) .

Fig. 2.Plasti c enclosures anchored in the harbour ofDe nHelder .

2.2. FILLING THE BAGS

Inmos t experiments 4-8 bags were simultaneously filled via abranche d pipe (Figure 3) with natural seawater collected a few miles off shoreb y a small tanker. This branched pipe device is of prime importance if random distribu­ tion of organisms over the different bags is to be achieved. In the first year (Al,2) a Begemann pump (typeK Z 120-40)wa s used. Tominimiz e damage to (for instance) diatom chains or copepods aVanto nFlex-i-Line rpum p was used in later experiments. Large predators, such as Ctenophora, Cephalopoda or fish larvae,wer e prevented from entering thebag s by filtration of the water through a 2m mnet .

An alternative method, which has been used to fill bags up to 20m deep,wa s to fix the bag onto the frame without the floats and then to sink the combi­ nation to the desired depth. By lifting the frame quickly to the surface the enclosure was filled and the floats could thenb e installed.

- 13- Fig. 3. Branched pipes to fill different enclosures simultaneously.

2.3.ADDITIO N OF CONTAMINANTS

The enclosures were used to study the fate of chemicals and their effects on the structure and function of marine plankton communities.Th epollutan t was added, mostly in single doses, at the start of the experiment. To this end about 100 1o f water was pumped out of eachba g into a PVC container in which it was rapidly mixed with a concentrated solution of the test compound. The mixture was at once pumped back in the enclosure via a sprinkler which was slowly lowered to ensure thorough mixing. Controls received the same treat­ ment with the exception of the addition of a pollutant. With the exception of cadmium, concentrations of the pollutants inth e controls were much lower than inth e contaminated systems.

Single doses were added, since this practice more closely simulates the "normal" field situation, where the source of a pollutant usually is an outfall, a river or a dumping event (Menzel and Case 1977). In the experiments performed in Den Helder,nutrient s topromot e phytoplank- ton growth were never added. The mineralization rates in the enclosed eco­ system were sufficiently high to enable thephytoplankto n to grow.

- 14 2.4. SAMPLING

The development of the phytoplankton, zooplankton and bacteria, together with a series of physico-chemical parameters (salinity, oxygen, pH, nutrients, temperature, concentrations ofpollutant s inwate r and sediment)wa s followed in the enclosed system (Figure4) .

/ i \ light

phytoplankton nutrients

chlorophyll-a P04-P,N0~-N ,N0~-N , phaeopigments NH.-N,reactiv eS i ,4C-assimilation 02 numbero falga lcell s speciesdistributio n D43 0/ D 66 5 • decomposers

relativenumbe ro f organismsi nwate r 1 andsedimen t zooplankton

numbero forganism s L___r speciesdistributio n m systemcharacteristic s

temperature,pH ,salinity , Secchidis cvisibility , concentrationadde d pollutanti nwate r andsedimen t

Fig. 4. Simplified diagram of the interrelations between the different tro­ phic levels and the abiotic factors influencing the development of the organisms on these trophic levels of the plankton community en­ closed by the plastic bags. The parameters by which changes in the system aremonitore d are also indicated.

- 15- During the first experiments all samples, except those of zooplankton, were taken daily with a non-metallic sampler consisting of two 1250 ml chambers that could be opened at any desired depth between 0 and 3 m. In later ex­ periments a non-metallic sampler was used containing up to three stoppered sampling bottles.Th e stoppers were removed by a rope at the desired sampling depth. This type of sampler has been used successfully at depths down to 35 m; at greater depths the bottles are compressed by the water pressure.

During the first experiment the zooplankton was sampled by a pump,bu t this was not successful. In later experiments the zooplanktonwa s sampled from the entire head of water (0 -2. 5 m)b y means of apip e with abal lvalv e at the lower end. The internal diameter of thepip e was 4o r 4.5 cm, and one sample consisted of 4 or 5 random lowerings of the pipe into the enclosure. The contents of the pipe were filtered through a 55 |Jmne t and the samples at once fixed and preserved in a 4%formaldehyd e solution in filtered seawater. In deeper bags zooplankton samples were collected by vertical hauls with a netprovide d with a conical opening to increase the efficiency.

The fouling of the walls by algae was monitored by following the algalbio - mass collecting on polyethylene sheets or on glass slides fixed ina plexi ­ glass frame and submerged in thebags . Itwa s found that growth on glass was twice that onpolyethylen e (Grollean d Kuiper 1980).

The sediment was sampled by means of sediment traps, which hung in the en­ closures at a depth of 2.8 m. They were routinely emptied at weekly inter­ vals. In most experiments the sediment, which had settled on the bottom of the bags,wa s collected at the end of the experiments.

2.5. EXPERIMENTS

Table I summarizes data on the model compounds used, the initial concentra­ tions planned, the duration of the experiments and other data concerning the experimental set-up. The duration of the experiments was related to the generation time of the organisms in the enclosed system. Copepods have a generation time of 1 - 2month s (Raymont 1977). Therefore 3-6 weeks seemed tob e aminimu m to detect significant influences on their development.

- 16 TableI .Summar yo fdat ao nth eset-u po fth eexperiments .

experiment duration additiono f onda y initial number details (days) concentrations ofbag s in March-April 1975 39 4 Al May-June197 5 29 4 Al August-September 31 HgCl control 2 1975 5p gHg.l- 1 2 A2 March-May197 6 43 HgCl. control 2 0.5 pgHg.l- 1 1 5 2 A3 50 1 May-August197 6 78 CdCl 3 control 2 1|j gCd.l- 1 1 A5 5 2 50 " 1 September-October4 8 CdCl 10 control 2 1976 l 1(J gCd.l- 1 1 5 2 A5 50 " 1 August-September 35 DCA control 1 1977 2p gDCA.l- 1 1 10 2 A7 25 1 May-June197 8 42 'DCA,4CP,DCP 5 control 2 0.1m gDCA.l- 1 1 A7 1.0m gDCA.l- 1 1 A7 0.1m g4CP.1- 1 1 A8 1.0m g4CP.1- 1 1 0.1m gDCP.l- 1 1 A8 1.0m gDCP.l- 1 1 August-September 28 4CP,DCP,NFA 3 control 2 1978 0.3m g4CP.1- 1 1 A8 1.0m g4CP.1- 1 1 0.3m gDCP.l- 1 1 A8 1.0m gDCP.l- 1 1 0.3m gNFA.l- 1 1 A6 1.0m gNFA.l- 1 1 A6

POSER1 8 HgCl2 2 control 4 March197 9 1p gHg.l- 1 2 A4.A9 4

POSER2 20 HgCl2 2 control 5 March-April 1979 5p gHg.l- 1 4 A4.A9 August-September 28 4CP.DCP 2 control 2 1980 0.3m g4CP.1- 1 1 1.0m g4CP.1- 1 1 A8 0.3m gDCP.l- 1 1 1.0m gDCP.l- 1 1 DCA= 3,4-dichloroanilin e DCP 2,,4-dichloropheno l 4CP= 4-chloropheno l NFA 5-•nitrofuroicacid-: :

17 2.6.ANALYTICA L METHODS

Here only general information on the analytical methods is given,detail s on methods for the determination of concentration of the added chemicals,aswel l as other details,pertinen t to individual experiments is given in thepapers . Mercury and cadmium concentrations were measured with atomic absorption spectrometry (AAS) and neutron activation analysis (NAA). The organic com­ pounds were analyzed with gas chromatography or highpressur e liquid chroma­ tography (HPLC).

Chlorophyll concentrations were measured according to Strickland and Parsons (1968) in 1 litre samples. In the first experiments extraction was improved by sonification, after 1976 a BraunMS K cell homogenizerwa s used (Hoornsman and Kuiper 1979). Pigment concentrations were measured with a Pye Unicam or a Vitatron photometer. Concentrations of chlorophyll and phaeopigments were calculated according to equations givenb y Lorenzen (1967).

The samples ofphytoplankto n were preserved with Lugol's iodine (Vollenweider 1969) and selected samples were examinated with a Zeiss inverted microscope. The main species were identified where possible, and named according to Drebes (1974), Hendey (1964)an d Lebour (1925). The small naked flagellates (diameter 3-8 pm),ofte n found in large num­ bers, could in most cases not be identified to species and are referred to as p-flagellates. The concentration and size distribution of suspended par­ ticulate matter was measured with an electronic particle counter. In 1975a Coulter counter model A was used, in the other experiments a Coulter counter model TA II with population accessory was used (Sheldon and Parsons 1967). Primary production was measured by Steemann Nielsen's (1952) C-14 method. 14 1 ml of a NaH CO. solution (ca 4 pCi)wa s added to 100m l samples in 125m l light and dark bottles. The bottles were incubated in-situ at depths of 0.5 and 2.0 m. During the first experiments samples were incubated for 6 hours (9-15 h), from 1977 onwards the incubation period was shortened to 4 hours (10-14 h).Afte r the incubation the samples were transported to the labora­ tory and filtered. Each filter was placed in a scintillation vial with a scintillation solution (Anderson and Zeutschel 1970, Pugh 1973). The vials were counted with a Nuclear ChicagoMar k 1o r aPackar d Tricarb liquid scin­ tillation counter. The inorganic carbon content of the water was determined by titration according to Strickland and Parsons (1968).

18 Oxygen concentrations at a depth of 0.5 m were measured with aYello w Springs model 54 or anOrbispher e Laboratories model 2603 oxygenmeter .

Concentrations of orthophosphate, ammonia, nitrite, nitrate and reactive silicate were measured according to Strickland and Parsons (1968), from 1976 onwards With aTechnico n AutoAnalyzer.

The zooplankton was counted, identified and measured by the procedures des­ cribed byFrans z (1976). Nauplii,copepodite s and adults of each species were divided into at least six size classes and the adults separated by sex.Sub - samples of the sample were examined with amicroscop e and at least 150orga ­ nisms counted. Changes in the population densities at the various stages of copepod development - copepods always form the major part of zooplankton biomass in the enclosures -wer e used to estimate development and mortality rates of selected species, using multiple regression analysis of abundance of size classes (stages)a t thevariou s sampling dates (Fransz 1976). Produc­ tion of organic matter by copepods was estimated by multiplying the means between zero and the upper limit of the 95%confidenc e interval ofdevelop ­ ment rates by the mean density and theweigh t increment for each time inter­ val between sampling dates (Fransz 1976). Dry weights of the copepods were derived from regressions of dry-weight on céphalothorax length given by Robertson (1968)an d Nassogne (1972).

The development of the bacteria was followed in 1975 by ATP measurements. Thereafter plate counts (medium 2216 E of Oppenheimer and Zobell 1952) or epifluorescence counts (Daley and Hobbie 1975)wer e used.

Other parameters measured included water temperature,light ,salinity , Secchi disc visibility andpH .

19- 3.REPLICATIO N OF PLANKTON DEVELOPMENT IN IDENTICALLY TREATED MODELECOSYSTEMS

The aim of the first experiments, carried out in 1975,wa s to establish the variability between the development of natural plankton communities insepa ­ rate plastic bags, filled simultaneously and exposed to identical conditions. High replicability is essential for application of the method to ecotoxico- logical problems, because high natural variability would make discrimination between the effects of added chemicals and natural behaviour impossible.

In these experiments four enclosures were used simultaneously (Al). Although some bags were lost due to accidents during the experiments results showed that the development of phytoplankton and zooplankton was very similar in separate plastic bags. Figure 5 and 6 show as examples the development of thephytoplankto n and the zooplankton during the experiment inMay-Jun e 1975.

As in all experiments, a succession of phytoplankton species was found. A first bloom was generated mainly by diatoms, which were probably limited by lack of nutrients. A second phytoplankton peakwa s formed by small p-flagel- lates. This succession of larger species (diatoms or large flagellates) to small |J-flagellateswa s found inman y following experiments.

chlorophyll (mg/m*) May-June 1975

o bag 1 • .. 2 .. 3 ù .. i samples 05 m

Fig. 5. Concentrations of chlorophyll at a depth of 0.5 m in the different enclosures during the experiment ofMay-Jun e 1975.

- 20- As in all experiments which followed, calanoid copepods were themos t impor­ tant species in the zooplanktonwit h respect to the biomass. During the first experiment sampling of the zooplankton with a pump failed, but in the sub­ sequent experiments several species developed in the enclosures from egg to adult and at the end of the experiment large numbers of copepods were present (Figure6) .

100cnumber of organisms per liter 100

0.1

time (daysl timef days) 10 15 20 25 30 15 20 25 30

Fig. 6. Development ofnauplii (a)an d copepodites (b)o f rémora longicornis during the experiment ofMay-Jun e 1975.

The high replicability found in these experiments made a series of toxicolo- gical experiments with enclosed plankton communities possible. Innearl y all experiments which followed two enclosures were used as controls (TableI) , and sometimes a model compound was added to duplicate systems in the same concentration. Results from these experiments confirmed thatplankto n commu­ nities exposed to identical experimental conditions inside simultaneously filled plastic bags developed in very similar ways. In the later experiments bacteria were also quantified and the results showed that bacteria also developed similarly in duplicate enclosures.

These conclusions have also been reached by other investigators (Takahashi et al. 1975,Davie s et al. 1975,Brockman n et al. 1977).

21 4.FAT E AND EFFECTS OF SELECTED CONTAMINANTS ON ENCLOSED MARINE PLANKTON COMMUNITIES

4.1. MERCURY

Mercury is one of the heavy metals most intensively studied and thiswa s the main reason why itwa s chosen as amode l pollutant in the first toxicological experiments and in POSER. In all experiments single doses of mercury(II) chloride were added to the enclosed systems to give initial concentrations ranging from 0.5 to 50 |JgHg. l .Thes e concentrations were chosen because _i the lowest concentrations having toxic effects are around 1 to 10|J gHg. l (Taylor 1977).

4.1.1.Fat e of the added mercury

In all experiments concentrations of mercury in the water decreased during the experiments by 3 to 50%pe r day (A2,A3,Kuipe r 1980). Measurement of the amounts of mercury on the walls of the enclosure and in the sediment,whic h collected at thebotto m of thebag s explained, onlypar t of the loss from the water column (3-25% remained in the system). A subsequent laboratory experi­ ment (A3,Kuipe r et al. 1980)showe d that large amounts of the added mercury were volatilized and disappeared into the atmosphere.Methylatio n of mercury to methylmercury was found in the experiments, but this mechanism seemsun ­ likely to explain the loss to the atmosphere. Transformation of the added mercury to volatile metallic mercury is more probable. This transformation can be carried out by bacteria, and also occurs as apurel y chemical process in seawater (A3,A4). These results support hypotheses about atmospheric transport of mercury put forward by Wollast et al. (1975) and also support observations on the West coast of Sweden (quoted in A3),whic h indicated that theNort h Sea was a source for atmospheric mercury (A3,A4).

4.1.2.Effect s of the added mercury

_l Addition of 0.5 pg Hg.l in one experiment (A3) resulted in no detectable significant effects on the enclosed plankton community. Detailed observations of the community fouling the walls of the enclosure showed that addition of 0.5 Hg.l changed the species composition of the periphyton (Grolle and Kuiper 1980). Addition of 1 pgHg.l " during POSER resulted in lower numbers

- 22- of bacteria compared with the controls directly after the addition, indica­ ting inhibition ofbacteria l growth (A4,Kuipe r 1982). _1 Single doses of 5 |Jg Hg.l were added in all experiments and resulted in comparable effects. Directly after the addition phytoplankton growth was ! inhibited. When mercury concentrations decreased phytoplankton growth re­ sumed, the overall effect being a lag phase in the occurrence of thephyto ­ plankton maxima. Whether adaptation or detoxification was the main factor ( for this resumed growth was not investigated. Similar effects were found on the copepod development in the enclosures. As an example Figure 7 shows the development of the phytoplankton and one species of copepods during one of the experiments(A3) .

23 1.0.0 Chlorophyll (mgm" 3) 20.0

10.0

5.0

2.0

1.0

0.5

0.2

0.1

0.05 March-May 1976 J controls + 05 /jg Hg.I" 0.02 is o 50 ..

10 20 30 iO *• time (days)

30 U0 *• time (days)

Fig. 7. Chlorophyll concentrations (a) and numbers of adults of Pseudocaliunus elongatus (b) in the different enclosures during the experiment in March-May 1976.

24 Addition of 5 |Jg Hg.l also influenced the development of the bacterial populations (Figure8) .

Fig. 8.Relativ e number of bacteria in the different enclosures during the experiment inMarch-Ma y 1976.

Addition of 50 pg Hg.l in one experiment (A3) has similar, but stronger effects on phyto- and zooplankton and on the bacteria. The copepods were nearly eliminated from the system, the phytoplankton was inhibited for a long time, resulting in a lag phase of a month, and the species composition of thephytoplankto n changed markedly.

In addition of these results, other effects typical for the ecosystem character of the experiments were shown. In the experiments carried out in Den Helder the species composition of the phytoplankton altered after the additions such that larger species were relatively more abundant in theen ­ closures which were contaminated with mercury. In all experiments in which the zooplankton densities in contaminated systems were lower than in the controls, the phytoplankton species composition was affected in the same way as in the mercury experiments. The effects on the phytoplankton species composition can be explained as an indirect effect of themercur y via selec-

25 tive feeding of the copepods. Copepods prefer larger species,s o that in the controls, where grazing pressure is higher, larger species have less chance of surviving than small species.

The increased population densities of the bacteria (Figure 8) can also be regarded as a food chain effect: dying phytoplankton cells served as a suitable substrate for thebacteria .

In A3 and A4 these results were compared with those from other workers using natural marine plankton communities. This comparison showed that comparable results were obtained using different plankton communities under different conditions at quite different locations.

The results of the experiments also indicated that the toxicity of mercury is related to the chemical form, this has also been found for copper and cadmium (Sunda and Guillard 1976,Sund a et al..1978) .

4.2. CADMIUM

Cadmium was used as a second model compound because much information on its toxicity was already available in the literature, it is a known pollutant in theDutc h coastal environment (Duinker and Nolting 1977), and also because itsbehaviou r in themarin e environment differs from that of mercury. Results arepresente d in detail inA5 . Cadmium was added to the enclosures in single doses of 1- 5 0 pg Cd.l _i Inon e experiment additional doses of 100 and 250p g Cd.l were added.

4.2.1. Fate of the added cadmium

The added cadmium remained in the experimental system, and accumulated very slowly (< 0.1% per day) in the sediment due to adsorption and subsequent settling of suspended particles (a combination of abiotic particles,phyto ­ plankton cells,dea d zooplankton and faecal pellets). Adsorption of cadmium to thewall s was negligable.

4.2.2.Effect s of the added cadmium

_l Addition of 5 or 50 |Jg Cd.l led to higher chlorophyll concentrations in comparison with those in the controls in two experiments. Ina third experi-

26- ment, inwhic h cadmium was added just after aphytoplankto n bloom,n o effects were found on the phytoplankton. .1 The higher phytoplankton biomass after addition of 5 or 50 pgCd. l could be the result of a number of different factors. Reduced grazing after addition -l -1 of 50 M8-1 is possible, but improbable after addition of 5 |JgCd. l be­ cause the grazers developed similarly to those in the controls. The possi­ bility that addition of5 pg Cd.l stimulatedphytoplankto n growth cannotb e excluded, although the possible mode of action remains unclear. Stimulation of activity as an initial effect on functioning of organisms has been found for many chemicals and may be the normal response to stress (cf. Stebbing 1981). _l In one experiment additional doses of 100 and 250 |jgCd. l were added. No significant effects of these additions on the phytoplankton were found, probably because the frequency of sampling was too low after these additions, and because the interactions between the different trophic levels became increasingly complicated in this experiment. During this experiment (May- August 1976) organisms on a third trophic level developed in the bags (the ctenophore Pleurobrachia pileus). It appeared that all concentrations of _i cadmium (1- 25 0 pg.l ') resulted in a dose-related inhibition of the development of this ctenophore (Figure9) .

WO numb er per bag :

Mav-Auaust 1976 o control 10 + ; jjg Cd.r'.bag -A 5 • 50

1 -

-time .days

10 20 30 iO 50

Fig. 9. Numbers of the ctenophore Pleurobrachia pileus in the different en­ closures during the experiment inMay-Augus t 1976.

27 The increased mortality rate of copepods in the controls due to grazing by P. pileus made demonstration of a possible increase in mortality rates re­ sulting from the cadmium additions more difficult. 50 |Jg Cd.l probably influenced the copepod development, 100 M8-1 certainly increased the _l mortality rate of some species, and 250 \lg.l killed almost all copepods. _l In the second experiment P. pileus was absent and 50 [Jg Cd.l clearly inhibited the development of the copepods (Figure 10). Moreover not all species were influenced similarly, resulting in changes in the species dis­ tribution. In nature such shifts in species composition may exert important effects onhighe r trophic levels via selective feeding.

300 r- biomoss dry weight. September - October 1976 3 mg.m' o control • control + 1/jgCd.r' a SugCd.r' » StjgCd.r' ° 50 /Jg Cd. I" 200

100

30 iO time,days

Fig. 10.Developmen t ofbiomas s of copepods during the experiment in September-October 1976.

The differences in response of the phytoplankton in the experiments were attributed to differences in cadmium speciation. Many literature data also show the importance of the speciation of cadmium for its toxicity (Sunda et al. 1978,Hardstedt-Rome o and Gnassia-Barelli 1980).

28- 4.3. ORGANIC COMPOUNDS

After the initial series of experiments with heavy metals the following organic compounds were studied: 4-chlorophenol (4CP), 2,4-dichlorophenol (DCP) and 3,4-dichloroaniline (DCA) (A7 and A8). Some attention was also paid to phenol. These aromatic compounds were chosen because laboratory ex­ periments had shown that they differed widely in degradability,pheno l being the least,an d DCA themos t resistent tobiodégradation .

A second aim of the experiments was to compare the degradation inth e model ecosystem with the results of laboratory biodégradation tests, carried out withwate r from the enclosures.

As a case study fate and effects of 5-nitrofuroicacid-2 (NFA)wer e studied (A6).

4.3.1. Fate of the added organic compounds

DCA was not degraded in the laboratory die-away tests. Although concentra­ tions in the model ecosystems decreased during the experiments to about50 % of the initial values, there were several indications that this decrease was not caused by biodégradation. It was shown that at least part of the added DCA could have diffused through the walls. This diffusion is awea kpoin t in the experiments. The problem of diffusion through the walls of thebag s was not important with the phenols. Phenol,4C P and DCP were degraded within one to three weeks in both the enclosure experiments and the laboratory tests. Shorter lagphases and faster degradation rates were found after repeated addition of 4CP,whic h showed adaptation of the bacterial community to 4CP.

Generally degradation rates were comparable to those found in the laboratory tests carried out simultaneously. Sometimes degradation rates weremuc h lower in the plankton community than in the laboratory test (4-CP, A8;phenol ,D e Koek and Kuiper 1981). Inbot h these cases concentrations of the contaminants decreased linearly with time, indicating that a factor, other than 4CP or phenol limited the growth rate of the bacteria degrading these compounds. Very low inorganic nutrient concentrations probably limited the growth of the bacteria, since addition of nutrients in the simultaneous laboratory experiments increased the degradation rate.

- 29 Limitation of degradation rates by lack of nutrients has been foundprevious ­ ly. The experiments with the plankton communities showed that apart from the nutrient status as such, ecological factors such as competition for these nutrients can influence the degradation rate. In experiments carried out in 1979-1981 4-nitrophenol (NP) and tetrapropy- lenebenzenesulphonate (TPBS) were used. In this experiments the differences in degradation rates in the enclosed plankton communities and in the labora­ tory die-away tests using the same water were even larger. BothN P and TPBS were more or less easily degraded in the laboratory and were persistent in the model ecosystem. These results confirmed the importance of ecological interactions in the determination of degradation rates under more natural conditions.

In one experiment with 4CP and DCP indications were obtained that stable and toxic intermediates were formed during the degradation of these compounds. This finding showed that it can be important to study fate and effects of contaminants simultaneously, as was done in the model ecosystem. This was also shown in the case-study with NFA. NFA appeared to loose its nitro-group within a day of addition to the bags, probably as a result of exposure to light. No adverse effects of the remaining molecule on the enclosed plankton community could be shown(A6) .

4.3.2.Effect s of the added organic compounds

_1 In the first experiment with DCA initial doses of 2, 10 and 25 |JgDCA. l _l were given, after a NOEC (no observed effect concentration) of 5.6 M8-1 had beenmeasure d in a laboratory testwit h the reproduction of Da.phn.ia magna as a criterion.N o effects of these additions on the enclosed ecosystem could _l be shown. In a later experiment initial doses of 0.1 and 1.0 mgDCA. l were given. Addition of 0.1 mg DCA.l resulted in inhibition ofphytoplankto n and bacteria, and addition of 1 mg DCA.l further resulted in ahig h mortality and changes in the species composition of the copepod community. The relative species composition of the phytoplankton was changed after the addition of both 0.1 and 1.0 mg DCA.l .Figur e 11 shows as an example the sizedistri ­ bution of the suspended particles (i.e.phytoplankto n cells)o n day 40 of the _i experiment. It was found that lowered grazing pressure (1 mg DCA.l vs.

controls) increased the relative importance of larger phytoplankton species.

4CP and DCP were added in three experiments in concentrations ranging from

30 _1 _1 0.1 to 1.0 mg.l . After addition of 0.1rag 4CP. 1 no effects on the en- _1 closed community could be recorded. Addition of 0.1 mg DCP.l resulted in a temporary reduction of numbers of bacteria as compared with the controls. _l Addition of 0.3 mg 4CP orDCP. l slightly inhibited thephytoplankto n growth rate. 1 mg 4CP or DCP.l caused a stronger inhibition ofphytoplankto n and also of the zooplankton. The phytoplankton species composition differed from that in the controls after addition of 1 mg 4CP orDCP. l (Fig. 11).Agai n lower zooplanktonpopulation s occurred simultaneously with larger phytoplank­ tonspecies .

Another effect of the addition of DCP on interactions of phytoplankton and zooplankton was shown in the third experiment, inwhic h ahighe r phytoplank­ tonbiomas swa s found in the enclosure with less zooplankton.

»- cumulative vol., %

control

control

^^ diameter 2.5 - U.O ym

I I .. i.O - 6i fjm %Z& .. 6i-10.1fjm ^ 10.1 - 160 fjm WE 16.0 -320fjm 32.0 -50.8/jm

Fig. 11. Size distribution of suspended particles (average of samples taken at depths of 0.5 and 2.0 m) on day 40 in an experiment in which DCA, 4CP and DCP were tested (May-June 1978).

- 31 Inon e experiment strong inhibition effects were found after that 4CP and DCP had disappeared from the water. This might indicate the formation of a toxic intermediate during the degradation of these compounds. Unfortunately no samples were available for identification of this intermediate at the moment that this indicationwa s obtained.

Concentrations causing effects in the different experiments were quite similar, but the intensity of the effects differed. Stronger effects were found at lower degradation rates.

32 5.REPRODUCIBILIT Y OFECOTOXICOLOGICA L RESULTS

Mercury, cadmium, DCP and 4CP were used in the same concentrations indiffe ­ rent experiments and thus yield information on the reproducibility of the experimental results.

The concentrations of mercury having effects on the enclosed plankton com­ munities were very similar in the different experiments carried out with Dutch coastal water. The results of enclosure experiments, carried out in other parts of the world (Saanich Inlet,Canada;Loc h Ewe,Scotland; Rosfjord, Norway, quoted in A4) were also comparable, showing that the high reproduci­ bility of the toxicological results in Den Helder was not related to the quality ofDutc h coastal water(A4) .

Inexperiment s with cadmium as amodelpollutan t minimum concentrations having effects on the enclosed plankton were within one order of magnitude in two experiments (1-5 MS Cd.l ). In another experiment, in which cadmium was added just after a phytoplankton bloom the NOEC (No Observed Effect Concen­ tration) found was between 5 and 50 pg Cd.l .Thi s difference was probably due to the different speciation of cadmium in the last experiment(A5) .

The concentrations of 4CP and DCP having effects on the enclosed community were also within one order of magnitude. In all experiments the type and intensity of the effects found depended very much on the situation of the plankton community. If for example the phytoplankton was not very active, no large inhibition effects were recorded (e.g.POSE R 2,A4) .

The fate of the added compounds was similar in the different experiments. Notwithstanding the fact that the starting conditions differed very much between experiments (temperature, light, nutrients, biomass, species compo­ sition), i t can be concluded that the results obtained in themode l plankton communities are reproducible. These large differences in starting conditions do notpreven t application of themetho d forecotoxicologica l problems.

33 6. POSSIBILITIES FOR EXTRAPOLATION OFRESULT S TO THE FIELD SITUATION

If the method is to form a contribution to bridging the gap between the laboratory and the field, it should be possible to extrapolate results of the experiments with some confidence to natural marine systems.

Some investigators have compared the development of a plankton community inside a bag with that in the surrounding water from which the contents of thebag s was derived, employing enclosures ranging from 68 to 1300m 3 (Davies et al. 1975,Takahash i et al. 1975, Takahashi and Whitney 1977,Gambl e et al. 1977). The phytoplankton in the bags developed similarly to that in the sur­ rounding water, at least for several weeks.Th e numbers of zooplanktonorga ­ nisms were often higher inside the bags, having less prédation than the waters outside.Brockman n et al. (1977),workin g withmuc h smaller enclosures (4 m3), state that during the firstpar t of their experiment,th e time course of nutrient and chlorophyll concentrations in their bags agreed well with that in thewate r from which the contents were taken.

Due to strong tidal currents in Dutch coastal waters, it is impractical to compare the development ofplankto n inside ourbag s with that at the location were the water was taken. The investigations during POSER (A9) showed that irrespective of the data of filling, a phytoplankton bloom of the same spe­ cies occurred in all enclosures (large and small) and in the fjord during the sameperiod , indicating that adominan t bag influence was absent.Natura l factors such as light,nutrient s and temperature were regulating the develop­ ment of the enclosed (and free)community .

Apart from the structural similarity in time between the enclosed plankton and the surrounding water,th e functioning of the system is also an important characteristic. In our experiments many important features of the enclosed system, such as rates of primary and secondary production and generation times of the copepods,ar e comparable to those found in theNort h Sea.

In nearly all experiments p-flagellates tended to dominate the phytoplankton after 4-6 weeks of enclosure, thus species composition became increasingly abberrant.Th ephytoplankto n and zooplanktonwer e sampled every two hours for a 24 h period in two experiments in 1977, to study the functioning of the system at the end of the experiments. Results showed (Kuiper et al. 1982) that primary production rates were not unusual for open sea systems. It was also shown that the copepods, which had developed in the bags during the

34- experiment, showed the same pattern of vertical migration as found in the open sea.

Information from these experiments and from literature suggests that the development of a plankton community inside a plastic bag is qualitatively similar to that of the "free" community at least for periods of up to four weeks (Menzel and Steele 1978). Research on freshwater plankton resulted in similar conclusions (Barica et al. 1980,Marshal l and Mellinger 1980).

With respect to the toxic effects found itwa s shown that the concentrations of mercury causing reactions of enclosed plankton communities, isolated in different locations (Canada, Scotland, Norway, The Netherlands) were very comparable. This indicates that results obtained in Dutch coastal waters canb e extrapolated to other sea areas(A4) .

It is obvious that the model ecosystems used are not identical to the North Sea or to the Dutch coastal water ecosystem. Because the enclosed systems, however, develop in a natural way, results may be extrapolated to natural systems with some confidence. The model systems must be seen in this respect as an intermediate between laboratory and field. The model system is more complex than the monoculture in the laboratory and less complex than the real field situation(s). Even if this less complex system is investigated in a relatively simple way, as was done in this study, concentrations of pollutants showing effects are among the lowest reported in the literature.

If concentrations having effects in de model ecosystems are found in the field, as for example is the case for cadmium (A5),i t seems probable that the ecosystems in question are already being influenced.

35 7. OPTIMALEXPERIMENTA L SET-UP

Ideally, a sufficiently large organization must be available tobac k up these experiments so that feed-back to the laboratory is possible to look forex ­ planations of observations in the complex model ecosystems. Ecotoxicology should focus on this synthesis of laboratory and (semi-)field research.

Apart from this organizational aspect, two factors are ofprim e significance for experimental design of enclosure experiments: the duration of the ex­ periment and the dimensions of the enclosure. In a field of applied research, such as ecotoxicology, the smallest possible enclosures used in the shortest possible experiments are preferable in terms of convenience of experimental handling,possibilitie s of replication and costs (Davies and Gamble 1979,A9).

Optimal duration. The duration of an experiment must be related to the generation time of the organisms in the system. Copepods have a generation time of 1- 2 monts. 3-6 weeks therefore seems to be a minimum to detect significant effects on their development. Experiments of up to three months duration have been carried out in small bags (A5).Mineralizatio n was intense enough to sustain sufficient primary production. The species composition of the phytoplankton does, however, change greatly with time, (j-flagellates becoming dominant. This succession was also found by others (e.g. Takahashi et al. 1975,Gric e and Menzel 1978).

It is clear that the system in the bags becomes increasingly unnatural with time. Since the experiments were designed to assess the influence ofchemi ­ cals on natural communities, long experiments (3 months) are not useful. Moreover duplicate controls diverge more and more from each other, after 4-6 weeks making the detection of differences between controls and con­ taminated systems more difficult, if only low numbers of bags are used. The optimal duration of an experiment is probably 4-6 weeks, depending on the water temperature and day length (i.e. rates of biological processes in the system).

Optimal dimensions. Enclosures ofwidel y differing dimensions havebee n used by different authors (0.3 - 16.000 m3). One of the aims of POSER was to compare the development of theplankto n in enclosures of different sizes (1.5 -3 0m 3)(A9). The develop­ ment of the phytoplankton in large enclosures was similar to that in small

36 enclosures with respect to biomass and species composition. The development of the bacteria and the zooplankton showed differences between large and small enclosures.

Mercury was added as a model pollutant and it was shown that results were very similar in large and small enclosures. Results were also comparable to those obtained by others in still larger enclosures (1300 m3, Grice and Menzel 1978).

The optimal dimensions of the enclosure appeared to be related to the aim of the experiment, the number of trophic levels included in the system, the species present and the population densities of the organisms. In A9 advan­ tages and limitations of different sizes are discussed. It was concluded that for ecotoxicological experiments in relatively eutrophic waters (e.g. Dutch coastal waters) bags containing 1- 2 m3 appear to be large enough. In more oligotrophic waters,wit h lowerpopulatio n densities of zooplankton, enclosures of 10 m3 are probably large enough to allow sufficiently large zooplankton samples(A9) .

37- 8. SUMMARYAN D CONCLUSIONS

Most investigations in ecotoxicology are carried out in the laboratory. Al­ though laboratory experiments are indispensable and yield useful information, it is difficult if not impossible to extrapolate results of short-term laboratory tests currently in use to real field situations. The need in aquatic ecotoxicology for experiments with complex systems, more closely approximating natural conditions, led to the use of large, flexible plastic bags, isolating part of the ecosystem, and suspended innatura l waters.Thi s approach has been used here to study marine plankton communities.Th e general aim of the study was to develop a method which could act as an intermediate between laboratory and field, for determining fate and effects of pollutants in low concentrations.

This thesis summarizes results of several experiments,reporte d in detail in different papers,whic h arepartl y reproduced in theAppendix .

Chapter 2 summarizes the materials and methods used.Durin g the experiments, which usually lasted 4-6 weeks the development of the phytoplankton, the zooplankton and the bacteria was followed (biomass and species composition), as well as several physico-chemical factors, affecting the organisms on the various trophic levels.

In the first experiments (Chapter 3) itwa s shown that Dutch coastal plankton communities, exposed to identical experimental conditions inside simulta­ neously filled plastic bags (contents 1.5 m3),develope d in very similar ways. Therefore thismetho d could be applied for toxicological research.

From 1975 onwards several experiments were carried out in which mercury, cadmium and selected organic compounds (phenol, 4-chlorophenol (4CP), 2,4- dichlorophenol (DCP) an 3,4-dichloroaniline (DCA) were added as pollutants. Most experiments were carried outwit hDutc h coastal water. In 1979w eparti ­ cipated in POSER (Plankton Observations in Simultaneous Enclosures inRos - fjorden), a project carried out in a south Norwegian fjord. There the fate and effects of mercury on plankton communities in enclosures of different dimensions were studied. Chapter 4 is devoted to the fate and effects of themode l pollutants in the different experiments.

The fate of the selected compounds was generally in accordance with the ex­ pectations. Cadmium remained in the system and accumulated very slowly into the sediment, which collected at the bottom of the bags. DCA was not degra-

- 38- ded, the other organic compounds were degraded inth e plankton system as well as in laboratory die-away tests. Some differences were, however, sometimes found between laboratory and semi-field experiments. Sometimes phenol and 4CP were degraded more slowly inth eplankto n community than in the simulta­ neous die-away tests, carried out using water from the enclosures.Thi s dif­ ference was probably caused by factors working on the ecosystem level,suc h as competition between different bacteria for substrate and inorganic nu­ trients and competition between bacteria and phytoplankton for inorganic nu­ trients.

This finding is of practical importance for the extrapolation of results of laboratory degradation tests to the field. Laboratory tests are often per­ formed in the dark so that competition between bacteria and algae for nu­ trients is generally absent. Moreover nutrients are often added in large amounts to laboratory cultures.Fo r this reason laboratory tests could easily overestimate the degradation rate in oligotrophic environments, such as most seas and oceans during a large part of theyear .

The fate of mercury in the modelecosystems was also unexpected. After addi­ tion of mercury(II)chloride to the systems concentrations decreased in the water and increased in 'the sediment and methylation of mercury was also found.A largepar t of the added mercury however,wa s lost to the atmosphere, probably asvolatil e metallic mercury.

Another important result was the indication in one experiment that toxic and stable intermediates were formed during the degradation of 4CP and DCP. In most cases biodégradation tests and toxicity tests are carried out separa­ tely. In the model ecosystem fate and effects were studied simultaneously. This also appeared to be important in the case of 5-nitrofuroicacid-2, which lost both the nitrogroup and its toxicity within one day after addition to the enclosures,probabl y as a result of exposure to light.

Inmos t experiments effects on the enclosed plankton community could be shown after the addition of the model pollutants. Concentrations causing these effects were relatively low as compared with results of laboratory toxicity tests. Apart from this sensitivity of the test system, it is important that it appeared to be possible to experiment with various organisms ona single trophic level, as well as with different trophic levels inon e system. Addi­ tion of contaminants led to effects on these interactions on, or between trophic levels.

39 In many cases differences were found in the species composition of the plankton community between controls and contaminated systems. Changes in the relative abundance of organisms on different trophic levels were also found-. These changes may be important in field situations, because small changes, such as the disappearance or the inhibition of one species, can cause large changes in the ecosystem by complex interactions.

Because mercury, cadmium, 4CP and DCP were used in the same concentrations in different experiments, information on the reproducibility of the experi­ mental results was generated. This information is given in Chapter 5. Not­ withstanding the fact that the starting conditions differed very much be­ tween experiments, it can be concluded that the toxicological results, i.e. fate of the added compounds and the concentrations having effects,ar e quite reproducible. The intensity of the effects depended very much on the situa­ tion in theplankto n community.

Chapter 6 is devoted to the possibilities for extrapolation of the results to field situations. Information from the experiments described here and from the literature suggest that the development of aplankto n community inside a plastic bag, is qualitatively similar to that of the "free" community at least for periods of up to 4 weeks. With respect to the effects found, it was shown that mercury had similar effects on plankton communities enclosed at different locations (Canada, Scotland, Norway, The Netherlands). These results indicate that results obtained with themode l plankton systems canb e extrapolated to other sea areas. On the other hand this simple method pro­ vides just a first step, and the model systems must be seen as an interme­ diate between laboratory and field situations.Th e model system is less com­ plex than the field situation. Even if this less complex system is investi­ gated in a relatively simple way, as was done in this study, concentrations showing effects are among the lowest reported in the literature.

In Chapter 7 some remarks aremad e on the optimal set-up of ecotoxicological experiments using enclosures. For ecotoxicological experiments, excluding fish or other larger carnivores, bags containing 1- 2 m3 appear tob e suf­ ficiently large in relatively eutrophic waters,suc h asDutc h coastalwaters . Inmor e oligotrophic environments enclosures with a contents of approximately 10 m3 are preferred to enable larger samples to be taken.Th e optimal dura­ tion of an experiment isprobabl y 4-6 weeks, because the enclosed community diverges more and more from the natural situation with time,makin g extrapo­ lation of results to field situations more questionable.

- 40- The enclosure method canb e applied to at least two field of ecotoxicological problems :t oasses s the impacto f specific dumping events and other environ­ mental problems, and to validate laboratory toxicity and biodégradation tests.

41 SAMENVATTING

Ecotoxicologisch onderzoek wordt doorgaans uitgevoerd binnen het laborato­ rium. Hoewel laboratorium experimenten onmisbaar zijn en waardevolle infor­ matie verschaffen, is het moeilijk zonie t onmogelijk om resultaten van kort durende laboratorium toetsen teextrapolere n naar concreteveldsituaties .

In de aquatische toxicologie bestond derhalve de behoefte om te kunnen ex­ perimenteren met complexe systemen, die in een aantal aspecten overeenkomen met natuurlijke ecosystemen. Dit leidde o.a. tot het gebruik van natuurlijke planktongemeenschappen opgesloten ingrot e plastic zakken.

In onderhavig onderzoek werd deze benadering toegepast op mariene plankton­ gemeenschappen. Algemeen doel van het onderzoek was de ontwikkeling van een methode om lot en effecten van vervuilende stoffen in een plankton systeem te bepalen. De methode zou een bijdrage moeten leveren de kloof tussen het laboratorium en de veldsituatie te overbruggen. Dit proefschrift is een sa­ menvatting van de resultaten van een aantal experimenten,welk e in detailbe ­ schreven zijn in een serie artikelen. Een deel van de artikelen is weerge­ geven inhe tAppendi x bij dit proefschrift.

Hoofdstuk 2 geeft een globale beschrijving van de gebruikte materialen en methoden. Gedurende de experimenten die gewoonlijk 4 tot 6 weken duurden, werd de ontwikkeling van het fytoplankton, het zoöplankton en bacteriën ge­ volgd. Daarnaast werden een aantal fysisch-chemische factoren gemeten, die van belang zijn voor de ontwikkeling van de organismen op de verschillende trofischenivo's . In demeest e experimenten werdenplanktongemeenschappe n af­ komstig uit het Nederlandse kustwater geïsoleerd in zakken met een inhoud van 1.5 m3.

De eerste experimenten toonden aan (hoofdstuk 3)da t de ontwikkeling vange ­ lijk behandelde planktongemeenschappen in gelijktijdig gevulde experimentele eenheden onderling grote overeenkomst vertoont gedurende deproefduur .Hier ­ door kon demethod e toegepast wordenvoo r toxicologisch onderzoek.

Vanaf 1975 werden een aantal experimenten uitgevoerd, waarin kwikchloride, cadmiumchloride en een aantal organische stoffen (fenol,4-chloorfeno l (4CP), 2,4-dichloorfenol (DCP) en 3,4-dichlooraniline (DCA)) als vervuilende stof werden toegevoegd. In 1979 werd geparticipeerd inhe tPOSE R (PlanktonObser ­ vations with Simultaneous Enclosures in Rosfjorden) projekt, dat uitgevoerd werd in een fjord in zuid Noorwegen. Hier werden lot en effecten van kwik-

- 42 chloride onderzocht op planktongemeenschappen geïsoleerd in zakken van ver­ schillende grootte (inhoud 1.5 -3 0 m3).Hoofdstu k 4 is gewijd aan het lot en de effectenva n demodelstoffe n ind everschillend e experimenten.

Het lot van de stoffen in het modelsysteem was in het algemeen in overeen­ stemming met de verwachtingen. Cadmium bleef in het systeem en accumuleerde zeer langzaam inhe t sediment,da t zichverzameld e op de bodem van de zakken. DCA werd niet afgebroken, de andere organische stoffen braken zowel in het planktonsysteem af, als in de laboratorium die-away toetsen. Soms werden echter verschillen gevonden tussen beide toetssystemen. Fenol en 4CP werden soms langzamer afgebroken in de planktongemeenschap dan in de die-away toets, welke tegelijkertijd werd uitgevoerd met water afkomstig uit het plankton systeem. Dit snelheidsverschil werd waarschijnlijk veroorzaakt door factoren werkzaam op ecosysteem nivo, zoals bijvoorbeeld competitie tussen bacteriën om substraat en competitie tussen bacteriën en fytoplankton om anorganische nutriënten.

Dit resultaat is van praktisch belang bij de extrapolatie van de resultaten van laboratorium biodegradatie toetsen naar de veldsituatie. Laboratorium toetsen worden vaak in het donker uitgevoerd, zodat competitie tussen bac­ teriën en algen afwezig Is. Bovendienworde nnutriënte nvaa k inovermaa t aan laboratorium cultures toegevoegd. Hierdoor zouden laboratorium toetsen ge­ makkelijk de afbraaksnelheden in oligotroof milieu kunnen overschatten. Vele zeeën en oceanen hebben een oligotroof karakter gedurende een groot deel van hetjaar .

Het lot van kwik in demode l ecosystemen was eveneens onverwacht.N a de toe­ voeging van het kwikchloride namen de concentraties inhe twate r af en inhe t sediment toe.Oo kwer d methylering van kwik geconstateerd. Een groot deel van het toegevoegde kwikverdwee n echter uit de systemennaa r de atmosfeer,waar ­ schijnlijk alsvluchti g metallisch kwik.

Een ander belangrijk punt was dat in één experiment aanwijzingen verkregen werden,da tbi j de afbraak van 4CP enDC P stabiele en toxische intermediairen werden gevormd. Meestal worden biodegradatie- en toxiciteitstoetsen geschei­ den uitgevoerd. In het modelecosysteem werden lot eneffecte n vanmodelstof - fe n tegelijkertijd bestudeerd. Dit bleek ook van belang bij 5-nitrofuraan- carbonzuur-2, dat zijn nitrogroep en zijn toxiciteit binnen een dag na toe­ voeging aanhe tmodelecosystee m verloor,waarschijnlij k als gevolgva nbloot ­ stelling aan licht.

43 In de meeste experimenten leidde het toevoegen van demodelsto f tot effecten op de opgesloten planktongemeenschap. De concentraties welke deze effecten veroorzaakten, waren relatief laag in vergelijking met de resultaten van laboratorium toxiciteits toetsen.

Naast deze gevoeligheid van het modelecosysteem is het van belang dat het mogelijk bleek experimenten uit te voeren met zowel vele soorten organismen op één trofisch nivo, als met verschillende trofische nivo's binnen éénex ­ perimenteel systeem. Toevoeging van vervuilende stoffen leidde veelal tot aantoonbare effecten op de interacties binnen één, of tussen trofische nivo's. In veel gevallen werden verschillen gevonden ind e soortensamenstel­ ling van deplanktongemeenscha p tussen de controles en devervuild e systemen. Ook verschillen in de relatieve dichtheden van organismen op verschillende trofische nivo's werden gevonden.Dez e veranderingen zoudenbelangrij k kunnen zijn in veldsituaties, omdat kleine veranderingen, zoals bijvoorbeeld het verdwijnen van een soort, grote veranderingen in het ecosysteem kan veroor­ zakenvi a complexe interacties.

Omdat kwik, cadmium, 4CP en DCP in dezelfde concentraties in verschillende experimenten werden getoetst, werd informatie verzameld over de reproduceer­ baarheid van de resultaten, welke gegeven is inhoofdstu k 5.Hoewe l de begin omstandigheden tussen de experimenten sterk varieerden, kan geconcludeerd worden dat de toxicologische resultaten, d.w.z. het lot van de toegevoegde stoffen, alsmede de concentraties welke in effecten resulteerden, reprodu­ ceerbaar zijn. De intensiteit van de gevonden effecten bleek sterk samen tehange nme t de situatie ind e planktongemeenschap.

De mogelijkheden om de verkregen resultaten te extrapoleren naar veldsitua­ tiesworde nbesproke n inhoofdstu k 6. Beschikbare informatie uit deverschil ­ lende experimenten alsmede uit de literatuur suggereert dat de ontwikkeling van een planktongemeenschap opgesloten in een grote plastic zak kwalitatief overeenkomt met die van het "vrije"plankto n gedurende tenminste vierweken . Verder werd aangetoond dat kwik vergelijkbare effecten heeft op plankton­ gemeenschappen die op verschillende plaatsen geïsoleerd waren (Canada, Schotland, Noorwegen, Nederland). Deze resultaten zijn een aanwijzing dat resultaten verkregen in de model ecosystemen geëxtrapoleerd kunnen worden naar andere zeegebieden. Aan de andere kant vormt de methode slechts een eerste stap en de model ecosystemen moeten gezien worden als een schakel tussen laboratorium- en veldsituatie. Het model ecosysteem isminde r complex

- 44- dan de veldsituatie. Zelfs indien ditminde r ingewikkelde systeem op eenre ­ latief oppervlakkige wijze wordt bestudeerd, zoals werd gedaan in onderhavige studie, dan nog zijn de concentraties welke effecten te zien geven relatief laagvergeleke n bij andere toetsen.

In hoofdstuk 7 worden enige opmerkingen gemaakt over de optimale opzet van dit type model ecosysteem experimenten. Voor ecotoxicologische toepassing, waarbij vissen of andere grotere carnivoren uitgesloten worden, blijken zakken met een inhoud van 1à 2m 3 groot genoeg in relatief eutrofewateren , zoals hetNederlands e kustwater. Inmee r oligotrofemilieu s verdienen grotere eenheden (ca 10m 3) devoorkeu r aangezien dan grotere monsters genomen kunnen worden. De optimale tijdsduur van een experiment is waarschijnlijk 4 tot 6 weken. De opgesloten gemeenschap gaat bij langere experimenten meer en meer verschillen van de natuurlijke situatie,waardoo r het steeds moeilijker wordt de resultaten te extrapoleren naar concrete veldsituaties.

De beschreven model ecosystemen kunnen binnen de ecotoxicologie tenminste toegepast worden op twee gebieden: in de eerste plaats om de invloed van specifieke dumping praktijken en andere milieuproblemen te bestuderen, in de tweede plaats tervalidati e van laboratorium toxiciteits- en biodegradatie toetsen.

45 9. CITED LITERATURE

Anderson,G.C . and R.P. Zeutschel (1970) Release of dissolved organic matter by marine phytoplankton in coastal and offshore areas of theN.E .Pacifi c Ocean. Limnol. Oceanogr. 15:402-407 .

Barica,J. , H. Kling and J. Gibson (1980) Experimental manipulation of algal bloom composition by nitrogen addition. Can. J. Fish.Aquat . Sei. 37:1175-1183 .

Boucher,G . (1980) Impact ofAmoc o Cadiz oil spill on intertidal and sublittoral meiofauna. Mar. Poll. Bull. 11:95-101 .

Brockmann,U . (1980) Biochemische Wechselwirkungen einer marinen Planktonpopulation in einem abgeschlossene Wasserkörper. Thesis Hamburg, 182pp .

Brockmann,U.H. ,K . Eberlein,H.D . Junge,H . Trageser and H.J. Trahms (1974) Einfache Folientanks zur Planktonuntersuchung insitu . Mar. Biol. 24:163-166 .

Brockmann,U.H. ,K . Eberlein,G . Hentzschel,H.K . Schöne,D . Siebers, K. Wandschneider and A. Weber (1977) Parallel plastic tank experiments with cultures ofmarin e diatoms. HelgoländerWiss . Meeresunters. 30:201-216 .

Brockmann,U.H. ,K . Eberlein,P . Hosumbek,H . Trageser,E .Maier-Reimer , H.K. Schöne andH.D . Junge (1977) The development of a natural plankton population in an outdoor tank with nutrient-poor sea water 1.Phytoplankto n succession,Mar . Biol. 43: 1-17.

Brockmann,U.H. ,E.H . Dahl and J. Kuiper Concept of POSER. Submitted toMar .Ecol . Prog. Ser.

Daley,R.J . and J.E. Hobbie (1975) Direct counts of aquatic bacteria by amodifie d epifluorescence technique. Limnol. Oceanogr. 20:875-882 .

Davies, J.M., J.C. Gamble and J.H. Steele (1975) Preliminary studies with a large plastic enclosure. In: L.E. Cronin (ed.) Estuarine Research, Vol. I, Academic Press, New Yorkpp .251-264 .

Davies, J.M. and J.C. Gamble (1979) Experiments with large enclosed ecosystems. Phil. Trans.R . Lond. B. 286:523-544 .

De Koek,W.C .an d J. Kuiper (1981) Possibilities ofmarin e pollution research at the ecosystem level. Chemosphere 10:575-604 .

46- Dortland,R.J . (1980) Toxicological evaluation of parathion and azinphosmethyl in freshwater modelecosystems . ThesisWageningen ,11 2pp .

Drebes,G .(1974 ) MarinesPhytoplankton . GeorgThiem eVerlag ,Stuttgart ,18 6pp .

Duinker,J.C .an dR.F .Noltin g (1977) Dissolved and particulate trace metals in the Rhine estuary and the SouthernBight . Mar.Poll .Bull .8 :65-71 .

Fransz,H.G . (1976) The spring development of calanoid copepod populations in the Dutch coastalwater sa srelate dt oprimar yproduction . Proc.10t hEur .Symp .Mar .Biol. ,Ostend ,2 :247-269 .

Gächter,R . (1979) MELIMEX,a nexperimenta l heavymeta lpollutio nstudy :goals ,experimenta l designan dmaio rfindings . Schweiz.Z .Hydrol .41 :169-176 .

Gamble,J.C ,J.M .Davie san dJ.H .Steel e(1977 ) LochEw eBa gExperiments . Bull.Mar .Sei .27 :146-175 .

Giesy,J.P .(1980 ) Microcosmsi necologica lresearch . Publ.Technica lInformatio nCente rU.S .Dept .Energy ,U.S .111 0pp .

Gray,J.S .(1974 ) Synergistic effects of three heavy metals on growth rates of a marine ciliateprotozoan . In: F.J. Vernberg and W.B. Vernberg (eds.),Pollutio n andphysiolog yo f marineorganisms ,Academi cPress ,Ne wYork ,etc. ,pp .465-485 .

Grice,G.D .an dD.W .Menze l(1978 ) Controlled ecosystem pollution experiment: effect of mercury. VIII. Summaryo fresults . Mar.Sei .Comm .4 :23-31 .

Grice,G.D .an dM.R .Reev e(1982 ) Marinemesocosms . SpringerVerlag ,Ne wYork ,43 0pp .

Grolle,T .an dJ .Kuipe r(1980 ) Development of marine periphyton under mercury stress in a controlled ecosystemexperiment . Bull.Environ .Contam .Toxicol .24 :858-865 .

Hardstedt-Romeo,M .an dM .Gnassia-Barell i (1980) Effect of complexation by natural phytoplankton exudates on theaccumu ­ lationo fcadmiu man dcoppe rb yth eHaptophycea e Cricosphaera elongata. Mar.Biol .59 :79-84 .

47 Hendey,N.I . (1964) An introductory account of the smaller algae of theBritis h coastalwaters . PartV : Bacillariophyceae. Her Majesty's Stationary Office,London , 317pp .

Hoornsman, G. and J. Kuiper (1979) Vergelijking van drie extractie methoden voorafgaande aan de spectrofoto- metrische bepaling van chlorogfyl. Report MT-TNO no.C L 79/45. 9pp .

Hueck,H.J . and E.H. Hueck-v.d. Plas (1976) A system of tests for potential effects of chemicals in the aquatic en­ vironment. In: Eds. R. Amavis and J. Smeets, Princples and methods for determining ecological criteria onhydrobiocoenoses . Pergamon Press,Oxford ,Ne w York,pp .467-528 .

Hueck-Van der Plas,E.H . and H.J. Hueck (1979) The prospective assessment of environmental effects of chemicals. Drug Design 8:311-354 .

Kerrison,P.H. , A.R. Sprocati,0 .Raver a and L. Amartini (1980) Effects of cadmium on an aquatic community using artificial enclosures. Env.Technol .Letter s 1: 169-176.

Ketchum, B.H.,V . Zitko and D. Saward (1975) Aspects of heavy metal and organohalogen pollution inaquati c ecosystems. In: A.D. Mclntyre and A.D. Hills (eds.) Ecological toxicology research. Effects of heavy metals and organohalogen compounds. Plenum Press, New York and London,pp . 75-90.

Kuiper,J . (1980) Continued investigations into pelagic microcosms subject to environmental stress by pollutants. III.Combine d effects of mercury and cadmium. Rapport MT-TNO no.C L 80/85, 67pp .

Kuiper,J . (1982) Ecotoxicological experiments with marine plankton communities in plastic bags. In: G.D. Grice and M.R. Reeve (eds.), Marine Mesocosms, New York, pp. 181-193.

Kuiper,J. , W. Brouwer enJ.H.L . Zwiers (1980) Het lotva n kwik(II)chloride toegevoegd aannatuurlij k zeewater opgesloten in eenplasti c zak. Rapport MT-TNO no.C L 80/29,9 pp .

Kuiper,J. , H. van het Groenewoud and G. Hoornsman (1982) Diurnal variation of some plankton parameters in an enclosed marine community. Neth. J. Sea Res. (in press).

Lebour,M.V . (1925) The dinoflagellates ofNorther n seas. Publ. by theMar .Biol . Ass.U.K. , Plymouth. 125pp .

- 48- Lorenzen,C . (1967) Determination of chlorophyll and phaeopigments. Spectrophotometric equations. Limnol. Oceanogr. 12:343-346 .

Marshall,J.S . andD.L .Mellinge r (1980) Dynamics of cadmium stressed plankton communities. Can. J. Fish.Aquat . Sei.: 37:403-414 .

McAllister, CD., T.R.Parsons ,K . Stephens and J.D.H. Strickland (1961) Measurements of primary production in coastal seawater using a large plastic sphere. Limnol. Oceanogr. 6:237-258 .

Menzel,D.W . and J. Case (1977) Concept and design: controlled ecosystem pollution experiment. Bull. Mar. Sei. 27:1-7 .

Menzel,D.W . and J.H. Steele (1978) The application of plastic enclosures to the study of pelagic marinebiota . Rapp. P.V. Réun. Cons. int.explor .Me r 173:7-12 .

Nassogne,A . (1972) Etudes préliminaires sur le rôle du zooplankton dans la constitution et le transfer de la matière organique au sein de la chaîne alimentaire marine enme r Ligure. Thesis Amsterdam.

Oppenheimer, C.H. and CE. Zobell (1952) The growth and viability of sixty-three species of marine bacteria as influenced by hydrostatic pressure. J. Mar.Res . 11: 10-18.

O'Sullivan,A.J . (1980) The Amoco Cadiz oil spill. Mar. Poll.Bull .9 : 123-128.

Oviatt, CA., H. Walker and M.E.Q. Pilson (1980) An exploratory analysis of microcosm and ecosystem behaviour using multivariate technique. Mar. Ecol. Prog. Ser. 2:179-191 .

Perkins,E.J . (1974) The biology of estuaries and coastalwaters . Academic Press,London ,Sa nFrancisco ,Ne wYork ,67 8pp .

Pugh,P.R . (1973) An evaluation of liquid scintillation counting techniques for use in aquatic primary productivity studies. Limnol. Oceanogr. 18:310-319 .

Ringelberg, J. and K.Kerstin g (1978) Properties of an aquatic microecosystem: I. General introduction to the prototypes. Arch. Hydrobiol. 83:47-68 .

- 49 Robertson,A . (1968) The continuous plankton recorder: a method for studying the biomass of calanoid copepods. Bull. Mar.Ecol . 6: 185-223.

Schelske,C.L . and E.F. Stoermer (1972) Phosphorus, silica and eutrophication of lakeMichigan . In: E.G. Likens ed.,Nutrient s and eutrophication. Spec. Symp.Amer . Soc. Limnol. Oceanogr.,pp .157-170 .

Schindler,D.W. , R.H. Hesslein,R . Wagemann and W.S.Broecke r (1980) i Effects of acidification and mobilization of heavy metals and radio­ nuclides from the sediments of a freshwater lake. Can. J. Fish.Aquat . Sei. 37:373-377 .

Sheldon,R.W . and T.R. Parsons (1967) A practical manual on the use of the Coulter counter inmarin e science. Coulter electronics comp.,Toronto ,Canada ,6 6pp .

Stebbing,A.R.D . (1981) Hormesis stimulation of colony growth inCampanuiari a flexuosa (Hydrozoa) by copper,cadmiu m and other toxicants. Aquatic Toxicol. 1:227-238 .

SteemannNielsen ,E . (1952) The use of radioactive carbon (C14) for measuring organic production in the sea. J. Cons. Perm. int.Explor . Mer 18:117-140 .

Stepanek,M . and M.D. Zelinka (1961) Limnological study of the reservoir Sedlice near Zeliv. XVII. The development ofphytonannoplankto n in silicon bags. Sei. Pap.Inst.Chem. Technol.o fWate r 5,par t 1,pp .321-323 .

Strickland,J.D.H . and T.R. Parsons (1968) A practical handbook of seawater analysis. Bull. Fish. Res.Bd . Canada 167: 1-311.

Strickland,J.D.H . and L.D.B. Terhune (1961) The study of in situmarin e photosynthesis using a large plastic bag. Limnol. Oceanogr. 6: 93-96.

Sunda,W.G . and R.R.L. Guillard (1976) The relationship between cupric ion activity and the toxicity of copper to phytoplankton. J. Mar.Res .34 :511-529 .

Sunda,W.G. ,D.W . Engel and R.M. Thuotte (1978) Effect of chemical speciation on toxicity of cadmium to grass shrimp, Palaemonetes pugio: importance of free cadmium ion. Environm. Sei. Technol. 12:409-413 .

Takahashi,M. ,W.H . Thomas,D.L.R .Seibert ,J . Beers,P .Koelle r and T.R. Parsons (1975) The replication ofbiologica l events in enclosed water columns. Arch. Hydrobiol. 76:5-23 .

50- Takahashi,M . andF.A .Whitne y (1977) Temperature, salinity and light penetration structures: Controlled Ecosystem Pollution Experiment. Bull. Mar. Sei. 27:8-16 .

Takahashi,M. , G.T.Wallace ,F.A . Whitney and D.W. Menzel (1977) Controlled Ecosystem Pollution Experiment: Effect of mercury on enclosed water columns. I.Manupulatio n of experimental enclosures. Mar. Sei. Comm. 3:291-312 .

Taylor,D . (1979) A review of the lethal and sublethal effects ofmercur y on aquatic life. Residue Rev. 72:33-70 .

Vollenweider,R.A . (ed.) (1969) A manual on methods for measuring primary production in aquatic environ­ ments. IBP Handbook no. 12. Blackwell Scientific Publ., Oxford and Edinburgh, 213pp .

Wollast,R. , G. Billen and F.T. Mackenzie (1975) Behaviour ofmercur y innatura l systems and its global cycle. In: A.D. Mclntyre and C.F. Mills (eds.), Ecological toxicology research. Effects of heavy metal and organohalogen compounds.Plenu m Press,Ne w York, London,pp . 145-166.

Zeitschel,B . (1978) Controlled environment experiments inpollutio n studies. Paperpresente d atOceanolog y Int. '78,Techn . Sess. B (Biol.Mar . Techn.) pp. 21-32.

- 51 APPENDIX

Part of the results discussed in this thesis are published in detail in the followingpapers :

Al Development of North Sea coastal plankton communities in separate plastic bags under identical conditions.Mar . Biol. 44:97-107 (1977). A2 An experimental approach in studying the influence of mercury on a North Sea coastal plankton community. Helgoländer wiss. Meeresunters. 30:652-665 (1977). A3 Fate and effects of mercury in marine plankton communities in experi­ mental enclosures.Ecotox .Environm . Safety 5:106-134 (1981). A4 Effects of mercury on enclosed plankton communities in the Rosfjorden during POSER. Submitted toMar . Ecol. Prog.Ser . A5 Fate and effects of cadmium in marine plankton communities in experi­ mental enclosures.Mar .Ecol . Prog. Ser.6:161-17 4 (1981). A6 Fate and effects of 5-nitrofuroicacid-2 (NFA) on a marine plankton community in experimental enclosures. Proc. 6th congress on aquatic toxicology of theASTM , October 1981,St .Loui s (in press). A7 Fate and effects of 3,4-dichloroaniline (DCA) in marine plankton communities in experimental enclosures (submitted toAquati c Toxicology) A8 Fate and effects of 4-chlorophenol (4CP) and 2,4-dichlorophenol (DCP) in marine plankton communities in experimental enclosures (submitted toAquati c Toxicology) A9 Influences of bag dimensions on the development of enclosed plankton communities during POSER (submitted toMar .Ecol . Prog. Ser.)

A4 and A9 were also published inBericht e aus dem Sonderforschungsbereich 94, Universität Hamburg,no . 21,Apri l 1982.

53- Development of North Sea Coastal Plankton Communities inSeparat e Plastic Bags under IdenticalConditions *

J. Kuiper

Central LaboratoryTN O Department of Marine Ecology; DenHelder ,Th e Netherlands MarineBiolog y44 ,97-10 7(1977 )

Abstract In two experiments lasting 4t o 6weeks , communities ofNort h Sea coastal plankton kept in separate plastic bags (ofabou t 1400 1) and exposed to the same environmen­ tal conditions showed very similar patterns of growth and decline. This result means that the method is suitable for the evaluation of toxic effects of environ­ mental pollutants at lowconcentration s on complex plankton systems.Th e phyto- plankton in the bags produced a succession of blooms,whic h were probably limited by shortage of nutrients.Th e dominant zooplankton organisms were various species of copepods which can develop in thebag s from egg to adult.Stron g indications were found thatmineralizatio n of organicmatte r occurs in the bags. Chemical parameters and phytoplankton biomasswer e found not tob e stratified, indicating that the contents of the bagswer ewel l mixed.

Introduction on structure and function of a marine plankton ecosystem. The main objective Toxicological investigations are usually of the first two experiments, reported performed in the laboratory..To inter­ here,wa s to find outwhethe r the method pret their results in terms of environ­ is suitable for the purpose in hand. The mental significance is difficult, one main question asked was whether separate reasonbein g that they usually are of but (nearly) identical plankton communi­ short duration and limited to single ties develop in the samewa y when exposed species. Long-term experiments with more to the same conditions. In case of a too complex systems that canb e regarded as large natural variability, itwoul d be a better approximation to field condi­ impossible todistinguis h between the ef­ tions are needed urgently (Ringelberg, fects of added chemicals and naturalbe ­ 1973; Lacaze, 1974; Kersting, 1975). haviour. Onewa y of bridging the gap between laboratory and natural conditions is to conduct experiments with large plastic bags housing plankton communities and Materials and Methods suspended in naturalwater s (Strickland and Terhune, 1961; Goldman, 1962; Fig. 1show s the construction of the Schelske and Stoermer, 1971, 1972;Horst ­ bags,whic h is based on that described mann, 1972;Brockman n étal., 1974;Par ­ by Brockmann etal .(1974) . sons, 1974). The bags (0.75m diameter) contain In 1975w e carried out threeexperi ­ about 14001 o f seawate r each, and are ments with North Sea coastal plankton made from aplasti c laminate ("Trikoron", communities kept in plastic bags in the Alkor-Oerlikon Plastic GmbH,Munich , harbour of Den Helder,Th e Netherlands. FRG), the inner layer consisting of The long-term aim of this investigation 100um-thic k polyethylene,whic h is bio­ and its sequel is to determine theef ­ logically inert,an d the outer layer of fects of pollutants at low concentrations 30um-thic k polyamide,whic h has good mechanical strength. The frames from which the bags are suspended in a shel­ *Workcarrie dou tunde rContrac tNo .110-75- 1 tered corner of the harbour of Den Hel­ ENVNo fth eE.C .Environmenta lResearc hPro ­ der are made from aluminium. PVC buoys gramme. serve as floats.Th e bags are shielded

Al-1 pipewa slowere d fivetime sint oth eba g ina narbitrar yway .Th econtent so fth e pipewer efiltere dthroug ha 5 5ur nnet , andth ezooplankto nfiltere dof fwa spre( - servedi na 4 %formaldehyd esolution . Thezooplankto nwa scounte daccordin gt o proceduresdescribe db yFrans z (1976). Thedevelopmen to fphytoplankto nwa s monitoredb ymeasurement s ofchlorophyl l aconcentratio naccordin gt oStricklan d andParson s (1968),b yinspectio nwit h aninverte dmicroscop eo fsample spre ­ servedwit hLugo l (Lund et al., 1958; Utermöhl, 1958),an db ymeasuremen to f particle-sizespectr awit ha Coulte r Counter,mode lA (Sheldonan dParsons , plastic 1967).Si xsiz eclasses ,havin gaverag e bag diameterso f 13,17 ,23 ,36 ,5 6an d7 8uih , weremeasured .I nth esecon dexperiment , sizeclasse si nth e3 t o1 1u mrang e j wereoccasionall yincluded . Thedevelopmen to fdecomposer s inth e waterwa smonitore db ymeasurement so f thedifference sbetwee nAT Pconcentra ­ aea wat tions (according toHolm-Hansen ,1969 )o h 0.45 and3 ur nfilter s (Derenbachan dWil| - Fig. 1.Diagra m of asuspende dplasti cba g liams,1 974) . Concentrationso forthophosphate ,ni ­ trate,nitrit ean dreactiv esilicat e weredetermine daccordin g toStricklan d fromcontaminatio nwit hrainwate ran d andParson s (1968).Othe rparameter sthaj t birddropping s (inputo fnutrients )b y weremeasure d includetemperature ,pH , Perspexshee tcovers . salinity,Secch idis cvisibility ,an d Thefirs texperiment ,carrie dou ti n oxygenconcentration .Value so fgloba l March-April,1 975 ,laste d6 weeks ,th e radiationpe rda ywer eprovide db yth e second,i nMay-June ,laste d 1month .I n RoyalDutc hMeteorologica lInstitute . eacho fthes eexperiments ,4 bag swer e Mosto fth estatistica lanalyse swer e simultaneously filledwit hnatura lse a performed accordingt oth eStandar dTSA M waterthroug ha speciall y contructed proceduresprovide db yth eCyphernetic s branchpip eusin ga Begeman npum p (type InformationCorporatio n (CIC). KZ 120-40).Th ese awate rwa scollecte d Walleffect swer estudie db ymean sof ! fromposition sa fe wmile sof fth eDutc h framescovere dwit hpolyethylen efilm . coastnea rDe nHelder .T opreven tlarg e Theframe swer elowere dt o0. 5 and2. 0n predators (Ctenophora,Cephalopoda )fro m depthan dalga lbiomas swhic hha dgrow n enteringth ebags ,th ewate rwa sfiltere d onthe mwa ssample deac hweek ,usin g througha 2 m mnet .Durin gth eexperi ­ chlorophylla a sa parameter . mentsn onutrient so rothe rchemical s wereadde dt oth econtent so fth ebags . Allsamples ,excep tthos eo fzoo - Resultsan dDiscussio n plankton,wer etake ndail ya tabou t 9.00hr swit hanon-metalli csample rcon ­ Development of the Phytoplankton sistingo ftw o 1250m lchamber stha t couldb eopene da tan ydesire ddepth .T o These awate rwit hwhic hth ebag swer e investigate thevariatio no fvariou s filledcontain sconsiderabl eamount so f parameterswit hdepth ,sample swer e siltan dsand ,whic har ehel di nsuspen ' takena tdepth so f0. 5 and2. 0m .I nth e sionb yturbulenc etha ti si npar tdu e firstexperimen tsample so fzooplankto n tostron gtida lcurrent s (Secchidis c weretake nb ypumpin g 1001 o fse awate r visibilitya tth estar to fth efirs te x fromeac hba gthroug ha 5 5u mne tint o périmentwa s0. 5m an da tth estar to f aPV Ccontainer .Th efiltere dwate rwa s thesecon d 2.4 m). Insideth ebag sther ^ returnedt oth ebags .I nth esecon dex ­ isles sturbulence ,an ds opar to fth e periment,zooplankto nsample swer etake n solidmatte rsoo nsettles ,wit ha resul ' fromth eentir ehea do fwate ri nth e tantincreas ei nSecch idis cvisibilit y bagsb ymean so fa pip e (3.21 volume ) Figs.2 an d3 sho wth eproductio no f providedwit ha bal lvalv ea tth elowe r phytoplanktonbiomas smeasure d aschlor end.Fo rcollectin ga sampl e (16 1), the phylla concentratio ni nth efirs tan d

Al-2 March- April 1975 chlorophyll a(mg/m 3) o tog 1 • ' 2 ' • 3 20 » " i samples 0.5 m

^

samples 2m

—J-;; •time (days) 30 1.0

Fig. 2. Concentration of chlorophyll a at depths of 0.5 and 2.0 m in different bags during first ex­ periment

chlorophyll atmg/mij May-June 1975

o bag 1 • •• 2 .. 3 A .. 1. samples 0.5m

• time (days! ^ >^a 0 5 10 15 20 25 Fig. 3. Concentration of chlorophyll a at'depth of 0.5 m in different bags during second experiment

Al-3 secondexperiments .Th ecurve ssho aw serieso fpeaks .Th e growtho falga ebe ­ Species ganafte r8 day si nth efirst ,an dafte r 1da yi nth esecon dexperiment ,reachin g Ceretaulina. &ZZ^&ZZZZZZZ2X'— bergonii a firstmaximu mon ,respectively , the Chaetoceros- 15thand -4t hdays .Th e differencebe ­ spp. Dytilum ppm byvolurrt i tweenthes etime-lag si sprobabl ydu et o brightwellii Eucampia— theligh tregim einsid eth ebags .I nMa y zoodiacus theamoun to fligh tenerg yenterin gth e Nilzschia bagswa snearl y twicetha ta tth een do f longissima sehgera March (duringth efirs t9 day si nMarc h Rhizosolenio- theaverag egloba lradiatio nwa s122 6J stotterfothii cm"2day~1a scompare dwit h225 6J cm- 2 Thalassiosira _ day-1i nth efirs t2 day so fMay) ;i n nordenskioldii addition,th einitia lloa do fdea dpar ­ ticlesi nth efirs texperimen twa smuc h largertha ni nth esecon d (initialSecch i /fhodomonas sp. discvisibilit y0. 5an d2. 4m ,respec ­ tively), wherea sth einitia lchlorophyl l Dinoflagellales — concentrationswer eth esam e (5.0m gm-3) . Aminimu mi sattaine di n2 5an d1 0days , respectively,afte rwhic hgrowt hstart s again,leadin gt oasecon dpea kin , re­ spectively,3 2an d1 6day safte rth ebe ­ ginningo fth eexperiment ,unfortunately , accidentscause dth elos so ftw obag so n the12t hda yo fth efirs texperiment , ando fon eba go nth e13t hda yo fth e Fig. 4.Developmen to fcel lvolum eo fdominan t second. phytoplankton speciesdurin g second experiment Fig.4 shows ,a safunctio no ftime , (averagefo rba gNos .2 ,3 an d4 ,an d0. 5 and thecel lvolum eo fth emai nspecie sde ­ 2.0m ) tectedwit hth emicroscop ei nth e second 4-

pprn by volume

7.0

March-April 1975 6.0 o bag 1 • • 2 * •• 3 5.0 i • U size range 12-69 p diometer

3.0

2.0

1.0- \>?

20 30 LO

Fig. 5.Concentratio n ofparticulat e matterdurin g firstexperiment .Averag e from samplestake naj t 0.5 and 2.0m dept h

Al-4 experiment.Th evolume swer ecompute d eterso f3 6an d5 6ur nwer ei nth emajor ­ fromcel lmeasurement susin gsimpl egeo ­ ity,wit hsmalle rparticle sbecomin gmor e metricshapes ,suc ha scylinders ,spheres , abundanti nth ecours eo fth esecon d etc.I nbot hexperiment sth efirs tpea k peak.Th epattern sfoun dar esimila rt o wasgenerate db ydiatoms ,i nth efirs t thoseshow nb ychlorophyl lconcentra ­ by chaetoceros diadema and Thallassiosira tionsan db ymicroscopi ccel lcounts . nordenskiöldii, inth esecon db yth elast - Thegrowt ho fphytoplankto n isofte n namedalone . limitedb ya deficienc yo flight ,nu - "* Thesecon dpea ki nbot hexperiment s trientso rboth ,o rb yzooplankto n wasgenerate db ymicroflagellates . grazing.T oasses sth eimportanc eo fon e Figs.5 an d6 sho wth econcentratio n ofthes efactors ,w eestimate dth econ ­ ofparticulat ematte ri nth edifferen t centrationo fvariou snutrient si nth e bags,a smeasure dwit ha Coulte rCounter . firstan dsecon dexperiment s (Figs.7 (Iti sassume dtha tth edensit yo fth e and 8). cellsequal stha to fth esurroundin gwa ­ Paasche (1973)studie dth ekinetic s ter.)I nth ecours eo fth efirs tgrowt h ofsilicat euptak ei n5 marin ediato m peako fth efirs texperiment ,mos to fth e speciesan dfoun dhalf-saturatio ncon ­ particlevolum ewa sfoun di nth esiz e stants (Ks) of0. 8 to3.3 7ug-a ts i1" 1. classeswit haverag eparticl ediameter s Heals odiscovere dtha tpar to fth ere ­ of2 3an d 36um .N osecon dpea kwa sde ­ activesilicat ecoul dno tb eutilize db y tected,sinc esiz eclasse swit ha diam ­ diatoms (0.3t o1. 3ug-a t I"1). Davis et etero f4 t o8 ur nwer eno tmeasured .A t al. (1973)use da KSo f0. 5 ug-atS i1- 1 thebeginnin go fth esecon dexperimen t tocomput emaximu mgrowt hrate so f skele- thesiz ecategorie swit haverag ediam - tonemacostatum. VanBenneko m et al. (1975), workingi nDutc hcoasta lwaters ,foun d thatwhe nsilicat ewa sdeplete d to0. 3 ug-atl" 1,artificia lenrichmen twit h ppm by volume silicatestimulate ddiato mgrowth . Ast ophosphat ea sa limitin gnu ­ . 1.0 trient,Goldber g et al. (1951)foun d that Asterionella japonica could not grow much furthera tconcentration sbelo w 0.15 to0.2 5 ug-at1-1 .I nthei rstud y of Chaetoceros gracilis, Thomas and Dodson May-June 1975 (1968)foun da rat elimitin gconcentra ­ o bag 1 2.0 tiono f0.2 2 ug-atPO4- P1-1 ,an da Ks • • 2 of0.1 2 ug-at1-1 .Va nBenneko m et " 3 al. * "*• (1975)conclud etha ti nDutc hcoasta l 1.0 size range 3-11 fidiameter \\ watersphosphat econcentration s aslo w as0. 1 ug-at1~ 1 couldb egrowth-limiting . *• timeldays) Ast onitrate ,Epple yan dThoma s 10 20 30 (1969)foun d KS valuesrangin gfro m0. 1 to1. 5ug-a t1-1 .Othe rinvestigator s (Eppleye t al., 1969;Maclsaa can dDugdale , 1969)obtaine dsimila rresults . Theabov eliteratur edat asugges t thatth egrowt ho fphytoplankto ndurin g ppm by volume thefirs tpea ki nth esecon dexperimen t islimite db ya deficienc yo fsilicat e size range 12- 69 fi diameter orphosphat eo rboth .Th efirs tpea ki n thefirs texperimen tma yb elimite db ya lacko fsilicate ,bu tthi si sles sclear . Iti sfurthe runcertai nwhethe rgrazin g byzooplankto norganism swa sa signifi ­ cantgrowth-limitin gfactor .I ti sre ­ markablethat ,althoug hnutrien tconcen ­ trationsafte rth efirs tphytoplankto n peaksta ylow ,othe rpeak sfollow .Podam o (1974a),workin gi nth esluic edoc ko f Ostend,Belgium ,obtaine dsimila rresult s (see"Developmen to fDecomposers "fo r anexplanation) . Duringthes eexperiments ,n osignifi ­ Fig. 6.Concentratio n ofparticulat e matter*dur ­ cantdifference s (P<0.05 )wer efoun dbe ­ ing second experiment.Averag e from samples tweensample stake na tdepth so f0. 5an d taken at0. 5 and 2.0m depth 2.0m .Thi swa sth ecas efo rth eparam -

Al-5 30 N03-N(pgat/l 1 March-Acril 1975

20 • /kl ^^{ 10 -

time(day s) . i

'°r NOrNtßga)/ll

Fig. 7.Concentratio n ofnutrient sdurin g firstexperiment .Averag e of twodepth s and allbags .Ver ­ ticalbar s indicate ± 1standar d deviation I

NOyNtßgat/l) 10

-»• time/days I

20 25

N02-N{pgal/l)

Fig. 8. Concentration of nutrients during second experiment. Average of two depths and all bags. Vertical bars indicate + 1 standard deviation

Al-6 etersmeasurin gphytoplankto nbiomas s tob eth esam e (Odum,1971 ;Parson san d aswel la sth ephysicochemica lparameter s Takahashi,1 973) . (excluding temperaturean dlight) . Asregard salga lgrowt ho nth ewalls , Figs.2 ,3 ,5 an d6 clearl yrevea la nochlorophyl lcoul db edetecte do nth e highdegre eo fsimilarit ybetwee nth e polyethylene filmsuspende di nth ebag s growthpattern so fth ephytoplankto ncom ­ after2 week si nth efirs tan d3 week s munities inseparat ebags .Th egrou po f inth esecon dexperiment .A tth een do f Brockmann,workin gi nHeligolan dwit h thefirs texperiment ,1. 4mg.m -2wa s plasticenclosures ,obtaine d similarre ­ founda t0. 5m an d0. 2mg.m -2a t2. 0m sults (Brockmann,1974 ,persona lcommuni ­ depth.A tth een do fth esecon dexperi ­ cation).Takahash i et al. (1975),wh o ment,0. 7 to1. 0mg.m- 2wa s detected. used fourver y largeenclosure s (70m3) , Thesevalue sindicat etha tfoulin gof ' alsofoun dver ysimila rgrowt hpattern s thewall so fth ebag sneve rbecam eseri ­ ofth ephytoplankto n ineach . ousdurin gth eexperiments .Th emai n Growthan dmortalit y rateswer ecom ­ foulingspecie swer epennat ediatom s putedfro mth eaverag echlorophyl lcon ­ (e.g. Nitzschia longissima and Navicula spp. centrationsa t0. 5 and2. 0m .N osignif ­ icantdifference swer efoun dbetwee ncor ­ responding slopesi ndifferen tbag s( P Development of the Zooplankton <0.05).Anothe rstron gindicatio ntha t thedevelopmen tpattern so fth esyste m Inbot hexperiments ,calanoi dcopepod s inth ebag sar esimila ri sth equalita ­ wereth edominan tspecies .I nth efirs t tivefac ttha tcorrespondin gpeak si n experimentth efollowin gspecie swer e differentbag sar egenerate db yth esam e found: Temora longicornis, Pseudocalanus sp., species.Theoretically ,condition si n Centzopages hamatust Acartia clausi and Eury- differentbag smus tb elargel yidentica l temora sp.I nth efirs texperimen tn o ifth epatter no fspecie ssuccessio ni s quantitative informationca nb egive no n

reactive silicate (ugat/l)

May-June 1975

Pig. 8 (continued)

Al-7 100r number of organisms per liter number of organisms per liter

May-June 1975 • bag 2 i - 3 » - 4

time (days)

10 15 20 25 30

-*- time (days! I I I

10 15 20 25 30

Fig. 9. Developmento f zooplanktondurin g second experiment.Numbe r of nauplii (a),copepodite s (b)an d adult copepods (c)pe r liter

0.11

Fig. 10.reraor a longicornis. Developmentduring k second experiment.Numbe ro fnaupli i (a),cope ­ podites (b)an d adults (c)pe r liter 0 5 10 15 20 25 30

Al-8 100c number of organisms per liter A the growth of these species,becaus e the method of sampling themwa s inadequate. The organisms probably managed to avoid being sucked into the pump hose (J. Gamble,Aberdeen , Scotland, personal com­ munication). Figs. 9-11 show the development of the zooplankton during the second ex­ periment.Apar t from the main species (Temora longicornis and Centropages hamatus), Acartia clausi and Paracalanus parvus devel­ oped in small numbers. Thewate rwit h which thebag s were filled also contained larvae of bivalves andworms ,a swel l as a few barnacle nauplii (whichwer e found as adults on the bottoms of the bags at the end of the experiment). Since the eggs of copepods hatch with­ in 2 days (Marshall and Orr, 1972), and since the number of nauplii increased during the experiment, it is probable that copepods can develop from eggs to adults in the bags. The very large num­ ber of nauplii and copepodites found in thebag s can be explained by the absence of predators.Th e plastic bags in Loch Ewe, Scotland, contained larger numbers of these organisms than the surrounding water. This was also attributed to less prédation (Davies et al'., 1975). Likewise, Podamo (1974b) found similar numbers of the organisms in the sluice dock atOst - end,whic h can alsob e regarded as a more or less closed system,wit h herbi­ vores at the end of the trophic chain. The growth patterns in separate bags of Temora longicornis and Centropages hamatus and other copepods are very similar.Mos t of the differences can be attributed to counting errors.

Development of Decomposers

5 10 15 20 25 30 Decomposers are organisms,e.g .bacteria , which decompose and mineralize organic matter. The differences between ATP con­ centrations on filters of0.4 5 and 3u m showed such awid e scatter that nocon ­ clusions could be drawn from them. Jassby (1975) has shown theoretically that estimation ofbacteria l biomass by means of fractional filtration techniques is usually impossible.

0.1

^Fig.11 . Centropages hamatus. Developmentdurin g secondexperiment .Numbe ro fnaupli i (a),cope ­ 5 10 15 20 25 30 podites (b)an dadult s (c)pe rlite r

Al-9 However,th efac ttha tth efirs t Literature Cited phytoplanktonpeak ,whic h isaccompanie d bya depletio no fnutrients ,i sfollowe d Antia,N.J. ,G.D .McAllister ,T.R .Parsons ,K . bya secon dpea kaccompanie db ymuc h Stephensan dJ.D.H .Strickland :Furthe r mea­ lessnutrien tdepletion ,strongl ysug ­ surementso nprimar yproductio n usinga laig e geststha tth eorgani cmatte rha sbegu n volumeplasti c sphere.Liranol .Oceanogr . 8, tounderg omineralization .Podam o (1974a) 166-184 (1963) obtained similarresult s fromhi sex ­ Bennekom,A.J .van ,W.W.C .Gieske san dS.B . periments inth esluice-doc k atOstend . Tijssen:Eutrophicatio no fDutc h coastalwa ­ Hecalculate d thatth enitroge ni nth e ters.Proc .R .Soc . (Ser.B ) 189, 359-374 sluicedoc kmus tb erecycle d 10time sa (1975) yeart omak eth esuccessiv ephytoplank ­ Brockmann,U.H. ,K .Eberlein ,H.D .Junge ,M . tonbloom spossible . Trageserun dK.J . Trahms:Einfach eFolien ­ Thefac ttha tth esecon dphytoplank ­ tankszu rPlanktonuntersuchun g in situ. Mar. tonpea k isneve rgenerate db ydiatom s Biol. 24, 163-166 (1974) mayb eexplaine db ydifference s inmin ­ Davies,J.M. ,J.C .Gambl ean dJ.H . Steele:Pre ­ eralization rates:i ti swel lknow ntha t liminary studieswit ha larg eplasti c enclc- siliconi srecycle d ata muc h lowerrat e sure. In: Estuarine research.Vol .I .p p251 — thanphosphoru san dnitroge n (Antia et 264.Ed .b yL.E .Cronin .Ne wYork :Academi c al., 1963). Press 1975 Davis,CO. ,P.J .Harriso nan dR.C .Dugdale : Conclusions Continuous cultureso fmarin e diatomsunde i silicate limitation.I .Synchronize d life Communities of North Sea coastal plank­ cycleo fSkeletonema costatum. J.Phycol .5 , ton isolated at the same time and having 175-180 (1973) the same size and composition show very Derenbach,J.B .an dP.J .L eB .Williams :Auto ­ much the same patterns of development trophican dbacteria lproduction : fractiona­ when exposed to the same environmental tiono fplankto n populationsb ydifferentiee l conditions. This makes application of filtrationo fsample s fromth eEnglis h Chan­ the plastic bag method in toxicological nel.Mar .Biol .25 ,263-26 9 (1974) research possible. Eppley,R.W. ,J.N .Roger san dJ.J .McCarthy : The "normal" behaviour of an isolated Half-saturation constantsfo ruptak eo f ni-k plankton community in a period of 4 to 6 trätean dammoniu mb ymarin ephytoplankto n4 weeks is to produce a succession of Limnol.Oceanogr . 14, 912-920 (1969) phytoplankton blooms, which are probably andW.H .Thomas :Compariso no fhalf-saturatio r limited by a shortage of nutrients. The constants forgrowt han dnitrat e uptakeo f dominant zooplankton organisms in the marinephytoplankton .J .Phycol .5 ,375-37 ! bags are various species of copepods. (1969) From the fact that the successive growth Fransz,H.G. :Th esprin g developmento fcalanoi d stages of the copepods strongly increase copepodpopulation si nth eDutc h coastalwa ­ in number, it follows that the various tersa srelate dt oprimar yproduction .Proc . species can develop in the bags from egg Eur.mar .biol .Symp . 10, 247-269 (1976) to adult. Goldberg,E.D. ,T.J .Walke ran dH .Whisensan d Strong indications were found that Phosphate utilizationb ydiatoms .Biol .Bu:.l . mineralization of organic matter occurs mar.biol .Lab. ,Wood sHol e 101, 274-284 in the bags. (1951) Chemical parameters and phytoplankton Goldman,C.R. :A metho do fstudyin g nutrient biomass were found not to be stratified, limiting factors in situ inwate r columns| indicating that the contents of the bags isolatedb ypolyethylen e film.Limnol . were well mixed. Oceanogr.7 ,99-10 1 (1962) Holm-Hansen,0. :Determinatio no fmicrobia lb .o - massi nocea n profiles.Limnol .Oceanogr . f4, Acknowledgements. Thanks are due to my col­ 740-747 (1969) leagues W.C. De Koek, G.H. Hoornsman, P. Roele, Horstmann,U .: fibe rde nEinflu ßvo nhâusliche à B. Schrieken, J. Van de Eikhoff and H. Vanhe t Abwasserau fda sPlankto ni nde rKiele rBupht , Groenewoud, who helped with the chemical, bio­ Kieler Meeresforsch. 28, 178-198 (1972) logical and statistical analyses and with eval­ Jassby,A.D. :A nevaluatio no fAT Pestimation ^ uating the results. Acknowledgement is also due ofbacteria lbiomas si nth epresenc eo f for help given by other colleagues at Delft. phytoplankton. Limnol.Oceanogr . 20, 646-6^8 Finally, I thank Drs. R. De Vries, H.G. Fransz, (1975) W.W.C. Gieskes and R. Leeuwis of the Netherlands Kersting,K. :Th eus eo fmicrosystem s forth e Institute for Sea Research, Texel, for their evaluationo fth eeffec to ftoxicants .Hy d:o - help and advice on various aspects of the work biol.Bull . (Amsterdam) 9, 102-108 (1975) here reported.

Al-10 Lacaze,J.C. : Ecotoxicology of crudeoil s and Ringelberg,J. : Parameter dependant (temperature) theus eo f experimentalmarin e ecosystems. tolerance levelsan d the influence of the Mar. Pollut.Bull . 5, 153-156 (1974) complexity of thebiologica l system.Hydro ­ Lund,J.W.G. ,C .Kiplin g and E.D.L eCren :Th e biol. Bull. (Amsterdam) 7, 106-114 (1973) inverted microscopemetho d of estimating Schelske,C.L . andE.F . Stoermer:Eutrophication , algal numbers and the statistical basiso f silica depletion andpredicte d changes in estimationsb y counting.Hydrobiologi a 11, algalqualit y inLak eMichigan .Science ,N.Y . 143-170 (1958) 173 (3995), 423-424 (1971) Maclsaac,J.J .an d R.C.Dugdale :Th ekinetic s of — Phosphorus, silica and eutrophication of nitrate andammoni a uptakeb ynatura lpopula ­ LakeMichigan .Spec .Symp .Am . Soc.Limnol . tions ofmarin ephytoplankton .Deep-Se aRes . Oceanogr. 1, 157-170 (1972) 16, 45-57 (1969) Sheldon,R.W . andT.R . Parsons:A practica l manual on theus eo f theCoulte r Counter in Marshall, S.M. and A.P.Orr :Th ebiolog y ofa marine science,6 6pp .Toronto :Coulte r marine copepod, 195pp . (Edinburgh,London : Electronics SalesCo . 1967 Oliver &Boyd) . Reprinted by Springer Verlag, Strickland,J.D.H . andT.R . Parsons:A practical Berlin: 1972 handbook of seawate r analysis.Bull .Fish . Odum,E.P. :Fundamental s of ecology, 574pp . Res. Bd Can. 167, 1-311 (1968) Philadelphia: Saunders 1971 andL.D.B . Terhune:Th e studyo f in situ Paasche,E. :Silico n and the ecology ofmarin e marinephotosynthesi s using a largeplasti c plankton diatoms. II.Silicate-uptak ekine ­ bag. Limnol.Oceanogr .6 , 93-96 (1961) tics in fivediato m species.Mar .Biol . 19, Takahashi,M. ,W.H .Thomas, D.L.R .Seibert ,J . 262-269 (1973) Beers,P . Koelleran dT.R . Parsons:Th erepli ­ Parsons,T.R. : Controlled ecosystem pollution cation ofbiologica l events inenclose d water experiment (CEPEX).Envir .Conserv . p.22 4 1, columns.Arch .Hydrobiol . 76,5-2 3 (1975) (1974) Thomas,W.H . andA.N .Dodson :Effect s ofphos ­ andM .Takahashi :Biologica l océanographie phate concentration oncel l division rates processes, 186pp .Oxford :Pergamo n Press andyiel do f atropica l oceanic diatom.Biol . 1973 Bull.mar .biol .Lab. ,Wood s Hole 134, 199- Podamo,J. : Essai debila n annuel du transfer de 208 (1968) l'azotedan s lebassi n dechass ed'Ostende . Ütermöhl,H. : ZurVervollkommnun g derquantita ­ I.Utilisatio n de l'azotepa r lephytóplanc - tivenPhytoplankto n Methodik.Mitt .int . tondan s lecycl ed'azote .Hydrobiol .Bull . Verein,theor .angew .Limnol .9 , 1-38 (1958) (Amsterdam) 8, 46-52 (1974a) - Essai debila n annuel dutransfe r del'azot e JanKuipe r dans lebassi n de chassed'Ostende . II.L e CentralLaborator y TNO rôle duzooplancto n dans lecycl ed'azote . Department ofMarin e Ecology Hydrobiol.Bull . (Amsterdam) 8, 53-66 (1974b) Ambachtsweg 8a Postbox 57 DenHelde r 1800 The Netherlands

Date of finalmanuscrip t acceptance:Jul y 8, 1977.Communicate d by0 .Kinne ,Hambur g

Al-11 Helgoländer wiss.Meeresunters . 30,652-66 5 (1977)

An experimental approach in studying the influence of mercury on a North Sea coastal plankton community

J. KUIPER

Centraal Laboratorium TNO; Delft, The Netherlands

ABSTRACT: The development of North Sea coastal plankton communities in four simultane­ ously filled plastic bags was followed for one month. To obtain a concentration of 5 ppb in the water phase a single dose of mercuric chloride was added to two of the bags.Thi s addition had a close impact on the development of the phytoplankton, while that on the zooplankton and the decomposers was less clear. In the course of the experiment, methylation of the added mercury proceeded in the sediment in the bags. The "plastic bag method" seems to be a suit­ able tool in toxicological research.

INTRODUCTION

Interpretation of the results of traditional toxicological laboratory experiments with single species, in terms of environmental significance, is difficult. Long-term experiments with more complex systems are urgently needed (Ringelberg, 1973; Lacaze, 1974; Kersting, 1975). Experimentation in the field, using toxic agents, encounters a number of practical difficulties. One way of bridging the gap between laboratory and natural conditions is to conduct experiments in large plastic bags housing plankton communities, the bags being suspended in natural water (Strick­ land & Terhune, 1961;Goldman , 1962; Schelske & Stoermer, 1971, 1972; Horstmann, 1972;Brockman n et al., 1974; Parsons, 1974). In 1975, we carried out, in the harbour of Den Helder (The Netherlands), three experiments with North Sea coastal plankton communities enclosed in plastic bags (contents approximately 1400 1/bag). The aim was to determine the influence of low concentrations of a pollutant on the structure and function of a marine plankton system. In the first two experiments, the development of the plankton community in four simultaneously filled bags showed, during 4-6 weeks, very similar patterns (cf. Takahashi et al., 1975), indicating that the method could be applied in toxicological research. The aim of the third experiment was to investigate the development of the plankton system, when subjected to a single dose of mercury. Only the results of this experiment are reported here.

A2-1 Influence of Hg on a plankton community 653

MATERIAL AND METHODS

Figure 1 shows the construction of the bag; it is derived from that used by Brock- man et al. (1974). Detailed information on the construction and operation of the bags will be published elsewhere. The experiment started August 25, 1975, and lasted one month. On the second day of the experiment, a single dose of mercuric chloride was added to two of the four bags in order to obtain a concentration of 5 ppb HgCU in the waterphase. To achieve this concentration, about 100 1wate r had been pumped from each bag into a PVC container, in which it was rapidly mixed with 0.7 1o f a 10pp m HgCls solution in water. Next, the mixture was at once pumped back into the bag through a perforated ring of PVC tubing (diameter 60 cm) which was slowly lowered into the bag to ensure thorough mixing of the HgClä solution with the original content of the bag.

plastic bag

aaa waear

Fig. 1 :Diagra m of aplasti c bag used for toxicological research

All samples, except those for zooplankton analysis, were collected daily at approximately 9 a. m., a non-metallic sampler being used at depths of 0.5 and 2.0 m. Zooplankton samples were collected twice a week; the sampling technique then used had been applied before (Kuiper, in prep.). The development of the phytoplankton was monitored by measuring chlorophyll a concentrations according to Strickland & Parsons (1968) and by inspection of Lugol- preserved samples, using an inverted microscope (Lund et al., 1958; Utermöhl, 1958). Cell volumes were computed from cell measurements assuming simple geometric forms (Vollenweider, 1969). The pigment extract for chlorophyll analysis was also used to measure the pigment index (D430/D665) described by Margalef (1965). In his opin­ ion, that index would be a measure for the diversity of the phytoplankton community.

A2-2 654 J. Kuiper

Primary production was measured by the 14C-method (Steemann-Nielsen, 1952); 100 ml samples were added to 1 ml of a NaHuCC>3 solution* in 125 ml bottles. After 6 h of incubation in the bags (a set of one light and one dark bottle was used at a depth of 0.5 m and another such set at 2.0 m), the bottles were transported to the laboratory in a dark box. Next, their contents were filtered through 0.45 /xm filters. Each filter was put in a scintillation solution (Anderson & Zeutschel, 1970; Pugh, 1973) and the activity measured with a Nuclear Chicago Mark I liquid scintillation counter. Correction for quenching was made by the channels ratio method. Computations of the 14C assimilation were corrected for ampoule activity and isotope discrimination. The development of zooplankton was followed by counting the organisms in the samples; we used the same procedures as Fransz (1976). With respect to the copepods, the nauplius, copepodite and adult growth stages of the different species were identi­ fied; growth and mortality rates were computed using the model of Fransz (1976). The development of decomposers in the waterphase was followed by measuring the differences in ATP concentrations (according to the method of Holm-Hansen, 1969) on 0.45 and 3^ m filters (Derenbach & Williams, 1974). Analyses of the concentrations of ortho-phosphate, nitrate, nitrite and reactive silicate were carried out as described by Strickland & Parsons (1968). The mercury concentration in the waterphase was measured by atomic absorption spectrometry, with an IRDAB HGM 2300 spectrophotometer. The mercury adsorbed on the walls of the bags, and the mercury concentration of the sediment at the end of the experi­ ment, were measured following Tjoe et al. (1973). Methylmercury concentrations in the sediment were determined according to Houpt and Compaan (1972) and Houpt (in preparation). The Royal Dutch Meteorological Institute (KNMI) was so kind as to provide us with the integrated daily values of global radiation, measured at the meteorological station "De Kooy". Standard deviations were estimated from the measurements using analysis of variance. Time series in different bags were compared by means of regression analysis. The significance of differences between the coefficients of these regression lines was tested using Students t. Students t was also used for testing the significance of differences between different measured values. Most statistical analyses were performed by the standard TSAM procedures of the Cyphernetics Information Corp. (CIC).

RESULTS AND DISCUSSION

Development of phytoplankton

The mercury concentrations measured in the morning of the third day of our experiment were the same as those calculated. Table 1give s the mercury concentrations in the two bags, as a function of time. Five days after the addition of mercury, the Hg concentration dropped below

* Ampoules with an activity of 3.8 /uCi ml"1 had been provided by the International Agency for 14Cdetermination , Denmark.

A2-3 Influence of Hg on a plankton community 655

Table1 Concentration of mercury in the waterphase (in ppb). Averages of duplicate samples taken at 0.5 and 2.0 m respectively, and their standard deviations (s.d.)

Day of ex­ Bag 1 Bag 4 periment HgCl2 concentration gCl2 concentration 0 <0.3 _ <0.3 _ 3 5.1 + 0.1 5.2 + 0.1 4 1.7 + 0.4 1.5 + 0.4 5 0.8 + 0.2 1.0 + 0.2 6 0.7 + 0.3 <0.3 - 7-32 <0.3 + - <0.3 - Addition of the HgCla occurred on the 2nd day of the experiment. the detection limit (0.3 ppb). As will be shown later, most of the mercury had probably been adsorbed by particles (cf. Smith et al., 1971) that settled on the bottom. Figure 2 shows the development of chlorophyll a in treated and untreated bags at a depth of 0.5 m.

chlorophyll olmg/m *] August -Seple mber 1975

o bag 1 mercury polluted • bog J. » bog 2 control a bag 3 samples 05m

Fig.2: Chlorophyl l aconcentration , at a depth of 0.5 mi n mercury-polluted bagsan d non-polluted controls

Figure 3 gives the cell volume per liter of water for the dominant species in the phytoplankton. The water that was used to fill the bags contained a bloom of different diatom species from which Skeletonema costatum and Bacteriastrum kyalinum con­ tinued growing in the bags, reaching a maximum on the second day, when mercury was added to the bags. Chlorophyll a concentrations in the polluted bags declined faster on the 3rd day of the experiment than in the unpolluted controls. On the fourth day, the chlorophyll concentration in the polluted bags increases, while that in the controls continues to decline. In the controls, a second bloom begins on the 6th day and reaches a maximum on the 10th day. This peak is mainly due to the colony-

A2-4 656 J. Kuiper forming //-flagellate Phaeocystis pouchetii. In the polluted bags, the beginning of the second chlorophyll peak is delayed until the 9th day, reaching a maximum on the 14th day. This peak is mainly due to free-living //-flagellates and, for a minor part, to P. pouchetii. The second growth peak in the controls is followed by a minimum on the 15th day, and there is a third maximum on the 21st day. After this day, chlorophyll con­ centrations in the controls remain high. In the polluted bags, the second maximum is also followed by a less pronounced minimum and a new increase due to some species of //-flagellates.

Species Asterionellajaponica Bactenastrumhyalinum 0 Biddulphia spp ppm 2 °y £ Chaetoceros spp volume Euzampia zoodiacus 6 Rhizosolenia spp Hitzschia longissimo

Thotassiosira spp Skeletonema costatum

Dmoflogellates

ju flogeltotes

Phaeocystis pouchett, time , days

5 '0 15 20 25 30 *w\ controls >, /mercury polluted Fig.3 : Developmen t of cell volume of main phytoplankton species in mercury-treated bags and controls

Compared to the standard deviation (s.d. ) of the chlorophyll measurements itself (factor 1.08), the decline in chlorophyll concentrations on the 3rd day, and the in­ crease on the 4th day in the polluted bags, are significantly (p < 0.05) faster than in the controls. This means that, at concentrations higher than 1.5 ppb, mercury increases the rate of decline of the phytoplankton. In other words, mercury then reduces the growth rate of the phytoplankton. The small peak on the 5th day is probably caused by growth of nutrients not yet used on the 3rd day. Clendenning (1958) found inhibition of the growth rate of Macrocystis pyrifera to occur at a concentration of 50 ppb HgCfe. For Scenedesmus sp. Bringmann & Kuhn (1959) found growth rate inhibition at a concentration of 30 ppb. Ben Bassat et al. (1972) showed a lag of a few hours in the growth curve of Chlamydomonas reinhardii to be an effect of a con­ centration of 100 ppb mercuric chloride. Nuzzi (1972) reports inhibition of growth rates of various phytoplankton species by 7 ppb HgC^. The concentration of 1.5 ppb

A2-5 Influence of Hg on a plankton community 657 that we found is also low compared with those occurring locally in polluted areas (f'onds, 1971; Smit h et al., 1971; Kecke s & Miettinen, 1972). A second effect of mercury addition is that, in the polluted bags, the second maximum is delayed four days, i. e. when Hg concentrations are no longer measurable. Ben Bassat et al. (1972) and Tompkins & Blinn (1976) also found lag phases, as an effect of addition of HgClg. Ben Bassat et al. (1972) suppose that the lag phase ends because the algae need time to convert the mercury, adsorbed by their cell walls, into a form which is not harmful to them. In our experiment, the possibility that minerali­ zation is retarded in the mercury-polluted bags seems more likely. A third effect of the mercury pollution is that the composition of the community of the second peak in the polluted bags differs from that in the controls. This is important because shifts in the community composition of the phytoplankton can have effects on higher trophic levels (Fisher ScWurster , 1974). Analysis of variance showed that, during the experiment, concentrations of chlorophyll at depths of 0.5 and 2.0 m did not differ significantly (p < 0.05).

Fig. 4: Phytoplankton production (expressed as mgC production per ms water) in different bags at a depth of 0.5 mfro m 9-15h

Figure 4 shows the phytoplankton production in the individual bags at the 0.5 m level. The pattern corresponds to that shown by the chlorophyll a concentrations (Fig. 2). The s. d. of the measurements is a factor of 1.28 and, therefore, too large to detect significant differences in the rate of decrease of the first peak between the polluted bags and the controls. Contrary to results of the chlorophyll concentration, there is a marked effect of depth on the 14C-assimilation. Production at a depth of 0.5 m is almost always higher than at 2.0 m. Cadée & Hegeman (1974) state that the growth rate of wadden-sea phytoplankton is inhibited below light energy values of 0.84 J cm"ämin_1. Taking a

A2-6 658 J. Kuiper day at 10 hours, inhibition results at values smaller than 500 J cm~2dayn. The values at a depth of 2.0 m in the bags, computed from the light values and the secchi-disc readings, according to the formulas given by Parsons & Takahashi (1973) and Cadée & Hegeman (1974), are almost every day below this limit. If the growth rate of the phytoplankton varies with depth, and biomass and other physico-chemical parameters do not, this indicates that the water volume in the bags is well mixed. This mixing is probably caused by the small differences in temperature observed in the bags.

Margalef index mercury polluted i controls

-*- timefdoys I

0 S 10 15 20 25 30

Fig.5 : Margale f index in different bags.Averag e from 0.5 and 2.0 m

Figure 5 shows the value of the Margalef pigment index at different times of the experiment. No significant difference (p < 0.05) was found between the measurements at depths of 0.5 and 2.0 m. Therefore, the average value of these two measurements is presented. Thé same pattern is found as in earlier experiments: low values at chloro­ phyll maxima, high values at chlorophyll minima. Carreto & Catoggio (1976) obtained the same results working with cultures of Phaeodactylum tricornutum. They reported that the index describes in a simple way the variations in pigment proportions, in­ creasing in value as the proportion of phytosynthetically active pigments (like chloro­ phyll a and c) decreases. An interesting point is that, for a long period, values in the controls remain higher than in the polluted bags. This indicates that, apart from the physiological state, another factor plays a role in determining the value of the index. Probably the species structure (qualitatively, not quantitatively) is important. The development of the concentrations of the various nutrients (averaged for two depths and corresponding bags) is given in Figure 6. It shows the same pattern as in earlier experiments. The first phytoplankton peak consumes nutrients till some nutrient reaches growth-rate limiting values. This time, this role seems to be taken by the nitrogen-containing compounds (Mclsaac & Dugdale, 1968; Eppley et al., 1969; Epp- ley & Thomas, 1969; Kuiper, in prep.). After the first phytoplankton peak, nutrient concentrations stay low. The development of the phytoplankton in separate bags shows similar patterns under identical environmental conditions; this confirms earlier results (Kuiper, in prep.) and is in agreement with results of Takahashi et al. (1975).

A2-7 Influence of Hg on a plankton community 659

NO,-N

6.0 NOfNIttgot/ll

+ mercury polluted reocHwsilicote lfigol/11 50 I • controls

4.0 \ JO Y 20 \ 10 \ i -4^-k=£=4T

Fig.6 : Concentration of nutrients in the polluted bags and the controls. Average of twodepth s and replicate bags.Vertica l bars indicate ± 1 standard deviation

Development of zooplankton

Apart from the calanoid copepods Temora longicornis, Acartia clausi, Centropa- ges hamatus and Paracalanus parvus (the latter one only in small numbers), also a harpacticid copepod, Euterpina acutifrons, develops in the bags (no clear distinction could be made between copepodites and adults of E. acutifrons). Larvae of Spio sp. sometimes formed part of the zooplankton-biomass, which cannot be ignored. Counts of these larvae showed a much higher s. d. than could be expected, which is probably caused by the fact that these larvae are sometimes concentrated near the bottom, so that sampling was not adequate. Further the larvae of bivalves, periwinkles and nauplii of barnacles were found in small numbers, as well as a few Podon sp. The development of C. hamatus, on which the influence of the addition of the mercury was most clear, isshow n in Figure 7. In the controls and in the polluted bags, the numbers of organisms generally increased sharply. A development of the different growth stages was seen for most species. Table 2 gives the results of the analysis of the differences in population densi­ ties between the polluted bags and the controls. The densities of T. longicornis and C. hamatus are significantly lower in the polluted bags than in the controls; as from

A2-8 660 J. Kuiper the 11th day of the experiment, the density of E. acutijrons is significantly higher than in the controls. To investigate whether these differences in density are caused by either increased mortality or a lower growth rate, the latter rates were estimated using the model of Fransz (1976). A sex ratio of 1.0 was assumed, and in the event of possible

number per liter

nauplii 01 100

• mercury polluted - controls

copepodites

Fig. 7:Developmen t of Centropages hamatus. Average from replicate bags (number of nauplii, copepodites and adultspe r liter) negative development rates, the same corrections were applied as used by Fransz (1976). Table 3 gives the relative development and mortality rates of C. hamatus together with one half of the 95 °/oconfidenc e interval of these estimates. The estimates are of the same order of magnitude as Fransz (1976) found in the North Sea; this indicates that the copepods in the bags develop in a natural way. Due to the large s. d. of the individual counts on which the rate estimates are based, the s. d. of these rates is also very large. Differences between the development and mortality rates in different bags are never significant. However, the differences between densities of C. hamatus, in the polluted bags and in the controls, coincide

A2-9 Influence of Hg on a plankton community 661

Table 2 Analysis of differences between the average population density of different organisms in the polluted bags and the controls. Analysis done with the sign-test (Wijvekate 1972) + significant difference — no significant difference. Hg < C means: in mercury-polluted bags, the popu­ lation density is lower than in the controls, n = nauplii; c = copepodites; a = adults

Species Stages Differences Experimental conditions

Temora longicornis n + p = 0.02 Hg

Acartia clausi n _ c + p = 0.05 Hg C (since 11th day of experiment) Paracalanus parvus n + p = 0.10 Hg

Table 3 Estimates of the relative development and mortality rate of Centropages hamatus. In brackets, one half of the 95 °/o confidence interval is given

Bag Relative development rate Relative mortality rate no. nauplii copepodites nauplii copepodites adults

1 mercury-contaminated 0.16(0.27) 0.00(0.00) -0.05(0.31) 0.14(0.30) 0.30(0.51) 2 control 0.20 (0.14) 0.02(0.03) -0.04(0.41) 0.06(0.10) 0.40(0.84) 3 control 0.19(0.22) 0.02(0.04) -0.13(0.30) 0.04(0.14) 0.51(0.96) 4 mercury-contaminated 0.12 (0.06) 0.00(0.00) -0.03(0.15) 0.05(0.06) 0.25(0.57)

with lower development rates of nauplii and copepodites in the polluted bags. Accordingly, the added mercury probably inhibited the growth rate but did not raise the mortality of C. hamatus. Apart from the large s. d. of individual counts, the fact that mercury influences the zooplankton during a relatively short time only, may explain why the effects of added mercury are not more pronounced. If, for example, a small lag phase occurs in the growth of a copepod, which seems the case for C. hamatus and T. longicornis, then computation of development and mortality rates using all samples tends to hide the effect.

A2-10 662 J. Kuiper

Development of decomposers

The ATP measurements could not be used to evaluate the development of de- composers-biomass, which was foreseen by Jassby (1975) on theoretical grounds. How­ ever, the fact that the first phytoplankton peak, which in its growth depletes the nutrients, is followed by a second and a third, during which nutrient concentrations do not decrease considerably, indicates that, in the bags, mineralization occurs. The second phytoplankton peak in the polluted bags is delayed four days. This lag phase is probably not caused by the fact that the phytoplankton cells need time to transform the mercuric chloride into a chemically less harmful compound (Ben Bassat et al., 1972). Concentrations of mercury in the waterphase are very low and, already on the 5th day, there was phytoplankton growth in the polluted bags. Probably the lag phase in the phytoplankton is caused by a lag phase in the growth of the decom­ posers, i. e. by a reduced rate of mineralization. Inhibition of bacterial growth requires mercury concentrations in the ppm range (Greeson, 1970); these concentra­ tions were found of the experiment in the sediment of the polluted bags.

"Fate" of added mercury

The decrease of mercury concentrations in the waterphase of the treated bags is given in Table 1, while Table 4 gives the results of mercury concentration measure­ ments in the sediment of the bags, collected at the end of the experiment (collection of sediment in Bag 3 failed).

Table4 Concentrations of mercury and methylmercury in sediments of treated bags and controls at the end of the experiment „ ~ , Tj . Concentration Concentration of methyl Ba8 V?taI "f "\ Hg (ppm based Hg (ppb based onwe t no. sediment (me) 8V. • ,.., •IC.N / J ico/\ v b' on wet weight) weight); (s.d. 15 °/o) 1 mercury-contaminated 0.19 5.4 2.3 4 mercury-contaminated 1.10 24.5 13.7 2 control 0.003 0.06 < 0.5

During the experiment, the mercury concentrated in the sediment to the ppm level. However, most of the added mercury (7 mg to each bag) could not be recovered. On the walls of the bags about 40 /ugha d adsorbed. If we assume that, till the end of the experiment, the mercury concentration in the waterphase remained 0.3 ppb - which is not very likely (see Smith et al., 1971) - in each bag 0.4 mg remained in the waterphase. In the sediment, 0.19 mg and 1.10 mg, respectively, were recovered. This means that 6.4 mg from Bag 1an d 5.5 mg from Bag 4 are not accounted for. A possible explanation may be the methylation of mercury in the sediment. Measurable amounts of methylmercury were found in the Hg-contaminated bags (Table 4), and the control showed no detectable amounts of this substance. It must be

A2-11 Influence of Hg on a plankton community 663

assumed that methylmercury was formed as a result of metabolical processes. As dimethylmercury is highly volatile, it is feasible that the mercury disappeared in the atmosphere via this derivate. Another explanation for the loss of mercury from the bags may be volatilization of the added mercury, this being the transformation into metallic mercury by bacteria (Kushner, 1974; Schottcl et al., 1974).

CONCLUSIONS

A single dose of 5 ppb mercuric chloride added to a North Sea coastal plankton community, enclosed by a plastic bag, resulted in: (a) disappearance of the mercury from the waterphase within five days, caused by adsorption on particles that had settled on the bottom of the bags; (b) a reduced phytoplankton growth rate, as long as mercury concentrations were higher than 1.5 ppb; (c) a delay in the beginning of the second phytoplankton bloom, probably due to a lag phase in the mineralization of organic matter in the sediment; (d) a change in the community structure of the phytoplankton of the second bloom; (e) lower population densities of two copepod species, probably caused by a reduced growth rate in the period following the mercury addition; (f) methylation of the added mercury in the sediment. The "plastic-bag method" seems to be a suitable tool in toxicological research.

Acknowledgements.Thank s are due to my colleagues De Koek, Hoornsman, Roele, Schrieken, Van den Eikhoff and Van het Groenewoud; they assisted in carrying out the various analyses and in evaluating the results. Acknowledgement is also due to Dr Hueck, Head of the Centraal Laboratorium's Department of Biology, for comments. The assistance of research workers of the Departments of Biology, Physics and Analytical Chemistry is greatly appre­ ciated. I also like to thank Drs. Cadee, De Vries, Fransz, Gieskes, Hegcman and Leeuwis of the Netherlands Institute for Sea Research for their help and advice on the various aspects related to the work reported here. Finally, I would like to thank the authorities of the Royal Netherlands Navy for their kind cooperation in supplying experimental facilities. This work lias been carried out under Contract No. 110-75-1 ENV N of the E. C. Environmental Re­ search Programme.

LITERATURE CITED

Anderson, G. C. & Zeutschel, R. P., 1970. Release of dissolved organic matter by marine phytoplankton in coastal and off-shore areas of the N.E. Pacific Ocean. Limnol. Oceanogr. 15,402-407 . Ben-Bassat, D., Shelef, G., Grüner, N. & Shuval, H. J., 1972. Growth of Chlamydomonas in a medium containing mercury. Nature, Lond. 240,43-44 . ßringmann, G. & Kuhn, R., 1959. The toxic effects of waste water on aquatic bacteria, algae and small crustaceans.Gesundheitsingenieu r 80,115 . Brockmann, U. H., Eberlein, K., Junge, H. D., Trageser, H. & Trahms, K. J., 1974. Einfache Folientanks zur Planktonuntersuchung in situ. Mar. Biol.24 ,163-166 . Carreto, J. I. & Catoggio, J. A., 1976. Variations in pigment contents of the diatomPhaeo- dactylumtricornutum during growth. Mar. Biol.36 ,105-112 .

A2-12 664 J. Kuiper

Cadée, G. C. & Hegeman, J., 1974. Primary production of the benthic microflora living on tidal flats in the Dutch Wadden Sea. Neth. J. Sea Res. 8, 260-291. Clendenning, K. A., 1958. Laboratory investigations. Effects of discharging wastes on kelp 1957-58. A. Rep. Inst. mar. Res. Univ. Calif. 58-11, 29-35. Derenbach, J. B. & Le B. Williams, P. J., 1974. Autotrophic and bacterial production: frac­ tionation of plankton populations by differential filtration of samples from the English Channel. Mar. Biol. 25, 263-269. Eppley, R. W. & Thomas, W. H., 1969. Comparison of half-saturation constants for growth and nitrate uptake of marine phytoplankton. J. Phycol. 5, 375-379. Eppley, R. W., Rogers, J. N. & McCarthy, J. J., 1969. Half-saturation constants for uptake of nitrate and ammonium by marine phytoplankton. Limnol. Oceanogr. 14, 912-920. Fisher, N. S. & Wurster, C. F., 1974. Impact of pollutants on plankton communities. Envi- ronm. Conservation 1, 189-190. Fonds, A. W., 1971. Kwik in de Nederlandse oppervlakte-wateren. TNO-Nieuws 26, 375-377. Fransz, H. G., 1976. The spring development of calanoid copepod populations in the Dutch coastal waters as related to primary production. In: Proceedings of the 10th European Symposium on Marine Biology. Ed. by G. Persoone & E. Jaspers. University Press, Wette­ ren, 2, 247-269. Goldman, C. R., 1962. A method of studying nutrient limiting factors in situ in water columns isolated by polyethylene film. Limnol. Oceanogr. 7, 99-101. Greeson, P. E., 1970. Biological factors in the chemistry of mercury. Geol. Surv. Water Supply Paper 173, 32-34. Holm-Hansen, O., 1969. Determination of microbial biomass in ocean profiles. Limnol. Oceanogr. 14, 740-747. Horstmann, U., 1972. Über den Einfluß von häuslichem Abwasser auf das Plankton in der Kieler Bucht. Kieler Meeresforsch. 28, 17-198. Houpt, P. M. & Compaan, H., 1972. Une nouvelle méthode spectrographique (emission) pour l'identification de traces de matières organiques contenant des halogènes et du mercure dans les fractions obtenus par Chromatographie en phase gazeuse. Analusis 1, 27-33. Jassby, A. D., 1975. An evaluation of ATP estimations of bacterial biomass in the presence of phytoplankton. Limnol. Oceanogr. 20, 646-648. Keükes1, S. & Miettinen, J. K., 1972. Mercury as a marine pollutant. In: Marine pollution and sea life. Ed. by M. Ruivo. Fishing News Books, London, 276-289. Kersting, K., 1975. The use of microsystems for the evaluation of the effect of toxicants. Hydrobiol. Bull. 9, 102-108. Kushner, D. J., 1974. Microbial dealings with heavy metals. In: Proceedings of the Inter­ national Conference on Transport of persistent chemicals in aquatic ecosystems. Eds: A. S. W. de Freitas, D. J. Kushner, S. U. Qadri. Nat. Research Council of Canada, 2, 59-63. Lacaze, J. C, 1974. Ecotoxicology of crude oils and the use of experimental marine eco­ systems. Mar. Pollut. Bull. 5,153-156. Lund, J. W. G., Kipling, C. & Le Cren, E. D., 1958. The inverted microscope method of estimating algal numbers and the statistical basis of estimations by counting. Hydro- biologiall, 143-170. Mclsaac, J. J. & Dugdale, R. C, 1969. The kinetics of nitrate and ammonia uptake by natural populations of marine phytoplankton. Deep Sea Res. 16, 45-57. Margalef, R., 1965. Ecological correlations and the relationship between primary productivity and community structure. Mem. 1st. Ital. Idrobiol. 18, 355-364. Nuzzi, R., 1972. Toxicity of mercury to phytoplankton. Nature, Lond. 237, 38-39. Parsons, T. R., 1974. Controlled ecosystem pollution experiment (CEPEX). Environm. Con­ servation 1,224 . — & Takahashi, M., 1973. Biological océanographie processes. Pergamon Press, Oxford, 186 pp. Pugh, P. R., 1973. An evaluation of liquid scintillation counting techniques for use in aquatic primary production studies. Limnol. Oceanogr. 18, 310-319.

A2-13 Influence of Hg on a plankton community 665

Ringelberg, J., 1973. Parameter-dependent (temperature) tolerance levels and the influence of the complexity of the biological system. Hydrobiol. Bull. 7, 106-114. Schelske, C. L. & Stoermer, E. F., 1971. Eutrophication, silica depletion and predicted changes in algal quality in Lake Michigan. Science, N.Y. 173 (3995), 423-424. — — 1972. Phosphorus, silica and eutrophication of Lake Michigan. In: Nutrients and eutro­ phication. Ed. by G. E. Likens. Am. Soc. Limnol. Oceanogr., Lawrence, Kan., 157-170. (Spec. Symp. Vol. 1.) Schottel, J., Mandai, A., Toth, K., Clark, D. & Silver, S., 1974. Mercury and mercurial resist­ ance determined by plasmids in Escherichia coli and Pseudomonas aeruginosa. In: Proceed­ ings of the International Conference on Transport of persistent chemicals by aquatic eco­ systems. Eds: A. S. W. de Freitas, D. J. Kushner, S. U. Qadri. Nat. Research Council of Canada, 2, 65-71. Smith, J. D., Nicholson, R. A. & Moore, P. J., 1971. Mercury in the water of the tidal Thames. Nature, Lond. 232, 393-394. Steemann-Nielsen, E., 1952. The use of radioactive carbon (C14) for measuring organic pro­ duction in the sea. J. Cons.Perm . int. Explor. Mer. 18, 117-140. Strickland, J. D. H. & Terhune, L. D. B., 1961. The study of in situ marine photosynthesis using a large plastic bag. Limnol. Oceanogr. 6, 93-96. — & Parsons, T. R., 1968. A practical handbook of sea-water analysis. Bull. Fish. Res. Bd Can. 167,1-311. Takahashi, M., Thomas, W. H., Seibert, D. L. R., Beers, J., Koeller, P. & Parsons, T. R., 1975. The replication of biological events in enclosed water columns. Arch. Hydrobiol. 76, 5-23. Tjioe, P. S., de Goeij, J. J. M. & Houtman, J. P. W., 1973. Automated chemical separations in routine activation analysis. J. radioanalyt. Chem. 16, 153-164. Tompkins, T. & Blinn, D. W., 1976. The effect of mercury on the growth rate of Fragilaria crotonensis Kitton and Asterionella formosa Hass. Hydrobiologia 49, 111-116. Utermöhl, H., 1958. Zur Vervollkommnung der quantitativen Phytoplankton-Methodik. Mitt. int. Verein. Limnol. 9, 1-38. Vollenweider, R. A. (Ed.), 1969. A manual on methods for measuring primary production in aquatic environments. Blackwell, Oxford, 225 pp. Wijvekate, M. L., 1972. Verklarende statistiek. Het Spectrum, Utrecht, 251 pp.

Author's address: J. Kuiper Centraal Laboratorium TNO P.O.B. 57 Den Helder The Netherlands

A2-14 F.COTOX1COLOGY AND ENVIRONMENTAL SAFETY S, 106-134 (1981)

Fate and Effects of Mercury in Marine Plankton Communities in Experimental Enclosures1

JAN KUIPER

Laboratoryfor AppliedMarine Research,MT-TNO, P.O. Box 57, 1780 ABDen Helder, The Netherlands ReceivedJune 30, 1980

Thedevelopmen t ofa NorthSe acoasta l planktoncommunit yexpose dt odifferen t degrees ofmercur ystres si nsi xsimultaneousl y filled plasticbag swa sfollowe d for4 4days .Mercuri c chloridewa sadde dt ofou rbag st oyiel dconcentration so f0. 5(on ebag) ,5 (twobags) ,an d SO /ig Hg- liter-1 (onebag )i nth ewater . Twobag sserve d ascontrols . Itcoul d beshow n that a singledos eo f0. 5/x gHg -liter "' altere dth especie scompositio no fth ealga lcommunit yo nth e wallso fth ebag . Additiono f5 /u.g Hg liter' hada marke dimpac t onth edevelopmen t ofth e phytoplankton, as well as on that of the zooplankton and decomposers. Addition of 50 /ug Hg- liter -'cause dinactivatio no rdeat ho fth ephytoplankto nan dzooplankton .Th etoxicit yo f mercuricchlorid et oth ephytoplankto ndepend so n(a )th econcentratio n ofth emercury , and (b)th eparticl e concentration, i.e., thesurfac e areaavailabl efo radsorptio n ofmercury .Fo r this second reason the ratio between numbers of living cells and inanimate particles is an important factor influencing mercury toxicity in aquatic ecosystems. Methylation of the addedmercur y occurs inth e sediment ofth ebag safte ra la gphas e of 1 month. Mosto fth e added mercury disappears from the system by volatilization to metallic mercury. The remainder is absorbed by the sediment and the walls of thebags .

Contents. 1. Introduction.2. Materialand Methods. 2.1. Experimental handling. 2.2. Analytical methods. 2.3. Statistics. 3. Results. 3.1.Developmen t ofth ephytoplankton . 3.2. Development ofth e zooplankton. 3:3.Developmen t ofdecomposers . 3.4. Fate ofth eadde d mercury. 3.5. Algalgrowt ho nth ewalls .4 .Discussion. 4.1. Development induplicat ebags . 4.2.Effect so fmercur yo nth ephytoplankton .4.3 .Effect so fmercur yo nth ezooplankton .4.4 . Effects ofmercur yo nth edecomposers . 4.5. Phytoplankton-zooplankton interactions. 4.6. Fate of the added mercury. 5. Conclusions.

INTRODUCTION Toxicological investigations are usually performed in the laboratory and the interpretation ofthei r results interm so fenvironmenta l significance isdifficul t for severalreason s(e.g. ,Gray , 1974). Inorde rt oasses sth evalu eo fexperiment si nth e laboratory for determining safe "no-effect" levels inth efield, ther e is a need for experiments with more complex systems that can be regarded as approximat­ ing more closely field conditions (Ringelberg, 1973; Lacaze, 1974; Ketchum et al., 1975). Severalinvestigator s inaquati cecolog y andecotoxicolog y uselarg eplasti cbag s suspended innatura lwater st ostud yth erelationship s betweenth eplankto nan dit s environment, and sot o narrow thega pbetwee n laboratory and natural conditions (Strickland and Terhune, 1961; Horstmann, 1972;Reev eet al., 1976;Menze l and

1 This work was carried out under Contract 227-77-1-ENV N of the EC Environmental Research Programme.

0147-6513/81/010106-29$02.00 106 Copyright © 1981b y Academic Press, Inc. All rights of reproduction in any form reserved. A3-1 EFFECTS OF MERCURY ON MARINE PLANKTON 107

Case, 1977;Davie s and Gamble, 1979). In 1974thi s type of research was started using Dutch coastal water plankton communities. Experiments with nonpolluted systemsshowe dtha tth emetho do fenclosin ga plankto ncommunit y ina plasti cba g can be used in toxicological research (Kuiper, 1977a), and further investigations (Kuiper, 1977b) were aimed at determining the effects of pollutants in low concentrations on the development of the enclosed system. The model pollutants were added in single doses, since this practice more closely simulates the "normal" field situation, where the source of a pollutant usually is an outfall, a river, or a dumping event (Menzel and Case, 1977). This paper describes an experiment in 1976i n which the influence of mercuric chlorideo nth edevelopmen t ofa plankto n communitywa sstudied .A singledos eo f mercury was added to several bags at three concentrations. Since conversion of inorganic mercury compoundsint omor etoxi cmethylate dform s isknow nt ooccu r in natural waters (Bryan, 1976),som e attention was paid to the possibility of this process occurring in the experimental bags.

2. MATERIAL AND METHODS 2.1. EXPERIMENTAL HANDLING Theexperimen twa sstarte do n2 9Marc han dconclude do n 12Ma y 1976. Sixbag s were filled simultaneously with about 1400 liters of North Sea coastal water collected afe w milesoffshore . Thebag swer eanchore d neara raf t ina quie t corner of the harbor of Den Helder. Large predators (Ctenophora, Cephalopoda) were prevented from entering the bags by filtration of the water through a 2-mm net. Details of the construction of the bags and the operation procedures have been described by Kuiper (1977a). On Day 7 (the start of an experiment being designated Day 0) single doses of mercuricchlorid ewer eadde dt ofou ro fth ebag st ogiv einitia lconcentration so f0.5 , 5,5, and5 0/i gH g• liter "' , respectively.T othi send , 100liter so fwate rwa spumpe d from eachba gint oa PV Ccontaine ran drapidl ymixe dwit h 1 litero fa concentrate d solution of mercuric chloride inacidifie d water. This mixture wasa t oncepumpe d backint oth eba gthroug ha PV Csprinkler ,whic hwa sslowl ylowere dint oth eba gt o ensurethoroug h mixingo fth e solution withth eseawate r inth ebag .Th ebag swit h initial concentrations of 0.5, 5, 5, and 50/x gHgliter" 1 will be referred to as HI, H5a, H5b, and H50 respectively. Two bags serving as controls were given the same treatment, except that nopollutan t was added to their contents. These bags are denoted CI and C2. The development of the system in the bags was monitored by means of the parameters listed in Fig. 1.Al l samples, except those of zooplankton, were taken daily, as a rule at 9 AM, with a nonmetallic sampler consisting of two 1250-ml chambers that could beopene d and closed atan ydesire d depth. Toinvestigat eth e variationo fselecte dparameter swit hdepth ,sample swer ealway stake na tdepth so f 0.5an d2. 0m .Zooplankto n sampleswer ecollecte d bymean so fa pip e(lengt h3 m , diameter4 cm )wit ha bal lvalv ea tth eend ,samplin gnearl yth eentir ewate rcolum n (0-2.5 m). Each sample (16liters ) consisted offive rando m lowerings of the pipe intoeac h bag.Th e contents of the sampler werefiltered throug h a55-/u. mnet ,an d the samples at oncefixed an d preserved ina 4%formaldehyd e solution in filtered seawater.

A3-2 108 JAN KUIPER

vV -ja

phytoplankton nutrients

chlorophyll-* P04-P, NOj-N, N02-N, phtcopigntnts Jffl,-N,reactiv eS i "•C-MtiniUtion 02 number of algal cells •pteit s distribution U 430 /D 665 decomposers

relative nunbero f organisms inwate r and sediment

number of organisms species distribution system characteristics

tempereture,pH ,salinity , Secchi disc visibility, concentration added pollutant invate r and sediment

FIG. 1. Simplified diagram of the interrelations between the different trophic levels and the abiotic factors influencing thedevelopmen t ofth eorganism so nthes e trophiclevel s ofth eplankto ncommunit y enclosed by the plastic bags. Also indicated are the parameters by which changes in the system are monitored.

The amount and species composition of the periphyton fouling the walls of the bags were assessed from the algal biomass collecting on glass and polyethylene microscope slidesmounte d inframe swhic hwer elowere dint oth ebags .Slide swer e removed twice a week for microscopical inspection and determination of chlorophyll. Sedimenting material was collected in traps consisting of polythene funnels (diameter 12 cm), whose shafts had been stoppered, and which were suspended from the start of the experiment at a depth of 2.25 m. One sample consisted of the contents of one trap.

2.2. ANALYTICAL METHODS

The chlorophyll concentration of 1-liter samples was measured according to Strickland and Parsons (1968). Before the samples werefiltered throug h Whatman GF/Cfilters, 1 mlo fa 1%magnesiu mcarbonat e suspensionwa spoure do nth e filter. Thefilters wer etransferre d to 10 ml 90% acetonean dafte r 1 mino fsonificatio n with a Bransonultrasoni cgenerator , theremnan twa s extracted for2 4h r in arefrigerator . After centrifugation, pigment concentrations were measured with a Vitatron MPS photometer system using 1-cmcells . Concentrations of chlorophyll and phaeopig-

A3-3 EFFECTS OF MERCURY ON MARINE PLANKTON 109 ments were calculated according to the equations given by Lorenzen (1967). When chlorophyll concentrations were less than 1 mg-m~3 the fluorometer unit of the Vitatron was used, and the instrument calibrated with a sample of known chlorophyll concentration. To determine the amount ofchlorophyl l on the microscope slides collected from the bags, the backs of the slides were cleaned and the slides transferred to glass vessels and broken. To each vessel, 10m lo f 90%aceton e was added and after 1 min of sonification and 24 hr of extraction, the chlorophyll content of the extract was measured. The samples of phytoplankton were preserved with Lugol's iodine (Vollenweider, 1969) and examined with a Zeiss inverted microscope (Ütermöhl, 1958). The main species were identified where possible using nomenclature given by Drebes (1974) and Ingram Hendey (1964). The concentration and size distribution of suspended particulate matter were measured with aCoulte r counter, ModelT A II, witha populatio n accessory, usinga 100 or 280-Mm tube or both (Sheldon and Parsons, 1967; Gamble et al., 1977). Primary production was measured by Steemann-Nielsen's (1952) 14C method. H Samples of 100 ml were added to 1 ml of NaH C03 solution (ampoules with an activity of 3.6 /iCi-ml-1 were supplied by the International Agency for UC determination, Horsholm, Denmark) in 125-mlligh t and dark bottles. One light and one dark bottle were suspended at depths of0. 5 and 2.0 mi n the bags. After 6h r of incubation (9 AM-3 PM) the bottles were taken to the laboratory in a dark box, and their contents filtered through 0.45-ju.m membrane filters (Sartorius No. 11306). Each filter was put in a counting vial containing 10 ml of a scintillation solution (Anderson and Zeutschel, 1970;Pugh , 1973). The vialswer e counted witha Packard Tri-Carb liquid scintillation counter. The inorganic carbon content ofth e water was determined by titration according to Strickland and Parsons (1968). The concentrations of orthophosphate, ammonia, nitrate, nitrite, and reactive silicate were measured with aTechnico n autoanalyzer according to Strickland and Parsons (1968) and Technicon procedures. The zooplankton was counted, identified, and measured by the procedures described by Fransz (1975). Subsamples of the 16-liter sample were examined with a projection microscope until atleas t 150organism s had been counted. The changes of the population densities in the various stages of development of the copepods, whichalway sfor m the major part ofth e zooplankton biomassi nth e bags, were used to estimate development and mortality rates of selected species, as well as the production of organic matter by these organisms according to Fransz (1975). Development of the decomposers in the water and sediment was monitored by plating samples once a week on agar enriched with medium 2216 E according to Oppenheimer and Zobell (1952).Th e pH was measured with a Beekman Model 3550 pH meter. Other parameters measured include water temperature, salinity, Secchi disc visibility, oxygen concentration, and global radiation. The mercury concentration in the water was measured by atomic adsorption spectrometry, with an IRDAB HGM 2300 spectrometer, according to Spijk (1975). Mercury concentrations in the sediment were measured according to Tjoe et al. (1973). Methylmercury concentrations in the sediment were determined according to Houpt (1976). Bioassay tests were used toestablis h whether any of the nutrients measured was limitingphytoplankto n growth during peaks. Inthes e testsa mixe d sampleo f5 liter s

A3-4 110 JAN KUIPER

(2.5-liter samples from depths of 0.5 and 2.0 m)wa s distributed over sixteen 500-ml conical flasks in portions of 250 ml each. The water in 14flask s was enriched with -1 different combinations of NaN03 (+30 /Amol-liter ), Na2HP04 (+2.5 /umolli- -1 -1 ter ), and NaSi03-9H20 (+5 /nmol-liter ). Two flasks served as controls. The remainder ofth e 5-liter sample wastake n toth e laboratory for anumbe r ofanalyses . The flasks were left in a container on the raft and were cooled by harbor water pumped through the container. Two days after the start of the test, the contents of the flasks were sampled, and inspected with a microscope, a Coulter counter, or both for possible growth after addition of the nutrients.

2.3. STATISTICS Most statistical analyses were performed on the CDC 6400 computer of IWIS-TNO, The Hague. Computations for the zooplankton analyses were conducted with the CDC 6600o f the Nuclear Centre at Petten (with the help of Dr. Fransz of the Netherlands Institute of Sea Research, Texel). For all cases where no confidence level is given, P < 0.05 was used.

3. RESULTS

3.1. DEVELOPMENT OF THE PHYTOPLANKTON General Remarks and Phytoplankton Biomass The water with which the bags werefilled containe d aconsiderabl e amount of silt and detritus (Secchi disc-visibility 0.5 m). The salinity was 30.2%«, indicating that about 11% of the water came from freshwater sources. Figure 2 shows the mercury concentrations in the water in the bags. Concentrations in all bags rapidly fell below the detection limit except in bag H50.

, - —O-O-Q.

0— —S

March *MQy J9?6 05 vgHg. • 5 .-! ft 5 a* 50

30 1.0

FIG. 2. Development of mercury concentrations in the watero f the pollutedbag s(averag e0. 5 and 2.0 mi n depth).

A3-5 EFFECTS OF MERCURY ON MARINE PLANKTON 111

iOO

200

too

50

20

10

Marzh-Moy 1976 g controls t OS/ugHgf1 *S o 50

40

FIG. 3. Development of chlorophyll concentrations in the different bags (average 0.5 and 2.0 m in depth).

Moredetail sabou tth efat e ofth eadde d mercuricchlorid ear egive ni nSectio n 3.4. The development of phytoplankton measured by chlorophyll concentration is shown in Fig. 3.Th egenera l pattern issimila r totha t found inearlie r experiments (Kuiper, 1977a,b). Duet o diminished turbulence inside the bags, part of the solid matter settled onth e bottom after the bags werefilled , witha resultin g decreasei n chlorophyll concentration and increase in Secchi disc visibility to about 2.0 m. After 2day schlorophyl lconcentration s begant orise owin gt ogrowt ho fdiatom s (Chaetoceros diadema and other Chaetoceros spp.) and microflagellates. The addition of 5 and 50 /ig-liter -1 mercury on Day 7 resulted in a decrease in chlorophyll concentrations. Phytoplankton growth was inhibited. This was not observed in bag HI. Considering the standard deviation of the chlorophyll measurements (15%),chlorophyl lconcentration s inH I onDay s8,9 , 10an d 11 are not significantly less than those in the controls. In these, and in bag HI (0.5 /u,g Hg• liter -1), thechlorophyl lconcentratio n reachesa maximu mo nDa y 12,fall st oa minimumo nabou tDa y25 ,afte r whichi tslowl yrise sagai nowin gmainl yt ogrowt h of microflagellates and dinoflagellates. As mercury concentrations inth e water decrease further, growth starts againi n the 5an d 50jig -liter -1 bags.Th e steep growth maximum in bags H5a and H5b( 5 /tig-liter-1) is delayed by 9 days, and in bag H50 (50 /itg-liter-1) by 29day s with respect to the controls. In bags H5aan d H5bthi s maximum iscause d by different organisms. In one bag most of the biomass consisted of Thalassiosira nordenskioldiitogethe rwit hlarg eflagellates an dmicroflagellates . Inth eothe rba gi t consisted ofth elatte rtw ogroup so forganism sonly .I nba gH5 0th egrowt hpea ki s duet odiatom s( Th. nordenskioldii andSkeletonema costatum) an di nmino rpar tt o large flagellates. Figure 4 shows the concentration of phaeopigments in the course of the experiment. Phaeopigments are degradation products of chlorophyll. Their concentrationsa tth estar to fth eexperimen twa srelativel yhigh ,bu tdecrease dwit h a factor of 2pe r day down to0. 1 mg-m -3 on Day 4. An increase in phaeopigment

A3-6 112 JAN KUIPER

FIG. 4. Development of concentrations of phaeopigments in different bags (average 0.5 and 2.0 mi n depth). concentration always coincides with a decrease in chlorophyll concentration. Thedevelopmen t inH5 adiffer s from that inH5b . After the secondgrowt hpea k of the phytoplankton, phaeopigment concentrations in one bag rose to only 0.4 mg-m"3 on Day 28, while the concentrations in the other bag increased to 1.2 mg-m~ 3 and remained high until Day 37.Th edifferen t species compositions ofth e phytoplankton in these bags may have been the cause of this effect.

particulott motter mm'r'

FIG. 5. Concentration of particulate matter in different bags (average 0.5 and 2.0 m in depth).

A3-7 EFFECTS OF MERCURY ON MARINE PLANKTON 113

March- Moy 1976

bog day total volume

o control 11 106 >3 89 +• 05/jgHgl-' it. 127 » 5 22 99 * 5 23 65 o 50 12 7.1

760 30X am US29 votum*ptr porticl* t/j*l FIG.6 . Sizespectru mo fparticulat ematte ro nday so fmaximu mvolum ei ndifferen t bagsa ta dept ho f 2.0 m.

At the end of the experiment the concentration of phaeopigments suddenly increased in bag H50 to 13.7 mg • m~3, which is extremely high. The decrease of the chlorophyll concentration on the preceding days was at least 11 mgm"3. Further degradation of the phaeopigments seems to have been retarded. Figure 5 shows the concentration of particulate matter as measured with a Coulter counter. The pattern is similar to that shown by chlorophyll concentrations,

TABLE 1

MAIN SPECIES GENERATING CHLOROPHYLL MAXIMA IN THE DIFFERENT BAGS DURING THE EXPERIMENT

Cell volume range OugHg-1-') Phytoplankton species Day (/um3)

Control Chaetoceros diadema 12 150-500 Chaetoceros spp. Microflagellates 0.5 Chaetoceros diadema 12 150-500 Chaetoceros spp. Microflagellates Thallassiosira nordenskiöldii 20-22 200-1100 Large flagellates (Rhodomonas sp.) Microflagellates 50 Thallassiosira nordenskiöldii 41 2200-4500 Skeletonema costatum

A3-8 114 JAN KUIPER although maxima were often reached 1 day later, and the range between minima and maxima was much narrower. Addition of 5an d 50/i g Hg-liter-1 had a very marked effect; after addition of 0.5 fig Hg-liter-1, concentrations of particulate matter were not significantly lower than those in the controls. Interestingly, the concentration of particulate matter in bag H50 did not fall to less than about 0.5 mm3liter~'. The particles were all smaller than 10|ii m in diameter; they were probably nonliving and kept in suspension by turbulence. Figure 6 shows the distribution of particulate matter over the different size categories at times coinciding with phytoplankton maxima in the different bags. Addition of mercury influenced the.selectio n of species forming the major part of the subsequent phytoplankton blooms: at higher mercury concentrations, larger cellswer e selected. Table 1 liststh e speciestha tgenerate d thegrowt hpeak si nth e different bags, together with their cell volumes. These results are remarkable, because inearlie r experiments (Kuiper, 1977a,b)th e succession intim eo f species invariably proceeded from largeones , mostly diatoms,t o small microflagellates, a trend that hasals obee nfoun d byothe rinvestigator s (Takahashiet al. , 1975; Grice and Menzel, 1978;Thoma s and Seibert, 1977).

Primary Production The phytoplankton production between 9 AM and 3 PM measured with the 14C method inbag s at depths of0. 5an d 2.0 mi sshow n inFig .7 . Analysis of variance

production ,mgC i Morch-Mçy 1976 control 1 >V' O.SfjgHg t bag ii Ii / / V. 50 '.. ,. ... ".. I

'M v J

.-,.i-,.^.....»-... =3iii±!sBlSli. J'' 30 *-ttmt.doy%

A /A #\\J/V Jr Alj l\

FIG. 7. Carbon assimilation in different bags at0. S and 2.0 mo f depthfrom 9 t o 15 hr.

A3-9 EFFECTS OF MERCURY ON MARINE PLANKTON 115

TABLE 2

RELATIVE CARBON ASSIMILATION JUST BEFORE AND JUST AFTER ADDITION OF MERCURY

Depth Bag Day (WSHg-1-1) 0.5 m 2.0 m

4 All 5.0 (0.5)s 3.2 (0.5) 6 All 5.0 (0.5) 2.9 (0.5) 7 All 7.7 (0.5) 3.0 (0.5) 8 Controls 6.7 (0.8) 2.6 (0.8) 0.5 7.8(1.1) 2.5(1.1) 5 4.2 (0.8)* 1.3(0.8) 50 -0.1(1.1)* 0.1 (1.1)* 9 Controls 5.1(0.8) 3.1(1.1) 0.5 5.1(1.1) 2.6(1.1) 5 6.6 (0.8) 1.8(1.1) 50 -0.3(1.1)* 1.2(1.1)

"Values are expressed as mg C(mg chl)"'(6 hr)~', with the SD inparentheses . * Significantly different from controls (P < 0.1). givesa nestimat eo fth eS Do fth emeasurement so f 14%.Apar tfro m theadditio no f the mercury, resulting ina differen t pattern intim ei nbag sH5a ,H5b ,an d H50, the depth and the phytoplankton concentration have great influence on the development of primary production with time. Figure 8 shows the carbon assimilation from 9 AM to 3 PM per milligram of chlorophyll (relative carbon assimilation) at a depth of 0.5 m. Addition of 5ju g Hg-liter-1inhibit s this relative carbon assimilation on onlyon e day(Da y 8). After

t rtlolivt corbon assimilation tmq C. tmachlar'. IGhrl4

iO 30 ^ ttmitdoys) FIG. 8.Carbo nassimilatio n(fro m9 t o IS hr)pe rmilligra mo fchlorophyl la tdepth so f0. 5an d2. 0 mi n different bags.

A3-10 116 JAN KUIPER

motivecarbon ossirmlotion (mgC.lmg chlor' (Shrs)''}

March-May T976 2 Controls • OS/jgHgf' i 5

0 300 COO 600 «K 1000 1200 U.OD *• energyinput Ucm^doy''! FIG. 9. Relative carbon assimilation as afunctio n of light intensity (energy input). this 1-day inhibition the relative carbon assimilation is higher in 5/ug -liter -1 bags than in the controls and bag HI (see also Table 2). Therelativ ecarbo n assimilation inba gH5 0i sno t showni nFig .8 .Afte r addition ofth emercur yi twa sinhibite dfo ra tleas t6 days .Thereafte r thescatte ro fthi srati o wasver yhig h(rang e0-482) ,probabl ydu et oerror si nth emeasurement sa tth ever y low levels of chlorophyll and production during this period. Contrary to phytoplankton biomass (chlorophyll, particulate matter), which is evenly distributed overth e bags,relativ ecarbo n assimilation ata dept ho f2. 0 mi s nearly always lowertha n that at 0.5 m,probabl y owingt oth elowe r light intensity at 2.0 m. Figure 9show sth e relative carbon assimilation duringperiod so fgrowt h against available lightenergy . Theavailabl eligh tI d atdept hd wascompute d from kd \d = 0.95-lo-e~ (Parsons et al., 1977), wherel 0 = global radiation per day, d - depth (in m),k = verticalextinctio n coefficient. Thefacto r 0.95i s incorporated tocorrec t for 5%absorptio n ofth eligh tb yth e Plexiglascover so fth ebags .I nthi s formulae isestimate d tob eequa lt o 1.7/5 (Cadéean d Hegeman, 1974),wher e5 is the Secchidis cvisibilit y inmeters . Belowligh tenerg y inputso fabou t 500J • cm -2• day-1, the available light energy can inhibit the relative carbon assimilation rate.

Nutrients Figures 10A-E showth edevelopmen t ofth ephytoplankto n nutrients. Asusua l thefirst phytoplankto n bloom depletes thenutrien t poolunti lth econcentratio n of some nutrient (not necessarily one of those measured) falls to a value that limits growth.I ti sclea rtha tth ephytoplankto nfirst utilize sth eavailabl eammoni a before broaching the supply of nitrate. The most prominent result of adding mercury is the time lag in the growth of phytoplankton in H5a, H5b, and H50. In addition, the development of nitrogen compounds in these bags proceeds along different lines. Figure 10B shows that nitrate concentrations in the bags to which 5/u. gHg-liter -1 was added decreased further than those in the controls; Fig. 10C shows that the nitrite concentrations

A3-11 EFFECTS OF MERCURY ON MARINE PLANKTON 117

- ., March-May 1876 « . r control

OifigHgl' bog

5 50 •\ \ \IX

i*_^«i« V i \'^< i . _ i .

\ V.

v'v." W-.

S', wjoi

/ I w /\ — y ,.

Wj-Af. f/grol f' \ 10 « » 40

JO 40 FIG. 10.Concentratio n of selected phytoplanktonnutrient s inth edifferen t bags(averag e of0. 5 and 2.0 m indepth) . remainhig hi nH5a ,H5b ,an d H50;an dfinally , Fig. 10Ashow stha ti nH5 aan dH5 b the increase in ammonia concentrations is delayed longer than can be expected from the time lag in phytoplankton growth. Thequestio n whetheron eo fth emeasure d nutrients waslimitin gphytoplankto n growth was investigated inbioassa y experiments on three occasions. On Day 11,

A3-12 118 JAN KUIPER

nuiDAtr p*r tittr number p*r hlwr !

"> ,' """" " -r -• /V V OSf^gHg I bag 5 .VV-W^' —- » KK4' /s---. np// noupfii / \ / \ ./•

•'T.'' . S.v oi Y i i J / A j\/ I \ —-l»"..<*>/s

10 X> X u>

FIG. 11. Development of Temoralongkornis. Numbers of small nauplii (a), large nauplii (b), copepodites (c), and adults (d) per liter in the different bags. samples from the controls were incubated. Coulter counter counts gave no significant differences between the controls and the flasks to which nutrients had been added. On Day 15 (after the chlorophyll maximum) a second test was done with a sample from one control bag. Again there were no significant differences, indicating that none of the added nutrients was limiting phytoplankton growth during the first peak in the controls and 0.5 /ig liter-1 bag. Considering the low concentrations of phosphate and silicate, this finding is remarkable. On Day 23, samples from H5a were incubated. Coulter counter results and microscopic inspection showed that in this bag phosphorus and silicon limited the growth of Th. nordenstkiôldii; there is probably a different limiting factor for the flagellates.

3.2. DEVELOPMENT OF THE ZOOPLANKTON General The water that was used to fill the bags contained a variety of zooplankton species, with calanoid copepods predominating. The main species was Temora longicornis (60-100%o fth etota l number ofcopepods) , followed byAcartia clausi, Pseudocalanuselongatus,Centropageshamatus,andEurytemorahirundoides. In

A3-13 EFFECTS OF MERCURY ON MARINE PLANKTON 119 addition to copepods, larvae of bivalves, worms, and nauplii and cyprids of barnacles wereals ofound . Byth een do fth eexperiment , thelatte r hadgrow n into adults living at the bottoms of the bags. Figures 11-13sho wth edevelopmen t ofth emai nspecie sdurin gth eexperiment . T. longicornus was always the main species. During the first 10 days there was a decrease in numbers of small nauplii, which developed into large nauplii and copepodites.- After this period new nauplii developed from eggs produced in the

Fio. 12.Development so fAcartia clausi.Number so fnaupli i(a) ,copepodite s(b) ,an dadult spe rlite r in the different bags.

A3-14 120 JAN KUIPER bags, sotw ogeneration s occurred together. The same seemst ohav ebee ntru e for A. clausi and P. elongatus. There was no significant difference between the development ofT. longicornis inth econtro l bagsan d that inba gHI . Addition of5 Mg Hg• liter -1had ,however , amarke deffect . OnDa y 11,4day safte r dosing,ther e werefewe r large nauplii, copepodites, and adults inbag s H5aan d H5btha n inth e controls, indicating a lower development rateo r ahighe r mortality rate orbot hi n the preceding period. After Day 11, the development of small nauplii was comparablet otha ti nth econtrol sunti lDa y25 ,whe na ne wminimu mi nth enumbe r

° » ' ' * »—r» » «I 1

FIG. 13. Development ofPseudocalanus elongatus.Number s of nauplii (a), copepodites (b),an d adults perlite r inth e different bags.

A3-15 EFFECTSO FMERCUR YO NMARIN EPLANKTO N 121 of small nauplii was reached. This may be an indirect effect of mercury addition, through mercury accumulation inth e food chain. The differences indevelopmen t of T. longicornis in H5a and H5b are interesting. Effects of mercury on the develop­ ment were most pronounced in the bag in which the phytoplankton bloom after the mercury addition consisted ofThalassiosira and flagellates. Summarizing, addition of 5 /ig Hg-liter-1 retarded the development of T. longicornis by about 1 week. Addition,o f 5 jug-liter -1 showed no clear-cut effect on the development of A. clausi, probably owing to the large scatter in measured numbers. Figure 13 shows that addition of 5 /u,g Hg-liter-1 decreased the numbers of Pseudocalanus adults. Addition of a single dose of 50/x gH g• liter -1 killed most of the copepods almost at once; however, some survived, for their numbers rose again at the end of the experiment.

Sex Ratios The sexratio s ofT. longicornis, A. clausi, andP . elongatus, summed throughout their development, are shown in Table 3. Addition of 5an d 50/tt gHg -liter -1 clearly influenced the sex ratio in these species.

Development and Mortality Rates and Secondary Production Relative rates of development, mortality, and reproduction were estimated by multiple regression analysis using the model of Fransz (1975). Results for T. longicornis are given in Table 4. It is not clear whether the observed retardation effects after addition of mercury are due to an increased mortality rate or a decreased development rate. Secondary production was estimated using the development rates from Table 4 and dry weight-length relationships given by Robertson (1968) and Nassogne (1972). Figure 14 shows the production of T. longicornis, which is responsible for most of the secondary production during the experiment.

3.3. DEVELOPMENT OF DECOMPOSERS Figure 15 shows the bacterial concentration in the water in the different bags. These numbers are similar to those found in Dutch coastal waters. The numbers of

TABLE 3 SEX RATIOS OFTemora longicornis,Acartia clausi, ANDPseudocalanus elongatus EXPOSED TO MERCURY, SUMMED THROUGHOUT THEIR DEVELOPMENT

Initial mercury Sex ratio [9/(9 +

(MgHg-l-') T. longicorr is A. clausi P. elongatus

0 0.41 0.98 0.38 0 0.51 0.88 0.66 0.5 0.56 0.93 0.18 5 0.75 0.47 0.99 5 0.68 0.50 0.80 50 0.72 0.50 —

A3-16 122 JAN KUIPER

TABLE4

RELATIVE MORTALITY AND DEVELOPMENT RATÉS OF Temora longicornis (MARCH-MAY 1976)

Relative development rate Bag (MgHg'l-) "i n2 c, c2 Control 0.11(0.09) 0.11(0.09) 0.02 (0.04) 0.09 (0.06) 5 0.12(0.08) 0.13 (0.09) 0.02(0.05) 0.17(0.16) 0.5 0.15(0.16) 0.02 (0.01) 0.04 (0.03) 0.05 (0.08) 5 0.20(0.16) 0.01 (0.04) 0.09(0.11) 0.17(0.21) Control 0.18(0.18) 0.10(0.06) 0.02 (0.01) 0.03 (0.04) 50 0.41(0.13) 0.37 (0.48) 0.07 (0.22) 0.05 (0.08)

Relative mortality rate

n, n2 c, c2 aï ao" Control -0.09 (0.23) -0.01 (0.14) 0.04(0.11) -0.21 (0.13) 0.07(0.13) 0.14(0.26) 5 -0.01 (0.19) 0.01 (0.16) 0.17(0.19) -0.13 (0.35) 0.17(0.25) 0.06(0.16) 0.5 -0.06 (0.39) 0.10(0.16) -0.08 (0.05) 0.05 (0.19) 0.03 (0.09) 0.01 (0.13) 5 -0.21 (0.34) 0.21(0.19) -0.17(0.12) 0.08 (0.46) 0.10(0.25) 0.08 (0.22) Control -0.05 (0.39) 0.07 (0.20) 0.04 (0.08) 0.02(0.11) 0.02 (0.07) 0.03(0.12) 50 -0.34 (0.15) 0.46 (0.57) 1.18(1.74) 0.20 (0.95) 0.12(0.30) 0.17(0.42) Note. TheS Di sgive ni nparentheses ,n, = smallnauplii , n* largenauplii ,c , = smallcopepodites , c2 = large copepodites, a =adults . bacteria inbag s HSa, HSb,an d H50o n Day 10 (just after addition ofth e mercury) are significantly higher than those inth e controls and bag HI (the SD ofth e measurements isestimate d to be 100%). The number of aerobic bacteria in the sediment traps was measured on 5day s during the experiment. Numbers ranged from 0.5 x 10et o 109 bacteria per trap. No significant differences were found between the bags.

- Secondary production. mgathfno dry wight m'1day''

March-Mov I97C o« controls • OSM Ho r' «« 5 .. •• • 50 ..

FIG. 14. Production of Temora longicornis in the different bags.

A3-17 EFFECTS OF MERCURY ON MARINE PLANKTON 123

to1

10'

s March-Mov 1976 \ s' control V —- .. O.SfjgHg I'1 bag

5 ,. SO .. time, days

0 tO 10 30 tO FIG. IS. Development of the relative number of bacteria in the water in different bags.

3.4. FATE OF THE ADDED MERCURY The concentration of the added mercury in the water in the different bags as a function of time is shown in Fig. 2. Analysis of variance reveals that there is no significant difference between concentrations at depths of 0.5 and 2.0 m. The average SDo fth e measurements isestimate d tob e 18%,whic hi sprobabl y too high for the higher concentrations. Until about Day 30,th edecreas e inth econcentration s inH I and H5a/H5b obeys 028( _7> the equation C, = C0e" ' , where C, = concentration on day /, andC 0i s the -1 concentration in the water on Day 7. C0 being 0.43 and 5.2 ju.g Hg-liter , respectively, the concentration decreases by 32% a day. The concentration of mercury inba g H50decrease s more slowly and less evenly. Toward the end of the experiment, the concentrations in bags HI and H5a/H5b increase again to 0.3 and 0.7 //.g-liter-1, respectively. Figure 16 shows (on a dry weight basis) the mercury concentration throughout theexperimen t inth esediment , inth esedimen t traps,an di nth ebotto msedimen ta t the end of the experiment. Concentrations strongly increase after addition of mercury toth e water, andreac h amaximu m inth e middleo fth eexperiment . Table 5 lists the methylmercury concentrations in the sediment at the end of the

1000 j March-Moy 1976 Hg . "1 9*9 t • „. SHgHgt-'

too • .>" 0S^9_HSJ_1___. to

control m

"* control ^ tmo. days t ! 0 10 20 30 CO FIG. 16. Mercury concentrations in the sediment in the different bags during the experiment.

A3-18 124 JAN KUIPER

TABLE 5

METHYLMERCURY (MeHg) IN THE SEDIMENT AT THE END OF THE EXPERIMENT

Concentration of Total MeHg MeHg on adr y Total Hg Bag in sediment weight basis added to bags (MgHg-1-') (Mg) Oxg-kg-') (mg)

Control 0.06 19.2 0 Control 0.27 37.9 0 0.5 0.63 83.4 0.7 5 5.04 478.6 7 5 0.54 121.8 7 50 0.23 58.5 70 experiment. During the experiment no methylmercury could be detected inth e sediment traps, indicating a lagphas e of about a month before methylation starts. This was also found byD eKoe ket al. (MT-TNO, unpublished results) workingi n the Dutch Wadden Sea.

3.5. ALGAL GROWTH ON THE WALLS

Results of the investigations into the algal fouling of the walls of the bags are described in detail by Grolle and Kuiper (1980). Figure 17show s the amount of chlorophyll at a depth of 0.4 m in the different bags. Fouling of the walls ofth e bags byalga e isunimportan t during the first weeks ofth e experiment. Atth e end, algal biomass on the walls cannot be neglected (e.g., as a sink for nutrients). Addition of50/x g Hg-liter-1 strongly inhibited development ofth e attached algae. Addition of 5pg Hg-liter" 1 inhibited thegrowt h rate ofth e periphyton only atth e beginning of the experiment (Days 10an d13) .

March-May 1976 % control

30 60 lime fdoyst

FIG. 17. Chlorophyll on glass slides at a depth of 0.4 ma s a function of time in different bags.

A3-19 EFFECTS OFMERCUR Y ON MARINE PLANKTON 125

The species composition ofth e periphyton was also affected. Addition of 0.5/x g Hg• liter -1di dno tinfluenc e development of the biomass ofth e attached algae,bu ti t did alter species composition. Melosira nummuloides is absent from all mercury bags, while iti spresen t inal lsample s from thecontrol s after Day2 0a tdensitie so f 0.03-1.0/mm2.

4. DISCUSSION The results are compared with those oftw o similar experiments with much larger bags: Davies and Gamble (1979) using bags containing 95m 3 of seawater and, a Controlled Ecosystem Pollution Experiment (CEPEX) which was performed in 1300-m3 enclosures (results published by various authors inMar. Sei. Commun. 3, (1977)).

4. 1. DEVELOPMENT IN DUPLICATE BAGS Development of the plankton communities inth econtrol s proceeds according to very similar patterns. Although there aredifference s in species composition of the phytoplankton and in the response of the zooplankton to a concentration of 5/n g Hg-liter-1, the development of the plankton in H5aan d H5bals o obeys a similar pattern. This confirms earlier results (Kuiper, 1977a,b) and allows the differences between the bags to be attributed to addition of mercuric chloride.

4.2. EFFECTS OF MERCURY ON THEPHYTOPLANKTO N Addition of 50pg Hg-liter-1 resulted in complete inhibition of phytoplankton activity. This inhibition caused a decrease of phytoplankton biomass, due toth e cellssettlin gt oth e bottom ofth e bags. Growth resumed about 20day s after addition of the mercuric chloride, although mercury concentrations were still as high as1 8 /Ltg Hg-liter-1. This recovery may have been due to development of mercury- resistant species, or to inactivation of the mercury by chelation or adsorption to inanimate particles. Thomas etal. (1977) obtained similar results andfoun d indica­ tions for both hypotheses. Addition of 5 /ng Hg-liter-1 also reduced the phytoplankton growth rate. The phytoplankton biomass decreased initially, but began to increase again as soon as the mercury concentration had fallen to about 1.5 /u,g Hg-liter-1, which is comparable to results of an earlier experiment (Kuiper, 1977b). Interestingly, addition of 5 /u.g Hg-liter-1 delayed the appearance of the phytoplankton peak by about 9days , whereas itinhibite d therelativ e carbon assimilation (Table 2)b yonl y 1day .Thoma s et al. (1977) likewise observed that the ability of phytoplankton to photosynthesize was affected less than its ability to increase its biomass. Two explanations, notmutuall y exclusive, arepossible : (1)mercur y does notinhibi tth e uptake ofcarbo n dioxide,bu tdoe s inhibit celldivision ;(2 )carbo n assimilation rates (measured from 9A Mt o3 PM inbottles ) inth econtrol s and bags H5aan d H5b cannot be compared because of differences in species composition. Testing the first hypothesis in the laboratory showed that cell division is more sensitive tomercuri c chloridetha n iscarbo n assimilation (Hanstveit and Oldersma, unpublished results). Other investigators also showed that cell division isth e most sensitive parameter for measuring effects of mercury on phytoplankton (Berland

A3-20 126 JAN KUIPER

80 % inhibition

60

= 0 63* - ' S9lr'>0t3. n.tl

20

—^-mio' pgHg per cell 0 10 20 30 tO SO 60 70 80 90 100

FIG. 18. Inhibitionof Chlamydomonas sp. inrelatio nt oth equantit yo fmercur y absorbed byth ealgae .

et al., 1977;Bryan , 1976).Davie s (1974)an d Premazziet al. (1978)foun d that the cell sizeo funialga l cultures increased under mercury stress asa result ofth e fact that cell division is more inhibited than carbon assimilation. Addition of5 and50/i gHg-liter -1altere d the speciescompositio n ofth egrowt h peak following the addition, higher mercury concentrations favoring selection of larger species. Since mercury affects the phytoplankton bybein gadsorbe d on the cellwalls ,th esmalle rsurfac e tovolum erati oo flarge rcell sma yexplai nwh ylarge r cells are more resistant to higher mercury concentrations. Ifthi s is true it should also be true that the toxicity of mercury to algae, apart from depending on the mercury concentration, is related to the total surface area available for mercury adsorption. Figure 18show s the results of experiments with unialgal cultures of Chlamydomonas sp. on the toxicity of HgCl2 in concentrations of 50an d 100 jug Hg-liter-1, the initial cell concentration being varied (Meijer and Oldersma, unpublished resultsobtaine d atthi slaboratory) .Thes eresult ssho wth eimportanc e of particle concentrations, that is, surface area, for the toxicity of mercury. A dependenceo fth etoxicit yo fmercur yo nparticl econcentratio nha sals obee nfoun d by others (Hamdy and Wheeler, 1978; Rice et al., 1973; Kamp-Nielsen, 1971; Harriss et al., 1970;Davies , 1978). The shift to larger cells in mercury-polluted communities can, however, alsob e explained in a different way. Addition of mercury also decreases the numbers of zooplanktonorganisms ,an d thereby thegrazin gpressur e onth ephytoplankto n in the bags to which 5an d 50/u. gHg'liter -1 were added. Larger cells are probably preferred by large copepodites and adult copepods over small microflagellates (Gamble et al., 1977).Th e difference in zooplankton densities between bags H5a and H5bcoincide s with differences in the phytoplankton species composition. In theba gwit hles szooplankton ,Th. nordenskiöldii i sfound ; inth eothe rba gdiatom s could not be detected in large numbers, although silicate concentrations were reduced in a way similar to that in the other, indicating that diatoms were also

A3-21 EFFECTS OF MERCURY ON MARINE PLANKTON 127 growing in this bag. Thomas et al. (1977) also found that reduced grazing ledt o growth of phytoplankton consisting of larger cells. •From the results it is difficult to decide whether phytoplankton species composition governs the zooplankton growth rate (Sonntag andGrève , 1977),o r whetherzooplankto ngrazin glead st oa specifi c phytoplankton community. Results from bagH5 0 point toth ezooplankto n and indicate that iti s notth eenclosur e as such, withit sreduce d turbulence, which promotes apredominanc e ofsmall , motile cells inth e controls (cf. Eppley et al., 1978).

4.3. EFFECTS OF MERCURY ON THE ZOOPLANKTON Addition of 5ti g Hg-liter-1 causes adela y inth edevelopmen t ofT. longicornis and P. elongatus. Similar effects were demonstrated in other experiments with large bags (Davies and Gamble, 1979; Beerset al., 1977;Gric e and Menzel, 1978). Inthes e experiments addition of 1 /xg Hg-liter -1 didno thav e anyinfluenc e on the zooplankton either, although addition of 5 or 10/x g Hg-liter-1 led to lower zoo- plankton densities as compared with a control. FromTabl e4 i ti sno tclea rwhethe r thedifferen t retardation effects aredu et oa n increased mortality rateo rt oa decrease d development rate. Thisobscurit y mayb e due to the fact that in the regression analysis we used all points, and assumed constant mortality and development rates for the duration of the experiment. These rates are not, however, likely to be constant, a single dose of mercury probably affecting theplankto n fora limite d duration oftim e only. Therefore, we also computed the development and mortality rates of T. longicornis, for this purpose dividing the experiment into twoperiod s (Day0-15 , and Days 17-44). The large standarddeviation s ofth e resultant ratesdi dno tallo w any conclusions to be drawn. However, the numbers of nauplii and copepodites in the 5 tig Hg-liter-1 bags being similar (after the lagphase ) tothos e inth econtrols , itwoul d seem probable that the development rate is decreased. The estimates inTabl e 4ar eo fth e same order of magnitude as those of Fransz (1975)foun d in theNort h Sea. This indicates that thecopepod s inth e bags develop in anatura l way. The effects ofmercur y addition onth e sexratio s of the copepods are very pronounced. Inmarin e copepods sexdeterminatio n is mainly phenotypic (Takeda, 1950). Fransz(1975 )suggeste d thatth e unexpected sex ratios hefoun di n the North Sea were induced by what he called "environmental factors."

4.4. EFFECTS OF MERCURY ON THE DECOMPOSERS OnDa y 10,th e numbero fviabl e bacteria was higherafte r additiono f5 an d5 0/x g Hg-liter-1 than in the controls. This growth was probably caused by the high mortality of the phytoplankton. Azam et al. (1977), also working with a natural marine plankton community, found a strong decrease just after addition of 5 xig Hg-liter-1, followed by the development of a mercury-resistant bacterial population, which reached highernumber s thani n the control. Davies and Gamble (1979) found noeffect s of mercury onth e bacteria populations they used. Measurements of the nitrogenous phytoplankton nutrients showed interesting differences between the controlsan dth ebag st owhic h5 an d5 0fig Hg• liter -1 were added. Nitrate concentrations after addition of 5 /xg Hg-liter-1 decreased further than inth e controls, nitrite concentrations remained high in bags H5a,H5b , and

A3-22 128 JAN KUIPER

H50,an dth eincreas ei nammoni aconcentration s inbag sH5 aan dH5 bwa sdelaye d longer than the time lag in the growth of the phytoplankton would lead one to expect. Ammonia reappears in the water after degradation of organic matter by bacteria orzooplankton .A dela yi nth ereappearanc e ofammoni aindicate stha tth e rateo fconversio no forgani cmatte rint oammoni ai sreduced . Harrisonet al. (1978) found that in short-term laboratory experiments ammonia regeneration was inhibited by mercury concentrations greater than 1 ju.gHg -liter -1. The decreased rate of ammonia regeneration forces the phytoplankton to use other nitrogencompounds .Thi s mayb eth ecaus eo fth efurthe r decrease ofnitrat e concentrations in the bagst o which 5 ^.gHg-liter -1 was added, ascompare d with the controls. Koike et al. (1978) found that after addition of 5/x g Hg-liter-1 to a natural community, selection occurred for denitrifying bacteria. This may explain our high nitrite concentrations after addition of 5an d 50/u g Hg-liter-1. These indications for effects ofadditio n of5 and5 0/xg Hg• liter -1 on stepsi nth e mineralization process support a hypothesis put forward as a result of an earlier experiment (Kuiper, 1977b).Inhibitio n ofammoni aregeneratio n didno tresul t ina delay ofphytoplankto n growthi nth epresen t work, asi tprobabl y did inth eearlie r experiment, because nitrate concentrations in the water were higher than previously. Several reviews of the toxicity of mercury to aquatic life are available (Bryan, 1976; Taylor, 1977;Stebbing , 1976; Davies, 1978; Lelande?al., 1979).Laborator y experiments seldom show effects of mercury in concentrations lower than 1 /u,g Hg-liter-1. Moore and Stebbing (1976)foun d athreshol d concentration of0.1 7 ng Hg• liter -1fo rbiochemica leffect s ona marin ehydroid . Sigmonet al. (1977 )foun da reduction in the diversity of the epiphyton community in an artificial stream, and Sawardet al. (1974 ,cite d by Daviesan d Gamble, 1979) found minoreffect s onth e phytoplankton atconcentration s of0. 1/xg Hg-liter-1. Inth e last-mentioned study, theexperimenter skep tth emercur yconcentratio n atth edesire dleve lb y refreshing the water. The organisms were exposed to new doses of "ionic or reactive" mercury (Davies and Gamble, 1979). Results of this and other bag experiments suggest that the toxicity of mercury varies with the chemical form in which it is present.

4.5. PHYTOPLANKTON-ZOOPLANKTON INTERACTIONS Asshow n in Fig. 14 the production ofT. longicornis isth e major component of zooplankton biomass during the experiment. The total daily primary production canb eestimate dfro mth ecarbo nassimilatio n measurementsa ttwic eth emeasure d production at 0.5 m (unpublished results, this laboratory), or from the pH measurements(Kuiper , 1978).Secondar y production onlyform s animportan t part ofprimar yproductio n inth esecon dhal fo fth eexperiment . Fromthi sobservatio ni t follows that, during the first 20 days of the experiment, grazing was probably unimportant (the possible influence on the phytoplankton species succession by selective grazing of larger particles has already been discussed). After Day 20, however, itbecam eincreasingl yimportan t andi tseem sprobabl etha t atth een do f the experiment the phytoplankton wasgraze d down by the copepods. Toestimat e the influence ofth ezooplankto no nth e phytoplankton directly, we conducteda grazin gexperimen to nDa y44 .A sampl eo f3. 2liter sfro mba gH5 bwa s splitint osi xportion so f0. 5liter .Thes esubsample swer etransferre d toseru m flasks

A3-23 EFFECTS OF MERCURY ON MARINE PLANKTON 129 after three of them had been filtered through a 55-/n.m net to remove the zooplankton.Th eflask s wereincubate dfro m 11 AMt o3 PMi na coole dcontaine ro n the raft. After the incubation, the phytoplankton was counted with a Coulter counter. In the flasks with zooplankton, 6813 (SD = 103) particles• ml -1 were counted, and inth eflasks withou t zooplankton 8818(S D = 207)particle s ml-1. It follows that 23%o fth e particles had beenfiltered b y the zooplankton during4 hr , indicating'the great importance of grazing in this phase of the experiment.

4.6. FATE OF THE ADDED MERCURY

The decrease of mercury concentrations in the water and the accumulation of mercury in the sediment show that part of the mercury is adsorbed by particles whichsettl et oth e bottomo fth eba g(cf . Smiths a/., 1971;Takahashie/o/., 1977; Kuiper, 1977b).However , atth een d ofth eexperimen t only 21-27%o fth eadde d mercury wasrecovere d inth esedimen t andth ewater . Asi nth eearlie rexperimen t (Kuiper 1977b),a largepar t ofth eadde d mercury disappeared or waspresen t ina compartment which was not measured. Takahashi etal. (1977)foun d that nearly all the mercury lost from the water in their enclosures had passed into the sediment. Davies and Gamble (1979) found 25-30% of the added mercury in the sediment at the end of the experiment; the remainder was mostly found on the walls,an d small losses to the atmosphere and through the wallso fth ebag s were reported. Wedi dno t measure the mercury that was strongly bound to suspended particles. Estimating the concentration of these particles to be 1 mg-liter-1 (—0.5mm 3-liter-1, which remained suspended in bag H50)an dth emercur y concentrations at 20,200,an d 1000mg-kg -1(Fig . 16), 0.035, 0.35,an d 1.5m gmercur y(5,5 ,an d2 %o fth eamoun tadded )ca nb eaccounte dfo ra t the end of the experiment. This indicates that at the end of the experiment suspended particles are probably not important as a sink for the added mercury. Anotherpossibilit y istha tpar to fth emercur y inth ewate rwa sboun dt ochelate s (Bryan, 1976; Kayser, 1976; So, 1979), making it inaccessible to the analytical method used. This might also explain the reappearance of mercury inth e wateri n the bags to which 0.5 and 5/i g Hg-liter-1 were added. Kuiper (1977b) obtained indications that adsorption of mercury on the walls of the bag was unimportant. Toremov eth euncertaint yabou tth efat eo fmercur yi nthi ssemifiel d experiment, asmall-scal eba gexperimen twa sperforme d inth elaboratory . Aba gcontainin g10 0 liters of seawater was suspended ina larg e container containing another 200liter s water. The container wasclose d witha nairtigh t lid. Mercuric chloride (1mg ) was addedt oth eba gt ogiv ea ninitia lconcentratio n of 10/i gH g• liter -1an dth eamount s ofmercur y inth ewater ,th esuspende dparticle san dth esediment ,o nth ewall san d the nylon ropes,an d(nonquantitatively ) inth eai ri nth econtaine r were monitored for 4weeks .A tth een do fthi sexperimen t approximately 10%wa sadsorbe d byth e wallo fth ebag , 10% hadaccumulate d inth e sediment,an d25 %wa sstil lpresen ti n thewate ran d suspended particles.Th ever y highconcentration s ofmercur y inth e air in the container suggested that the remainder had disappeared to the atmosphere. No mercury could be detected in the water surrounding the bag. Of the 75% of added mercury which we could not account for in thefield experiment,probabl y lesstha n 10%ha dbee nabsorbe d toth ewall so fth ebags .Th e remainder probably had been lost to the atmosphere. This loss of mercury is

A3-24 130 JAN KUIPER probably not due to methylation and subsequent volatilization, because methylmercury was only found in very low concentrations at the end of the experiment. More probable is the transformation of the added mercury into the volatile metallic mercury; bacteria are ablet ocarr y out thisconversio n (Kushner, 1974;Schottele/a/., 1974; Nelsonan dColwell , 1975;Baierefa/. , 1975), whereasi t alsooccur s asa purel y chemical process(Newto n and Ellis, 1974;Ketchu met al., 1975; Glickstein 1979; Baier et al., 1975). In seawater containing low concentrations of mercury this conversion can be explained thermodynamically (Baier, et al., 1975). Davies and Gamble (1979) and Takahashi et al. (1977) found no or very small losses to the atmosphere. This isprobabl y caused by the much smaller water-air areai nrelatio nt oth eenclose d volumean dth ereduce d mixingi nver ylarg ebag sa s compared with the enclosures used in this study. It is not clear how much of the mercury woulddisappea runde rnatura lconditions ,sinc ethes econdition s willvar y in space and time. Brosset and Svedung (1977)showe d that the North Sea (rather than industrial air pollution) was the primary source of atmospheric mercury in Sweden, and Wedepohl (1970)als o reported that the sea isa source of mercury in the atmosphere.

5. CONCLUSIONS 1. Thedevelopmen t ofth eplankto n communities induplicat e bagsshowe d very similarpattern si ntime .Thi smean stha tlarge rdifference s between thebag sca nb e attributed to the addition of mercuric chloride. 2. The addition of a single dose of 0.5, 5, and 50 /xg Hgliter-1 as mercuric chloride to North Sea coastal plankton communities enclosed in plastic bags resulted in: a.a decreas eo fmercur yconcentration s inth ewate rphas eafte r additiono f0. 5 and 5 /xgH g- liter" 1a ta rat e ofon ethir d perda y(afte r addition of5 0/x g Hg• liter -1 this decrease was slower); b. a complete inactivation of phytoplankton activity after addition of 50 /xg Hg-liter-1, and inhibition of phytoplankton growth rates after addition of 5jx g Hg-liter-1 as longa s concentrations in the water were 1.5 /xgHg-liter -1 or higher (no significant influence on growth rates of the phytoplankton was found after addition of 0.5 /xg Hg-liter-1); c. adela y of9 day si nth eoccurrenc e ofth ephytoplankto n maximumafte r the additiono f5 /x gH g• liter -1,an da dela y of2 9day safte r additiono f5 0/x g Hg• liter -1 (the possible causes of this delay are discussed); d. indications that cell division is inhibited at lower mercury concentrations compared to cell growth (measured as C02 uptake); e. a change of the community structure of the phytoplankton bloom after addition of 5an d 50/x gHg-liter -1, higher mercury concentrations favoring larger species, probably as a result of reduced grazing pressure; f. an immediate mortality of most copepods after addition of 50/x gH g- liter -1 addition of 5/x gHg-liter -1 clearly affects the development ofT. longicornis,th e mainzooplankto n species inth ebags ,an d ofP . elongatus; addition of5 and5 0 /xg Hg-liter-1 changed the sex ratios ofT. longicornis, A. clausi, and P. elongatus;

A3-25 EFFECTS OF MERCURY ON MARINE PLANKTON 131

g. a change in thegrowt hpattern so fdecomposer si nth ewate rafte r additiono f 5an d5 0/i gH g• liter "' ,indication swer eobtaine dtha tth econversio nrat eo forgani c matter into ammonia is inhibited by addition of 5 /ugHg-liter" 1; h. methylation of the added mercury in the sediment of the bags, after ala g phase of about amonth ; about 75%o f the added mercury was lost 40day s after addition, probably for the most part by volatilization and subsequent loss to the atmosphere; i. adifferen t growth pattern of the periphyton after addition of 5 and 50/i g Hg-liter"1; the species composition of the periphyton is influenced inal l mercury-polluted bags. 3. The toxicity of mercury toth ephytoplankto n depends onth emercur y concentration, on the total surface area available for adsorption of mercury, and therefore onth erati obetwee n livingan dnonlivin gparticle s ina particula r system, and probably also on the metal species in which mercury is present. 4. The similaritybetwee n the resultso fthi sexperimen t andth eresult sobtaine d in experiments using much larger bags (95an d 1300 m3) is remarkable.

ACKNOWLEDGMENTS

Thanks are due to my colleagues Pirn de Koek, Gerard Hoornsman, Piet Roele, Ben Schrieken, Hannie van Noort, Janva nde n Eikhoff, and Henk van hetGroenewoud , whoassiste d in carryingou tth e various analyses and inevaluatin g the results. The assistance of research workers of the Department of Biology, Physics, and Analytical Chemistry of the Division of Technology for Society TNO is greatly appreciated. Iwoul d also liket othan k Drs. Fransz, Gieskes, and Van Arkelo fth e Netherlands Institute for Sea Research for their help and advice on the various aspects related toth e work reported here. Finally Iwoul d liket othan k Mr. Sjok Visseran d Dorus Pompert of B. V. Aquariusfo rthei renthusiasti c help, and the Royal Netherlands Navy for their kind cooperation in supplying experimental facilities.

REFERENCES ANDERSON, G. C, AND ZEUTSCHEL, R.P . (1970). Release of dissolved organic matter by marine phytoplankton in coastal and off-shore areas of the N.E. Pacific Ocean. Limnol. Oceanogr. 15, 402-407. AZAM, F., VACCARO, R. F., GILLESPIE, P. A., MOUSSALLI, E. I., ANDHODSON, R. E. (1977). Controlled Ecosystem Pollution Experiment: Effect of mercury on enclosed water columns. II. Marine bacterioplankton. Mar. Sei. Commun. 3, 313-330. BAIER, R. W.,WOJNOWICH , L., AND PÉTRIE, L. (1975). Mercury loss from culture media. Anal. Chem. 47, 2464-2467. BEERS, J.R. , REEVE, M. R., AND GRICE, G.D .(1977) . Controlled ecosystem pollution experiment: effect of mercury on enclosed water columns. IV Zooplankton population dynamics and production. Mar. Sei. Commun. 3,355-394 . BERLAND, B. R., BONIN, D. J., GUÉRIN-ANCEY, O. J., KAPKOV, V. I., AND ARLHAC, D. P. (1977). Action demétau x lourds àde s doses subléthales sur les caractéristiques del acroissanc e chezl a diatomée Skelelonema eostatum. Mar. Biol. 42,17-30 . BROSSET, C, AND SVEDUNG, L. (1977).Preliminary Study of the Possibility of a Relationship between High Acidity inLakes and High Mercury Content in Their Fish Populations. Swed. Water & Air Pollut. Lab. Rep. B378. BRYAN, G. W. (1976). Heavy metal contamination inth e sea. InMarine Pollution (R. Johnston, ed.), pp 185-302. Academic Press, London/New York/San Francisco. CADÉE. G. C, AND HEGEMAN, J. (1974). Primary production of the benthic microflora living on tidal flats in the Dutch Wadden Sea. Neth. J. Sea Res. 8,260-291 . DAVIES , A. G. (1974). The growth kinetics of Isochrysis galbana in cultures containing sublethal concentrations ofmercuri c chloride./. Mar. Biol. Assoc. U.K.54 , 157-169.

A3-26 132 JAN KUIPER

DAVIES, A. G. (1978). Pollution studies with marine plankton part. II. Heavy metals.Advan. Mar. Biol. 15, 381-508. DAVIES, J. M., AND GAMBLE, J. C. (1979). Experiments with large enclosed ecosystems. Phil. Trans. Roy. Soc. Ser. B 28«, 523-544. DREBES, G. (1974). Marine Phytoplankton. George Thieme Verlag, Stuttgart. EPPLEY, R. W., KOELLER, P., AND WALLACE, G. T. (1978). Stirring influences the phytoplankton species composition within enclosed columns of coastal seawater! J. Exp. Mar. Biol. Ecol. 32, 219-239. FRANSZ, H. G. (1975). The spring development of calanoid copepod populations in the Dutch coastal waters as related to primary production. In Proceedings, 10th European Symposium on Marine Biology (G. Persoons and E. Jaspers, ed.), Vol. 2, pp. 247-269. GAMBLE, J. C, DAVIES, J. M., AND STEELE, J. H. (1977). Loch Ewe Bag Experiment, 1974.Bull. Mar. Sei. 27, 146-175. GLICKSTEIN, N. (1979). The potential loss of dissolved mercury and selenium in marine experimenta­ tion. Mar. Pollut. Bull. 10, 157. GRAY, J. S. (1974). Synergistic effects of three heavy metals on growth rates of a marine ciliate protozoan. In Pollution and Physiology of Marine Organisms (F. J. Vernberg and W. B. Vernberg, eds.), pp. 465-485. Academic Press, New York. GRICE, G. C, AND MENZEL, D. W. (1978). Controlled Ecosystem Pollution Experiment: Effect of mercury on enclosed water columns. VIII. Summary of results. Mar. Sei. Commun. 4, 23-31. GROLLE, T., AND KUIPER, J. (1980). Development of marine periphyton under mercury stress in a controlled ecosystem experiment. Bull. Environ. Contam. Toxicol. 24, 858-865. HAMDY, M. K., AND WHEELER, S. R. (1978) .Inhibitio n ofbacteria l growth by mercury and the effects of protective agents. Bull. Environ. Contam. Toxicol. 20, 378-386. HARRISON, W. G., RENGER, E. H., AND EPPLEY, R. W. (1978). Controlled Ecosystem Pollution Experiment: Effect of mercury on enclosed water columns. VII. Inhibition of nitrogen assimilation and ammonia regeneration by plankton in seawater samples. Mar. Sei. Commun. 4, 13-22. HARRISS, R. C, WHITE, D. B., AND MACFARLANE, R. B. (1970). Mercury compounds reduce photosynthesis by plankton. Science 170: 736-737. HORSTMANN, U. (1972). Über den Einfluss von häuslichem Abwasser auf das Plankton in der Kieler Bucht. Kiel. Meeresforsch. 28, 178-198. HOUPT, P. M. (1976). Physical phenomena and analytical applications of helium microwave discharges. Anal. Chim. Acta 86, 129-138. INGRAM HENDEY, N. (1964).An Introductory Account of the Smaller Algae of British Coastal Waters, Part V, Bacillariophyceae (Diatoms). Her Majesty's Stationery Office, London. KAMP-NIELSEN, L. (1971). The effect of deleterious concentrations of mercury on the photosynthesis and growth of Chlorella pyrenoidosa. Physiol. Plant. 24, 556-561. KA YSER, H. (1976). Waste water assay with continuous algal cultures: The effect ofmercuri c acetate on the growth of some marine dinoflagellates. Mar. Biol. 36, 61-72. KETCHUM, B. H., ZITKO, V., AND SAWARD , D. (1975). Aspects of heavy metal and organohalogen pollution in aquatic ecosystems. In Ecological Toxicology Research, Effects of Heavy Metal and Organohalogen Compounds (A. D. Mclntyre and C. F. Mills, eds.), pp. 75-90. Plenum, New York/London. KOIKE, I., HATTORI, A., AND GOERING, J. J. (1978). Controlled Ecosystem Pollution Experiment: Effect of mercury on enclosed water columns. VI. Denitrification by marine bacteria. Mar. Sei. Commun. 4, 1-12. KUIPER, J. (1977a). Development of North Sea coastal plankton communities in separate plastic bags under identical conditions. Mar. Biol. 44, 97-107. KUIPER, J. (1977b). An experimental approach in studying the influence of mercury on a North Sea coastal plankton community. Helgolaender Wiss. Meeresunters. 30, 652-665. KUIPER, J. (1978).Continued Investigations into Pelagic Microscosms Subject to Environmental Stress by Pollutants. TNO-Report CL/78/100. KUSHNER, D. J. (1974). Microbial dealings with heavy metals. In Proceedings of the International Conference on Transport of Persistent Chemicals in Aquatic Ecosystems. (A. S. W. deFreitas, D. J. Kushner, and S. U. Qadri, eds.), Vol. 2, pp. 59-63. Nat. Res. Council of Canada, Ottawa. LACAZE, J. C. (1974). Ecotoxicology of crude oils and the use ofexperimenta l marine ecosystems. Mar. Pollut. Bull. 5, 153-156.

A3-27 EFFECTS OF MERCURY ON MARINE PLANKTON 133

LELAND, H.V. , LUOMA, S.N. , AND FIELDEN, J.M . (1979). Bioaccumulation and toxicity of heavy metals and related trace elements. J. Water Pollut. Contr. Fed. 51, 1592-1616. LORENZEN, C. (1967). Determination of chlorophyll and phaeopigments, spectrophotometric equa­ tions. Limnol. Oceanogr. 12, 343-346. MENZEL, D. W., AND CASE, J.(1977) . Concept and design: Controlled Ecosystem Pollution Experi­ ment. Bull. Mar. Sei. 27, 1-7. MOORE, M. N., AND STEBBING, A. R.D . (1976). The quantitative cytochemical effects of three metal ions on alysosoma l hydrolase ofa hydroid. J. Mar. Biol. Assoc. U.K.56 , 995-1005. NASSOONE, A. (1972).Etudes préliminaires sur le rôle du zooplankton dans la constitution etle transfer de lamatière organique ausein dela chaîne alimentaire marine enMer Ligure. NELSON , J. D., AND COLWELL , R. R. (1975). The ecology of mercury resistantbacteri a in Chesapeake Bay. Microbial Ecol. 1, 191-218. NEWTON , D. W., AND ELLIS, R. (1974). Losso fmercur y (II)fro m solution.J. Environ. Qual. 3,20-23. OPPENHEIMER, C. H., AND ZOBELL, C. E. (1952). The growth and viability ofsixty-thre e specieso f marine bacteria as influenced by hydrostatic pressure. J. Mar. Res. 11, 10-18. .PARSONS, T. R., TAKAHASHI, M., AND HARGRAVE, B. (1977).Biological Océanographie Processes, 2nd ed. Pergamon Press, Oxford. PREMAZZI, G., BERTONE, R., FREDDY, A., AND RAVERA, O. (1978). Combined effects of heavy metals and chelating substances onSelenastrum cultures. InProceedings, Research Seminar EEG und Umweltbundesamt, Berlin. UBA-Berichte 10/78, pp. 169-187. PUGH, P. R. (1973) . An evaluation of liquid scintillation counting techniques for use in aquatic primary production studies. Limnol. Oceanogr. 18, 310-319. REEVE, M. R., GRICE, G. D., GIBSON, V. R., WALTER, M. A., DARCY, K., AND IKEDA, T. (1976). A controlled environmental pollution experiment (CEPEX) and itsusefulnes s inth e study oflarge r marine zooplankton under toxic stress. In Effects of Pollution onAquatic Organisms (A.P . M. Lockwood, ed.), pp. 145-162. Cambridge Univ. Press, London/New York. RICE, H. V., LEIGHTY, D. A., AND MCLEOD, G. C. (1973). The effects of some trace metals on marine phytoplankton. CRC Crit. Rev. Microbiol., 27-49. RINGELBERG, J.(1973) . Parameter dependent (temperature) tolerance levels and the influence of the complexity ofth e biological system. Hydrobiol. Bull. 7, 106-114. ROBERTSON, A. (1968). The continuous plankton recorder: A method forstudyin g thebiomas s of calanoid copepods. Bull. Mar. Ecol. 6, 185-223. SAWARD , D., STIRLING, E. A., AND TOPPING, G. (1974).Experimental studies on the Effects of Copper and Mercury on a Marine Food Chain. Presented at Nato Science Committee Conference on ecotoxicity of heavy metals and organo halogen compounds Mont Gabriel, Canada. Cited by Davies and Gamble (1979). SCHOTTEL, J., MANDAL, A., TOTH, K., CLARK, D., AND SILVER, S. (1974). Mercury and mercury resistance determined by plasmids inEscherichia coli and Pseudomonas aeruginosa. In Proceedings of the Conference on Transport of persistent Chemicals in Aquatic Ecosystems (A. S. W. de Freitas, D. J.Kushner , and S.U .Qadri , eds.), Vol. 2,pp . 65-71. Nat. Res. Council ofCanada , Ottawa. SHELDON, R. W., AND PARSONS, T. R. (1967).A Practical Manual on the Use of the Coulter Counter in Marine Science. Coulter Electronic Sales, Toronto. SIGMON, C. F., KANIA, H. J., AND BEYERS, R. J. (1977). Reductions in biomass and diversity resulting from exposure tomercur y in artificial streams. J. Fish. Res.Canad. 34, 493-500. SMITH, J. D., NICHOLSON, R. A., AND MOORE, P. J. (1971). Mercury in the water of the tidal Thames. Nature (London) 232, 393-394. So, C.L . (1979). Mercury-water interactions inth e estuarine environment. Mar. Pollut. Bull.10 , 334-337. SONNTAG, N.C , AND GRÈVE, W. (1977). Investigation ofth e impact ofmercur y on enclosed water columns using azooplankto n simulation model. J. Fish. Res.Bd. Canad. 34, 2295-2307. SPIJK, S.J . (1975).Bepaling van sporen arseen, cadmium, chroom, koper, kwik, lood en zinkin zeewater d.m.v. atomaire absorptie-spectrometrie. TNO-Report CL 75/104. STEBBING, A. R. D.(1976) .Th eeffect s oflo w metallevel so na clona l hydroid.J. Mar. Biol. Assoc. U.K. 56, 977-994. STEEMAN-NIELSEN, E. (1952). The use of radioactive carbon (C14)fo r measuring organic production in the sea. J. Cons. Ferma. Int. Explor. Mer18 , 117-140. STRICKLAND, J.D . H., AND PARSONS, T. R.(1968) . Apractica l handbook ofseawate r analysis. Bull. Fish. Res.Bd. Canad. 167, 1-311.

A3-28 134 JAN KUIPER

STRICKLAND, J. D. H., AND TERHUNE, L. D. B. (1961).Th e study of in situ marine photosynthesis using a large plastic bag. Limnol. Oceanogr. 6, 93-96. TAKAHASHI, M., THOMAS, W. H., SEIBERT, D. L. R., BEERS, J., KOELLER, P., AND PARSONS, T. R. (1975). The replication of biological events in enclosed water columns. Arch. Hydrobiol. 76, 5-23. TAKAHASHI, M., WALLACE, G.T. , WHITNEY, F. A., AND MENZEL, D. W.(1977) . Controlled Ecosystem Pollution Experiment: Effect ofmercur y onenclose d watercolumns . I. Manipulation of experimental enclosures. Mar. Sei. Commun. 3,291-312 . TAKEDA, N. (1950). Experimental studies on the effect of external agencies on the sexuality of marine copepod. Physiol. Zool. 23, 288-301. TAYLOR, D. (1977).A Summary of the Data on the Toxicity of Various Materials to Aquatic Life, Vol.I . Mercury Rep., ICI Ltd., No. BUA/1784 . THOMAS, W. H., AND SEIBERT, D. L. R. (1977). Effects ofcoppe r onth e dominance and thediversit y of algae: Controlled Ecosystem Pollution Experiment. Bull. Mar. Sei. 27, 23-33. THOMAS, W.H. ,SEIBERT , D.L .R. , AND TAKAHASHI, M. (1977). Controlled Ecosystem Pollution Experiment: Effect of mercury on enclosed water columns. III. Phytoplankton population dynamics and production. Mar. Sei. Commun. 3,331-354 . TJOE, P. S., DE GOEU, J. J.M. , AND HOUTMAN, J. P.W . (1973). Automated chemical separationsi n routine activation analysis. J. Radioanal. Chem. 16, 153-164. UTERMÖHL, H.(1958) . Zur Vervollkommnung der quantitativen Phytoplankton Methodik. Mitt. Int. Ver. Limnol. 9, 1-38. VOLLENWEIDER , R.A . (1969).A Manual on Methods for Measuring Primary Production in Aquatic Environments. Blackwell, Oxford/Edinburgh. WEDEPOHL, K.H .(ed. ) (1970). Handbook of Geochemistry, Vol. II,2 ,p . 80-1-1. Springer-Verlag, Berlin.

A3-29 Publication P 81/40 1981-09-07

EFFECTS OFMERCUR Y ON ENCLOSED PLANKTON

COMMUNITIES INTH E ROSFJORD DURING POSER

+ Jan Kuiper Henk vanhe t Groenewoud + Gerrit Hoornsman + PietRoel e ++ Uwe Brockmann

+ Laboratory forApplie d Marine Research,Divisio n of Technology for Society TNO,P.O.Bo x 57,178 0A B Den Helder,Th eNetherlands . ++ Institut für organische Chemie u. Biochemie derUniversitä t Hamburg, Martin Luther King Platz 6, 2000 Hamburg 13,FR G

A4-1 ABSTRACT

During POSER two experimentswer e performed to study the effects of a single addition ofmercury(II)chlorid e on the development of plankton communities enclosed in large plastic bags, containing 1.5-30m 3 water and ranging from 3 to 40m in depth. The experiments lasted 1an d 3weeks ,respectively . In one of these experiments the influence of an additional dose of nutrients was studied.

During both experiments mercury concentrations in thewate r decreased after the addition. Part of themercur y was found in the sediment that collected on thebotto m of thebag s during the experiment,bu t most of themercur y probably volatilized and was lost to the atmosphere. _l Addition of 1o r 5 ugHg. l in the first experiment resulted in lower num­ bers of bacteria as compared with the numbers in the controls.Th e first experiment was too short to show effects on thephytoplankto n and the zoo- plankton. _l Addition of 5 ugHg. l in the second experiment inhibited the bacteria and the phytoplankton, and increased themortalit y rate of the copepods,whic h formed the principal part of the zooplankton.

The results of mercury addition to these marine plankton ecosystems were comparable with results obtained during other experiments with natural marine plankton communities also enclosed in large plastic bags in fairly different sea areas.

Addition of nutrients stimulated phytoplankton growth,bu t had neither effects on thebacteri a nor on the zooplankton.

A4-2 1. INTRODUCTION

' The extrapolation of the results of laboratory toxicity tests tomarin e eco- I I systems isdifficul t if not impossible. The reasons were discussed extensive- ) ly in the literature (Gray 1974,Menze l and Case 1977,Perkin s 1979). Experi­ mentswit h complex systems,whic h more closely approximate field conditions, arenecessar y tobridg e the apparent gap between the laboratory and the natural world. Many investigators used large,flexibl e plastic enclosures to study the fate and effects of pollutants on aquatic ecosystems (Schelske and Stoermer 1972, Menzel and Case 1977,Zeitzsche l 1978,Davie s and Gamble 1979). Whenw e use this approach topollutio n and dumping problems,correspondin g laboratory tests canb e validated (Hueck et al. 1978); and ecotoxicologists may be able ,_ tobette r predict the influence of pollutants in the field. In this way ; authorities canb e helped in setting standards tominimiz e damage to the eco­ system inquestio n (Hueck and Hueck-Van der Plas 1976;Hueck-Va n der Plas and ; Hueck 1979).

, Since197 4experiment s have been performed to study fate and effects of pol- i ] lutants on enclosed plankton communities inDutc h coastal waters.Th e inves- i tigations resulted in ametho d enabling one to experiment with natural plank- i ton systemswit h several trophic levels andwit h a large number of species within one experimental unit under semi-natural conditions.No-effec t levels found with thismetho d are among the lowest ones reported in literature (Kuiper 1980, 1981a,b). ! Dutch coastal waters,however ,ar e sediment-loaden and polluted. Ten to fif- j teen per cent of thewate r used in these bag experiments was freshwatar, one third ofwhic h came directly from the Rhine after a residence time of approximately onemont h in the North Sea, the remaining two thirds coming in­ directly from the Rhine via Lake IJssel (Zimmerman and Rommets 1974). Thebio - coenoses in thebag s mighthav e reacted differently toenvironmenta l stress ] 1 than those living in the open sea (cf.Fishe r et al. 1973,Fishe r 1977), because they had adapted themselves to pollution (Stokes et al. 1973,Jense n et al. 1974,Stockne r and Antia 1976,Moraitou-Apostolopoulo u and Verriopoulos 1979)o rbecaus e of other processes specific for this particular sea area. Results obtained would be of limited significance, if they could be applied only to aver y limited area of coastal waters near the Dutch coast.

A4-3 One of the aims of POSER (Plankton Observations in Simultaneous Enclosures inRosfjorden )wa s to investigate if ecotoxicological results obtained in the Rosfjord canb e compared with results obtained inDutc h coastalwaters . The Rosfjord was chosen because of its low freshwater input and because the waters were relatively unpolluted (Brockmann et al. 1981a).

During POSER two experiments were performed with mercury as amode l pollutant. Mercury was chosen since the results of many experiments with mercury and enclosed marine plankton communities were available for comparison (Grice and Menzel 1978,Davie s and Gamble 1979,Kuipe r 1977, 1980, 1981b,Toppin g et al. 1981).

A4-4 2. MATERIALS AND METHODS

2.1 SET-UP AND COURSE OF THE EXPERIMENTS

On March 6, 1979, the first experiment (POSER1 )wa s started. Six small bags (length 3m , contents 1.5 m )wer e filled simultaneously while we used a Vanton Flex-i-liner pump (Kuiper 1981b). The inflow hose of the pump was moved from a depth of 20m to0. 5 m during pumping. Four large bags (depth 40m , contents 30m 3) were filled while we used themetho d of Brockmann et al. (1981b).

OnMarc h 8, single doses ofmercury(II)chlorid ewer e added to the bags. Two smallbag s received a dose of 1u gHg. l ;tw o small and two large bags: 5 ugHg. l ;th e remaining bags served as controls (Table 1).T o add the mercury a stock solution ofmercury(II)chlorid ewa s rapidly mixed with 1001 of sea-water taken from thebag . This mixture was at once pumped back via a sprinkler,whic hwa s lowered in theba g during pumping to ensure thorough mixing with thecontent s of thebag .

The development of thephytoplankton , zooplankton and thebacteri a and a set of physico-chemical parameters influencing the development of thebiot a (nutrients,pH ,temperature )wer e monitored both in thebag s and in the fjord.

Table 1 Set-up of mercury experiments during POSER.

Experiment number Number of bags Initial mercury Nutri.ent s added and starting date concentration small large Ug Hg.l"1

1 6-3-1979 2 2 no 2 1 no 2 2 5 no

2 16-3-1979 2 1 no 1 1 5 no 2 - - yes 2 ~* 5 yes

A4-5 All samples,excep t thoseo fzooplankton ,wer e taken daily, as arul e at 9 a.m.,wit h anon-metalli c sampler (Meyer bottle principle). The small bags were sampled at depths of 0.5 and 2.0 m, the large bags and the fjord at depths of 1,3 , 10,2 0 and 35m . Zooplankton samples were taken twice aweek . In the small bags almost awhol ewate r column (0-2.5m )wa s sampled by means of a pipe (length 3m , diameter 4.5 cm) tipped with.abal l valve. Each sam­ ple (20 1)consiste d of five random lowerings of the pipe. The contents of the pipewer e filtered through a 55 urnne t and themateria l retained was immediately fixed and preserved in a 4% formaldehyde solution in filtered sea-water. Zooplankton samples in the largebag s were takenwit h a 55 urnne t with a conical opening,havin g a diameter of 80mm . Each sample (176 1)con ­ sisted of avertica l haul from a depth of 35m to the surface.

During the firstwee k of the experiment weather conditions were very bad and on March 12ther ewer e strong currents in Bóvéy Bight.Thes e currents caused a total exchange ofwate r masses in theB^r^ y Bight (Brockmann et al. 1981c). As a result a second aim,a direct comparison of thedevelopmen t of the plankton in the bags with that in the surrounding water,wa s difficult and the first experiment was stopped.

A second experiment (POSER2)starte d onMarc h 16 (day 0)b y filling seven small and two large bags (length 20m , contents 15 m3). It appeared that, within thene wwate r body now present in the Rosfjord,th ephytoplankto n spring bloom had already occurred as the nutrient concentrations in the water were very low. Since lowmineralizatio n rates would occur due to the low water temperatures (-1 to+ 1 C),w e did not expect large phytoplankton growth in thebags . Therefore,fou r of the seven smallbag s were spiked with nutri­ ents.

The amounts of nutrients added are considered tob e comparable to those present before the spring bloom (cf.Lännergre n and Skjoldal 1975). On day 2 -1 of the experiment we added single doses of 8uga tKNO3. I , 1.5 ugat

KH2PO4.I and 5 ygatNa 2SiFg.l to some bags (seeTabl e1) . Apart from thenutrient s single doses ofmercury(II)chlorid e were also added to some of the bags on the same day.A summary of the experimental design is shown inTabl e 1.

The same parameters weremeasure d as in POSER1 . At the end of the experiment (day 20) the sediment thatha d settled in the smallbag swa s collected for total mercury analysis.

A4-6 2.2 ANALYTICAL METHODS

Chlorophyll concentration was measured according to Strickland and Parsons (1968). Before 1-litre samples were filtered through Whatman GF/C filters, 1m l of a 1%magnesiu m carbonate suspension was poured on the filter. The filters were transferred to 10m l of 90%aceton e and thecell s were extrac­ ted after destruction in aBrau n MSK cellhomogenizer . After centrifugation pigmentconcentration s were measured witha VitatronMP S photometer system. Concentrations of chlorophyll and phaeopigmentswer e calculated according to equations given by Lorenzen (1967).

Phytoplankton samples from different depths of each bagwer e preserved with Lugols iodine for microscopical inspection with a Zeiss invertoscope. Selected samples were inspected and the main species were identified by means of thenomenclatur e givenb y Drebes (1974) and Imgram Hendey (1964).

The concentration and size distribution of suspended particulate matter in unpreserved samples were measured with a Coulter Counter model TAH with a population accessory,whil e we used a 100ur no r a 280 um tube or both (Sheldon and Parsons 1967).

Primary production was measured only in the smallbag s by Steemann Nielsen's (1952) llvCmethod ; 2m l of aNa ll+C03 solutionwa s added to 100m l samples in 125m l light and dark bottles.Ampoule s containing an activity of 2.70 uCi.ml were supplied by the International Agency of lf*C determination (H^rsholm,Denmark) .Th ebottle s were incubated in situ, two light bottles at a depth of 0.5 m, one light and one dark bottle at a depth of 2.0 m. After 4 hours of incubation (10 a.m. - 2p.m. ) the bottles were transported to the laboratory in a dark box,an d their contents were filtered through 0.45ur n membrane filters.Filter s were counted with aPackar d Tricarb liquid scintil­ lation counter.Th e inorganic carbon content of thewate rwa s determined by titration according to Strickland and Parsons (1968).

With aTechnico n auto-analyzer the concentrations of orthophosphate,nitrate , nitrite and reactive silicate were measured by the Strickland and Parsons' (1968) and Technicon procedures. The zooplanktonwa s counted, identified and measured by procedures described by Fransz (1975). The development of the bacteria in the waterwa s monitored by theDale y and Hobbie (1975)epifluorescenc e method.

A4-7 The mercury concentration in thewate r was measured by atomic absorption spectrometry with an IRDAB HGM 2300 spectrometer according to Spijk (1975). Mercury concentrations in the sediment were measured according to Tjoe et al. (1973). Themea n concentration in the upper 20 or 40m of thewate r columnwa s cal­ culated following a trapezial integration scheme. Most computations and statistical analyses were performed on the CDC 6400 computer of IWIS-TNO,Th e Hague.Fo r all caseswher e no confidence level is given p< 0.05 was tested.

A4-8 3. RESULTS

3.1TECHNICA L DIFFICULTIES

The strongwind s and the very low temperatures,wit h floating freshwater ice­ fields,highl y affected thematerials .Th e two layers of the bag foils had loosened from each other after one or twoweeks ,whic h reduced the mechanical strength of the laminate. Consequently, two large bags inPOSER 2showe d leak­ ages (the control after day 10,th e other after day 11)an don esmal l bag (mercury treated bag 6- after day10) .

During the experiments in the fjord (Brockmann et al. 1981c), strong water exchanges occurred,whic h made a comparison between the development of the plankton inside and that of the plankton outside thebag s only useful during the first phase of the experiments.Th ewate r exchanges in the Rosfjord led to strongvariation s in the salinity (30-34%). These salinity changes caused difficulties for the large bags. Whenever the salinity in the fjord was much higher than that in the bags, part of thebag s tended to float on the fjord. On one occasion thepoint s of the 20m bags were only a fewmeter s under the water surface.B y increasing theweigh t under the bags,w e caused the bottom of thebag s to sink again,bu t in this case theheavie r water outside pressed thebag s together at depths varying from 2 to 10m . For this reason sampling of the large bags after day 12durin g P0SER2was impossible (cf.Cas e 1978). The sampling of the small bagswa s not disturbed by the salinity variations.

3.2 MERCURY CONCENTRATIONS INWATE R AND SEDIMENT

The average mercury concentrations in thewate r during both experiments are shown in Figure 1.N o large differences were found between different depths, indicating that themercur y was well mixed after the addition. Only in the large bag during POSER2 th emercur y distribution with depthwa s nothomo ­ geneous. On day 3onl y 1.6 ugHg. l was found at adept h of 20m , in the upper layers (0-10m ) an average of 5.9 ugHg. l was measured.

Mercury concentrations in thewate r decreased in all bags. Owing toba d weather conditions during POSER1 samplin g was not possible on all days,bu t in the small bags concentrations seemed to decrease with more than 10%pe r day.Durin g the second experiment concentrations in the large bag decreased with 4%pe r day. In the smallbag , towhic h no nutrients had been added, concentrations decreased by 5%pe r day after day 4. From day 3 to day 4 a

A4-9 mercury /jg.t'

so • • SB - POSER 1 ° LB - • SB - POSER2 ' LB - so »A S8N-

i.0

time Idaysl

Fig.1 Mercur yconcentration s inth ewate ro fth ebag sdurin gPOSE R 1(initia l concentrations 1an d5 y gHg.l -1)an dPOSE R2 (initia lconcentration s 5p gHg.l -1)'.I nth esmal lbag sth eaverag eo f0. 5 and2. 0m dept hi s presented.I nth elarg ebag sth eaverag eo fal lsample si sgiven . SB= smal lbag ,L B= larg ebag ,SB N= smal lba gwit hnutrient sadded .

Table2 Mercur yconcentration so na wet-weigh tbasi si nth esedimen ti nth e small bags,collecte da tth een d ofPOSE R2

Bag Mercury Amounto f Amounto fmercur y Aa s% o fth e concentration sediment perba gu g mercuryadde d yg.kg" g (A) 1contro l 319 4.6 1.5 - 2contro l 274 5.5 1.5 -

35 UgHg.l " 1.280 5.5 7.1 0.1 4contro l+ nutrients 160 15.4 2.5 - 5contro l+ nutrients 7 28.7 0.2 - 6H g+nutri ­ ents 16.100 7.6 122.7 2.2 7H g+nutri ­ ents 18.600 16.5 307.5 5.6

A4-10 very rapid decrease was found. In the small bags, towhic h nutrients had been added, the average decrease inmercur y concentrations during the first weeks of the experiments was 11% per day. At the end ofPOSER 2th e sediment in the small bags was collected. Collection of sediment in the large bags failed. Table 2 shows the amount of sediment collected and themercur y concentrations found in these sediments. Mercury accumulated in the sediment,bu t only amaximu m of 5% of the added mercury could be recovered here.

3.3 PHYTOPLANKTON

At the start of POSER1 th e phytoplankton community consisted of diatoms (main species Skeletonema costatum,Thallassiosir a nordenskiöldii and Chaetocerosspp. ).Figur e 2 shows the chlorophyll concentrations in thebag s and those in the fjord during the first experiment. In allbag s chlorophyll concentrations decreasedlik ethe y didi nth e fjord. Since addition of mercury did not influence the development in the small bags, the average concentra­ tion for all bags ispresented . Constant pH and constant silicate concentra­ tions aswel l as the increasing importance of carotenoid pigments relative tochlorophyl l indicate that thephytoplankto n was not active during this period.

POSER 1 0 Chlorophyll • SS mg.m'3 AA LC , + * LM o fjord

•

0. t 1 \ \ // ^

Start POSER 1 1 1 1 1 L

Fig. 2Developmen t of chlorophyll concentrations in thebag s and the fjord during POSER 1.Vertica l bars indicate ± one standard deviation. SB - smallbag ,L C = large control bag,L M = large mercury-polluted bag.

A4-11 At the start of the second experiment (POSER2)th ewate r had a relatively low salinity (29-30%) and a low temperature (0-2°C) (Brockmann et al. 1981c). After filling,th evertica l distribution of the phytoplanktonwa s homogeneous. Thallassiosiranordenskiöldi i was themai n species,othe r diatoms found included Thallassionemanitzschioides ,Chaetocero s debilis, Chaetoceros borealis,Nitzschi a seriata,Coscinodiscu s sp..Apar t from diatoms a few large flagellates were found (Peridinium sp. and Euglenasp.) .

Figure 3show s thever y similar chlorophyll concentrations in the small bags, _l not spiked with nutrients but one polluted with 5 ugHg. l .Concentration s of suspended particulate matter asmeasure d with the Coulter Counter (size range 4-128 um diameter) showed a similar pattern.Afte r day 6, chlorophyll concentrations increasedowin gt o growth of diatoms,mainl y Thallassiosira nordenskiöldii, and, toa lesse r extent,Chaetocero s debilis and Chaetoceros borealis. In the largebag s the development of average chlorophyll concen­ trations was very similar to that in these smallbag s (Kuiper et al. 1981).

Figure 4 shows the average chlorophyll concentrations in the bags that were spiked with nutrients.Th e addition ofnutrient s generated abloo m ofdia ­ toms. In these bags Thallassiosira nordenskiöldii againwa sth emai nspecie san d growth of Chaetoceros debilis and Chaetoceros borealis was also found. The maxima in the controls are approximately seven times the maxima in the bags without nutrient addition. The addition ofmercur y inhibited,for nutrient-enriche d bags,-the growth of the phytoplankton,whic h resulted ina dela y of themaximu m of four days as compared with the controls.Th e same species as in the controls generated this bloom. Concentrations of particulate matter in the nutrient-enriched bags showed the same patternwit h time as the chlorophyll.

InFigur e 5 the carbon assimilation at a depth of 0.5 m in thenutrient - spiked enclosures is shown.Agai n the inhibition of thephytoplankto n after the addition of mercury isver y clear.Becaus e the carbon assimilation rate waspartl ydetermine d by the biomass of thephytoplankton , the carbon assi­ milation perm g chlorophyll was computed. Figure 6 shows this relative carbon assimilation in themercury-treate d bags as relative to that in the controls. Since primary productivity at 0.5 m did not differ from that at 2.0m , the average relative carbon assimilation was plotted. In the bag that was not spiked with nutrients the relative carbon assimilation was in­ hibited during thewhol e experiment. In the bags towhic h nutrients had been

A4-12 POSER 2

15 20 *» lime Idaysl

Fig.3 Chlorophyl l concentrations during POSER2 i nth esmal lbag swhic hwer e not spikedwit h nutrients.

chlorophyll ; . mg.m''

(5 20 lime Idaysl

Fig.4 Chlorophyl l concentrations during POSER2 i nth esmall ,nutrient - spiked bags.

A4-13 100 carbon assimilation mgC.m'3. tt.hr'

60

Fig.5 Carbo n assimilationi n the smallbag s with nutrient additiona ta 10 IS time (days) deptho f0. 5m

200 % of relatie carbon assimilation

150 POSER2 o with nutrients • without nutrients

100

Fig.6 Th erelativ e carbon assimilation (average 0.5an d2. 0m depth )i n mercury-treated bagsa s apercentag e oftha ti n thecontrols .Vertica l I addition time Idaysl bars indicate± 1 s.d. .

A4-14 added,th e relative carbon assimilationwa s lower than in the controls till day 8. From day 10t oda y 15 the relative carbon assimilation was higher in themercury-treate d bags than in the controls.Nutrien t depletion in the controls probably caused this difference.

Figure 7a-d shows the concentrations of silicate,phosphate ,nitrat e and nitrite during the second experiment in the nutrient-enriched bags. Directly after the addition of nutrients phosphate,silicat e and nitratewer econsume d in the controls by the diatom bloom until one of these nutrients,probabl y silicate ornitrat e orboth , reachedgrowt hrat elimitin gconcentration . The addition ofmercur y delayed the consumption ofnutrients ,a s could be ex­ pected from the development of phytoplanktonbiomas s and productivity.

Fig. 7a nitrate /jmol.l Fig. 7b POSER2

;s 20 IS 20 time (days) time {days!

phosphate /jmoi. r' Fig. 7c Fig. 7d 2.5

2.0 nitrite jumot.l i 0.3 15 •i s 02 W

* i 0.5 •H i 15 20 15 20 -*• time (days) *• time (days)

Fig. 7Developmen t of the concentrations of silicate,nitrate ,phosphat e and nitrite in the smallbag swit h nutrient addition.

A4-15 3.4 ZOOPLANKTON

During the first experiment the zooplankton consisted mainly of calanoid copepods. Nauplii of Calanus finmarchicuswer e themos t important innumber s (cf. Brockmann et al. 1981c formor e details). Numbers were very low (1-2 per litre). Therefore the samples from the small bags (20 1)wer e not repre­ sentative for-detection of significant differences between the bags. In the large bags larger samples were taken (176 1), and thenumber s of copepods appeared tob e lower in themercury-treate d bags than in the controls.How ­ ever, the first experiment was too short to allow conclusions.

At the start of POSER2th e species composition of the zooplankton community resembled that of the preceding period. Nauplii of Calanus finmarchicus were the most important innumbers .Apar t from this copepod,Acarti a clausi, Centropages hamatus,Pseudocalanu s elongatus,Oithon a similis and an uniden­ tified harpacticoid were found.

Figure 8 shows the total number of copepods (all species) in the fjord and in the large bags. The changes in the fjord are related towate r exchanges;

Number of copepods POSER 2 , • LC 0-20m ° LM 0-20m 1- fjord 0-20m ;5 • fjord 0-35m

/\ / \\ 1 1 \ 10 °~~~ ~"\^^ \ \\

•*•*— \ \\ ~~+ \ V 5 \\ \ / V

I addition \ . - time (days) s * 5 10 IS IS march

Fig. 8 Development of the total number of copepods in the largebag s and the fjord during POSER 2. LC »contro l bag,L M =mercury-pollute d bag.

A4-16 at the start of the experiment numbers of copepods in thebag swer e similar to those in the fjord. In both largebag s numbers of copepods declined, but therewa sa grea t difference between the bags. On day 10 65%o f the number found on day 1wa s counted in the control,bu t in the mercury-polluted enclosure hardly 5%wa s found. The decline of the copepods in the control wasmainl ycause d by declining numbers of Calanus finmarchicus. The other species seemed to thrive well. In themercury-pollute d system all species

seemed tob e influenced, although numbers of Pseudocalanus elongatus and the harpacticoid copepod were too low to allow conclusions.Durin g the experimen­ talperio d no growth of nauplii to largernaupli i or copepodites was observed. In all small bags, numbers of all species of copepods decreased sharply (Kuiper et al. 1981),bu t inmercury-treate d bags this decrease was faster than in the controls.Figur e 9 shows thenumbe r of copepods (all species) in the mercury-treated bags as apercentag e of that in the controls. The varia­ tion of this parameter in the small bags is larger than in the largebag s due to the lowernumber s of organisms in the samples from the small bags, and the subsequent larger error in the estimation of thepopulatio n density. It is clear,however , that addition of 5 ugHg. l enhanced the mortality of the copepods.

POSER2 ° LB a SBN » SS

50

time fdaysi

Fig. 9Number s of copepods (all species) in themercury-treate d bags asa percentage of those in thecontrols . LB = largebag , SB = smallbag , SBN = smallba gwit h nutrients added.

A4-17 3.5 BACTERIA

Figure 10show s thenumbe r of bacteria in the small bags during POSER1 . In the controls,number s increased from 2 to 5.105 bacteria perml . _1 Addition of 1an d 5 ugHg. l inhibited this increase,bu t at the end of the experiment bacterial numbers also increased in the polluted bags. In the large bags too few sampleswer e taken during POSER1 t o allow any conclusions on the influence of the added mercury.

Figure 11 shows the number of bacteria in the small bags during POSER2 . In the controls,t owhic h nonutrient s had been added,number s increased slowly during the experiment from 5 to about 14.105 bacteria perml . Addition ofmercur y inhibited this increase till day 8.Afte r day 12num ­ bers ofbacteri awer e comparable to that in the controls. In thenutrient - spiked controls bacteria at first showed a small increase until day 6, then a decrease occurred and, after day 10,number s increased toa maximum on days 14an d 15,respectively . In the nutrient-spiked bags treated with mercury,number s of bacteria were lower than in the controls during the first days following the addition of mercury,bu t after day 8 already numbers increased to amaximu m comparable to that in the controls. In the large bags no effects ofmercur y on the bacteria could be detected during POSER2 .

POSER 1 bacteria y /V»!Os.m/"' /

2 -

Iaddition of Hg

t 6 S »- time f days)

Fig. 10Developmen t of thebacteri a in thewate r in the small bags during POSER 1.

A4-18 20 • o controls (ö) »»5 jugHg.l bac eria ® Nx 0s ml" i/ 16 • , i

12 i V--•/ /// S • J1 T^f POSER 2 • o controls i » 5/jgHg.r'

1 , 20 W 15 20 *• time (days)

Fig. 11a Fig. 11b

Fig. 11Number s ofbacteri a in the small bags without nutrient addition (a) and with nutrient addition (b).Averag e of samples taken at 0.5 and 2.0 m depth.

A4-19 4. DISCUSSION AND CONCLUSIONS

The development of the plankton in large and small enclosures is compared in a separate paper (Kuiper et al. 1981). The general plankton development in the Rosfjord and in the controls is discussed in detail by Brockmann etal . (1981c,d,thi s volume). Here,th e resultswil l bemainl y discussed inrela ­ tion tomercur y and to former experiments with enclosed marine plankton communities treated with mercury (Grice and Menzel 1978,Davie s and Gamble 1979,Kuipe r 1977, 1980a,b,Toppin g et al. 1981).

4.1FAT E OFTH E MERCURY ADDED

Informe r experiments mercury concentrations decreased by 3-30%pe r day (Kuiper 1977, 1980, 1981b). Davies and Gamble (1979) found a decrease of 6% per day,Takahashi et al. (1977)3 % per day in theirba g experiment. The rate atwhic h mercury disappearedfro mth ewate r did not seem tob e related toba g size,bu t to the amount of particles available for adsorption and subsequent settling (cf.Toppin g et al. 1981). If more nutrientswer eavailabl e forpar ­ ticle production,mercur y concentrationsdecrease dfaster .Hig h phytoplankton activity also results inhighe r production of residuals of phytoplankton cells and extracellular, soluble organic matter,whic h could form organic complexes with mercury. These complexes could have a different adsorption behaviour to thewall s of thebag s and react differently with organisms.

At the end of POSER2a maximu m of only 5.6% of themercur y added could be recovered from the sediment. This low recovery must partially be attributed to incomplete sediment sampling,bu t the sampling errorwil l not have exceeded 100%.I n experiments inDe nHelde r approximately 25%wa s recovered at the end of experiments, amaximu m of 10%wa s adsorbed to thewall s and the remainderwa svolatilize d and had disappeared into the atmosphere. Davies and Gamble (1979)recovere d 25-30%,Toppin g et al. (1981) found 6-25% in the sediment and Takahashi et al. (1977)recovere d nearly all the added mercury in the sediment. Davies and Gamble (1979) suggest that the largerwal l surface :volum e ratio in their bags as compared with Takahashi's caused relatively higher adsorp­ tion to thewalls ,bu t Topping et al. (1981)di d notmeasur e more than20 % of the added mercury on the walls,whic h were made of PVC foil. Experiments with enclosures containing 68m ^ in Saanich Inlet showed that,pe r day, 1% of the added mercury was lost to the atmosphere (Azam,pers . comm.).

A4-20 The cause for thehighe r rate of mercury loss from small bags is probably related to the reduced mixing and to the relatively smaller surface area available for gas exchange with the atmosphere inver y largebag s incompari ­ sonwit h the small bags. It isno t surewhic h part of themercur y in the sea would disappear in the atmosphere under natural conditions,sinc e these con­ ditions will vary strongly with time and place. Our results supporthypo ­ theses about atmospheric mercury transport put forward by Wollast et al. (1975); Brosset and Svedung (1977)als o found strong indications that on the west coast of Sweden the polluted water of the North Sea,rathe r than indus­ trial air pollution,wa s the primary source of atmospheric mercury.

4.2 EFFECTS OF THE MERCURY ADDED

?ÏÎY.£°Eϧ2ÎS£22 During POSER1 n o influence of mercury on thephytoplankto n could be detected, which was probably due to the inactivity of thephytoplankto n in thecon ­ trols. Brockmann et al. (198ld)hypothesiz e that recently upwelled water had been enclosed and that unchelated trace metals inhibited phytoplankton growth.

During POSER2 th e relative carbon assimilation ratewa s inhibited after addition of 5 ugHg. l in theba gwithou t nutrient addition during the whole experiment (finalH g concentrations about 1.3 Ug.l ).I n thenutrient - spiked bags phytoplankton growthwa s inhibited as long as mercury concentra­ tions werehighe r than 2-2.5 ug Hg.l .Davie s and Gamble (1979)an d Thomas et al. (1977), alsoworkin g with enclosed plankton communities, found a transient reduction of the relative carbon assimilation rate after addition _1 _1 of 1u gHg. l .Additio n of 5 or 10u gHg. l had stronger,bu t also tran­ sient inhibitory, effects on the phytoplankton in theseexperiments .Fiv e days aftermercur y addition Thomas et al. (1977) found that phytoplanktonbio - mass increased again inbot h treated enclosures,mercur y concentrations being still ashig h as 1an d 5 ugHg. l ,respectively . During this time the water from the enclosures was still toxic to the outside phytoplankton, indicating that the phytoplankton in themercury-treate d bags showed some adaptation. On day 45 of the same experiment mercury concentrations in the enclosures were not toxic any more to the outside phytoplankton, although _1 the abovemercur y concentrations were still ashig h as 0.4 and 2.9 ugHg. l , respectively.

A4-21 In former experiments with Dutch coastal plankton addition of 0.5 or 1y g Hg.l had no significant effects on the enclosed phytoplankton. Addition of 5 ugHg. l inhibited the phytoplankton in all experiment as long as mercury concentrations in thewate r werehighe r than 1.5-2 ygHg. l (Kuiper 1977, 1980, 1981b). In one experiment a single dose of 50 ygHg. l was added, resulting ina very strong inhibition of thephytoplankto n during 29 days. After this period phytoplankton growth occurred again,mercur y concentrations _1 in thewate r being as high as 18y gHg. l Results of the present experiments may indicate inactivation of mercury by chelation or adsorption toparticle s (cf.Brian d et al. 1978). The fact that the relative carbon assimilation was not inhibited in the nutrient-spiked bags a few days after the addition of mercury,whil e in the bagwithou t nutrient addition the carbon assimilation was inhibited during thewhol e experiment,wa s probably due to lack of complexing substances which were not produced is such amounts as in thenutrient-spike d bags. Another explanation,no t excluding the former,ma y be that in theba gwith ­ out nutrient addition the phytoplankton was also stressed by thever y low nutrient concentrations. The additional mercury stress had a stronger influ­ ence thanwithou t nutrient stress.Cloutier-Manth a and Harrison (1980) found that the threshold of mercury toxicity to Skeletonema costatum decreased by an order of magnitude when NHit-limited cultures were starved of ammonia.

In former experiments Kuiper (1977, 1980, 1981b) showed changes in species composition after addition ofmercury ,whic hwer e probably related to reduced grazing after mercury addition. During POSER probably less than 2% was filtered per day by the copepods (using filtration rates givenb y Sonntag andParson s 1979). Thiswa s probably the reason that the species composition of the phytoplankton was not changed in these experiments. Neither did Davies and Gamble (1979) and Thomas et al. (1977) find changes in the species composition as adirec t effect of mercury addition.

During POSER2, addition of 5y gHg. l reduced thenumber s of copepods con­ siderably in comparison with the controls.Thi s was also found by Davies _1 and Gamble (1979) and Beers et al. (1977) after addition of 10o r 5y gHg. l . Beers et al. (1977) found also a delay in the development of the copepod populations. In former experiments inDutc h coastal waters addition ofmer ­ cury always delayed the development of the copepods (Kuiper 1977b, 1981).

A4-22 During POSER a delay could not be shown,sinc e the copepods did not develop in the controls either,a t this early spring bloom phase (Kuiper et al.1981 , Brockmann et al. 1981c). The presence ofnauplii indicates that the develop­ ment of Calanus finmarchicus had started in aprecedin g period,becaus e this species hibernates in the copepodite V stage in thesewater s (Matthews et al. 1978).

Bacteria During both POSER experiments numbers ofbacteri a were lower immediately after addition of 1o r 5y g Hg.l .Thes e small organisms showed the quick­ est response. Six and ten days aftermercur y addition in thebag s with and without nutrients bacterial numbers were however similar to those in the controls. Just aswit h phytoplankton, this finding indicates that nutrient addition accelerates the inactivation of the mercury via the increased production of complexing material,althoug h the development of adapted species cannot bé excluded. Davies and Gamble (1979)di d not find any effects of addition of 1o r 10 ugHg. l onnatura l bacterial populations,whic h may alsohav e been caused by their methodology. Azam et al. (1977), alsoworkin g with natural communi­ ties, founda stron g decrease of bacterial activity and numbers directly "_1 after addition of5 y gHg. l ,followe d by the development of amercury - resistant bacterial population,whic h reached higher numbers than in the control. In former experiments with Dutch coastal plankton the bacterial biomass was estimated by counting colony-forming units (CFU)o n agarplates .A decrease after addition of 5y gHg. l has never been detected,bu t after aperio d of approximately onewee k numbers of bacteria inpollute d enclosures were always higher than in the controls (Kuiper 1980, 1981b). The inhibitory effects during POSERwer e comparable to the results of Azam et al. (1977), the stronger increase of bacteria after addition ofnutrient s and mercury was comparable toresult s ofAza m et al. (1977) and Kuiper (1980, 1981b). This stronger increase isprobabl y caused by the increased availability of substrate by died phytoplankton and the absence of inhibitory substances, sometimes released by exponentially growing phytoplankton (Brockmann et al. 1977). In this respect, thebacteria ldecreas ewithi n thenon-polluted , but nutrient-spiked bags,betwee n day 7an d day 11, canb e interpreted as an inhibition effect by the exponentially growing phytoplankton at thistime .

A4-23 Com2arison_with_other_studies From the results available of enclosure experiments itca nb e concluded that mercury effects onnatura l plankton communities, isolated from fairly differ­ ent parts of theworl d (Saanich Inlet,Canada ;Loc h Ewe,Scotland ; Dutch coastal waters;Rosfjord ,Norway )ar e very similar and that themercur y con­ centrations at which effects were found, are also comparable (0.5-2 ygHg. l ) This indicates that results obtained inDutc h coastal waters can,wit h some confidence,b e extrapolated to other seaareas . It is interesting to compare the results of enclosure experiments with those of other toxicity studies. Several reviews concern the toxicity of mercury to aquatic life (Bryan 1976,Stebbin g 1976,Taylo r 1979,Davie s 1978,Lelan d et al. 1979). Laboratory experiments seldom show effects at concentrations lower than 1u gHg. l .Th e lowest concentration atwhic h effects were found _l _1 in enclosure experiments was 0.5 ugHg. l ;a single dose of 0.5 ug Hg.l changed the species composition of the periphyton (Grolle and Kuiper 1980).

Sigmon et al. (1977) found a reduction of the diversity of an epiphyton com­ munity inartificia l streams,an d Saward et al. (1974)detecte d small effects at concentrations of 0.1 ugHg. l .I nbot h studies themercur y concentrations were kept constant at the desired level in a flow-through system. The orga­ nisms were constantly exposed tone w doses of "ionic or reactive"mercur y (Davies and Gamble 1979). For copper and cadmium itha s been shown that their toxicity is correlated with the concentration of free metal ions and not with the total metal con­ centration (Sunda et al. 1980,Sund a and Lewis 1978,Enge l and Fowler 1979). Results of different bag experiments indicate that mercury is detoxified during these experiments.Additio n of nutrients accelerates this process, indicating a role for organic matter as a complexing agent. These experiments areno t conclusive,however , as towhic h forms ofmercur y are themos t toxic, or as to themechanism s by which mercury is detoxified.

4.3 EFFECTS OFNUTRIEN T ADDITION

The addition of nutrients to the small bags strongly stimulated phytoplank- ton growth. The rate of increase and theheigh t of the chlorophyll maximum was comparable to that of a spring bloom inLindaspollene , afjor d north of

A4-24 theRosfjor d (Lännergren and Skjoldal 1975). As secondary effects of this enhanced primary production the amount of sediment and the rate of decrease of the added mercury were higher in the nutrient-spiked enclosures. The species composition was not affected, although percentages changed. Stimulation of primary production is often found after nutrient addition (Berland and Bonin 1975,Parson s et al. 1977, 1978). Another secondary effect of the enhanced primary productionwa s that the rate of decrease of the added mercury was higher in thenutrient-spike d enclosures. This is important,sinc e inman y enclosure experiments which explicitly aim at the fate of the added pollutant,nutrient s were added. If apollutan t is added which adsorbs toparticles ,rate s of decrease will be found which are probably toohig h when compared with naturally occurring rates of removal (cf. Lee et al. 1978). In these cases addition of nutrients to the enclosed plankton communities should be avoided. f By comparison of the effects ofmercur y in the nutrient-enriched enclosed ecosystemswit hth e enclosures left in the original nutrient condition,i t canb e concluded that those plankton ecosystems would be affected themos t by heavy metals,whic h arewithi n the regenerative mode with lownutrien t concentrations.

Cloutier-Mantha and Harrison (1980) concluded the same from their laboratory j studieswit h Skeletonema costatum. This applies to the temperate regions J during the summer timewit h succeeding anabolic and catabolic phases.Bu t also during thewinte r phase when the phytoplankton isno t very active, production of dissolved organic substances is reduced so that in particular protection of bacteria that are still active will not occur.

ACKNOWLEDGEMENTS

Wewoul d like to thank Marijke van der Meer,Maria n Pullens and Dr J.H.L. Zwiers for their analytical work.W e gratefully acknowledge the enthousiastic help and advice of our German and Norwegian colleagues during and after POSER. The captain and crew of the RV Victor Hensen are thanked for their assistance during our stay in theRosfjord . The Netherlands Ministry of Foreign Affairs and theNorwegia n authorities are thanked for their help in providing experi­ mental facilities in the Rosfjord. Part of these investigations,carrie d out by Dr Brockmannwa s supported by the Deutsche Forschungsgemeinschaft via the Sonderforschungsbereich 94 "Meeresforschung"Hamburg .

A4-25 5.LITERATUR E CITED

Azam, F., R.F. Vaccaro,P.A . Gillespie,E.I .Moussall i and R.E.Hodso n (1977). Controlled ecosystem experiment: Effect of mercury on enclosed water columns. II.Marin ebacterioplankton . Mar. Sei. Comm. 3: 313-329. Beers, J.R., M.R. Reeve and G.D. Grice (1977). Controlled ecosystem pollution experiment:Effec t of mercury on enclosed water columns.I V Zooplankton population dynamics and production. Mar. Sei. Comm. 3:355-394 . Berland, B.R., D.J. Bonin and S.Y. Maestrini (1975). Isolement in situd'ea u deme r naturelle dans des enceintes de grand volume. CNEXO rapport scien­ tifique et technique,21 , 18pp . Briand, F., R. Trucco and S. Ramamoorthy (1978). Correlations between speci­ fic algae and heavy metal binding in lakes.J . Fish. Res.Boar d Can.35 : 1482-1485. Brockmann,U.H. ,K .Eberlein , G. Hentzschel,H.K . Schöne,D . Siebers, K.Wandschneide r and A. Weber (1977). Parallel plastic tank experiments with cultures of marine diatoms. Helgoländerwiss . Meeresunt.30 : 201-216. Brockmann,U.H. , E. Dahl and J. Kuiper (1981a). Concept of POSER. This volume. Brockmann et al. (1981b). Methods of POSER. Thisvolume . Brockmann, U.H., P. Kolterman,E .Dahl ,A . Dahle,K . Eberlein,A . Gaertner, G. Gassmann,K . Hammer,J . Jahnke,G . Kattner,M . Krause,J . Kuiper, M. Laake and K. Nagel (1981c). Water exchange in Rosfjorden during POSER. Spring 1979.Proc . Symposium onNorwegia n Current Systems,Septembe r 1980. Geilo,Norg e (in press). Brockmann,U.H. , G. Kattner and E.Dah l (1981d). Plankton spring development in a south Norwegian fjord. In:Marin e Mesocosms, G.D. Grice and M.R. Reeve (eds.). Springer,Ne w York (in press). Brosset,C . and L. Svedung (1977). Preliminary study of thepossibilit y of a relationship between high acidity in lakes and high mercury content in their fish populations. Swed.Wate r &Ai r Poll. Lab.Rept . B378, 29pp . Bryan,G.W . (1976). Heavy metal contamination in the sea. In:Marin ePollu ­ tion,R . Johnston (ed.), Academic Press,London ,etc. ,pp . 185-302. Case, J.N. (1978). The engineering aspects of capturing amarin e environ­ ment, CEPEX and others. Rapp.P . V. Réun. Cons. int.Explor . Mer 173: 49-58.

A4-26 Cloutier-Mantha,L . and P.J. Harrison (1980). Effects of sublethal concen­ trations ofmercuri c chloride on ammonium-limited Skeletonema costatum. Mar. Biol. 56:219-231 . Daley,R.J . and J.E. Hobbie (1975). Direct counts of aquatic bacteria by a modified epifluorescence technique. Limnol. Oceanogr. 20:875-882 . Davies,J.M . and J.C. Gamble (1979). Experiments with large enclosed eco­ systems. Phil.Trans . R. Soc. Lond. B. 286:523-544 . Davies,A.G . (1978). Pollution studies with marine plankton. Part II.Heav y metals. Adv. Mar.Biol . 15:381-508 . Drebes, G. (1974). Marines Phytoplankton. Georg Thieme Verlag, Stuttgart, 186pp . Engel, D.W. and B.A. Fowler (1979). Factors influencing cadmium accumulation and its toxicity tomarin e organisms.Environm . Health Persp. 28:81-88 . Fisher,N.S . (1977). On thedifferentia l sensitivity of estuarine and open ocean diatoms toexoti c chemical stress.Th e Amer. Naturalist 111: 871-895. Fisher,N.S. ,L.B . Graham,E.J . Carpenter and C.F. Wurster (1973). Geo­ graphic differences inphytoplankto n sensitivity toPCB's .Natur e241 : | 548-549. j Fransz,H.G . (1975). The spring development of calanoid copepods populations J in theDutc h coastal waters as related to primary production. Proc. 10th Eur. Symp.Mar .Biol. ,Osten d 1975,vol . 2,pp .247-269 . ' Gray,J.S . (1974). Synergistic effects of three heavy metals on growth rates j of amarin e ciliate protozoan. In:Pollutio n and physiology of marine organisms,F.J . Vernberg and W.B.Vernber g (eds.),Academi c Press,Ne w ; York, etc.,pp .465-485 . Grice,G.D . and D.W. Menzel (1978). Controlled ecosystem pollution experi­ ment: effect ofmercur y on enclosed water columns.VIII . Summary of results. Mar. Sei. Comm. 4: 23-31. j Grolle,T . and J. Kuiper (1980). Development ofmarin e periphyton undermer - ' cury stress in a controlled ecosystem experiment. Bull. Environm. ; Contam. Toxicol. 24:858-865 . f Hueck,H.J . and E.H. Hueck van der Plas (1976). A system of tests forpoten ­ tial effects of chemicals in the aqueous environment. In:Principle s and methods for determining ecological criteria onhydrobiocoenoses . ' R. Amavis and J. Smeets (eds.),Pergamo n Press,Ne w York,pp .467-528 .

A4-27 Hueck,H.J. , D.M.M.Adema ,W.Chr . deKoe k and J. Kuiper (1978). Experiences with the validation of ecotoxicological tests.Proc . Research Seminar EEG u.Umweltbundesamt ,Berli n 1977.UB A Berichte 10:159-167 . Hueck-Vande rPla s, E.H .andH.J .Huec k (1979). Theprospectiv eassessmen to fenvi ­ ronmentaleffect so fchemicals .Dru gDesig n8 : 311-354. IngramHendey ,N . (1964). An introductory account of the smaller algae of British coastal waters.Par t V. Bacillariophyceae. Her Majesty's Stationary Office,London , 317pp . Jensen,A. ,B . Rystad and S.Melso m (1974). Heavy metal tolerance of marine phytoplankton. I. The tolerance of three algal species to zinc in coastal seawater . J. Exp.Mar . Biol.Ecol . 15:145-157 . Kuiper,J . (1977). An experimental approach in studying the influence of mer­ cury on aNort h Sea coastal plankton community. Helgoländerwiss .Meeres ­ unters. 30:652-665 . Kuiper,J . (1980). Continued investigations into pelagic microcosms subject to environmental stress by pollutants. III.Combine d effects of mercury and cadmium. Report MT-TNOno . CL 80/85,6 7pp . Kuiper,J . (1981a). Ecotoxicological experiments with marine plankton communi­ ties inplasti c bags.Proc . Symp. on enclosed marine experimental ecosystems Sidney,B.C .Canada .Publ . by Springer,Ne wYork , inpress . Kuiper,J . (1981b). Fate and effects ofmercur y inmarin e plankton communities inexperimenta l enclosures. Ecotox.Environm . Safety 5: 106-134. Kuiper,J. , U.H. Brockmann and E.Dah l (1981). The influence of bag dimensions on the development of enclosed plankton communities during POSER, this volume. Lännergren,C . and H.R. Skjoldal (1975). The spring phytoplankton bloom in Lindaspollene, a land-locked Norwegian fjord. Autrophic and heterotrophic activities in relation tonutrients . Proc. 10thEur . Symp.Mar .Biol. , Ostend,vol . 2:363-391 . Lee, R.F.,W.S . Gardner,J.W . Anderson,J.W .Blayloc k and J. Barwe11-Clarke (1978). Fate of polycyclic aromatic hydrocarbons in controlled ecosystem enclosures. Environm. Sei. Technol. 12:832-838 . Leland,H.V. , S.N. Luoma and J.M. Fielden (1979). Bioaccumulation and toxi­ city of heavy metals and related trace elements. Journal WPCF 51:1592 - 1616. Lorenzen,C . (1967). Determination of chlorophyll and phaeopigments.Spectro ­ photometry equations.Limnol . Oceanogr. 12:343-346 .

A4-28 Matthews,J.B.L. , L.Hesta d and J.L.W.Bakk e (1978). Ecological studies in Korsfjorden,Wester n Norway. The generations and stocks of Calanus hyper- boreus and Calanus finmarchicus in 1971-1974.Oceanologic a acta 1:277 - 284. Menzel, D.W. and J. Case (1977). Concept and design:Controlle d ecosystem experiment. Bull.Mar . Sei. 27: 1-7. Menzel, D.W. and J.H. Steele (1978). The application of plastic enclosures to the study of pelagic marine biota. Rapp.P.V . Réun. Cons.int . Explor. Mer 173:7-12 . Moraitou-Apostolopoulou, M. and G. Verriopoulos (1979). Some effects of sub­ lethalconcentration so f copper on amarin e copepod. Mar.Poll .Bull .10 : 88-92. Parsons,T.R. , K. von Brockel, P. Koeller, M.R. Reeve and 0. Holm-Hansen (1977). The distribution of organic carbon in amarin e planktonic food web following nutrient enrichment. J. Exp.Mar .Biol .Ecol . 26:235-247 . Parsons,T.R. , P.J. Harrison and R.Water s (1978). An experimental simulation of changes in diatom and flagellate blooms.J . Exp.Mar .Biol .Ecol .32 : 285-294. Perkins,E.J . (1979). The need for sublethal studies.Phil .Trans .R . Soc. Lond. B. 286:425-442 . Saward,D. , E.A. Stirling and G. Topping (1974). Experimental studies on the Effects of Copper and Mercury on aMarin e Food Chain. Presented at Nato Science Committee Conference on ecotoxicity of heavy metals and organo halogen compounds,Mon tGabriel ,Canada . Cited by Davies and Gamble (1979). Schelske,C.L . and E.F. Stoermer (1972). Phosphorus,silic a and eutrophica- tion of lakeMichigan . In:Nutrient s and eutrophication, G.E. Likens (ed.). Spec. Symp. Amer. Soc.Limnol . Oceanogr.,pp . 157-170. Sheldon,R.W . and T.R. Parsons (1967). A practical manual on the use of the coulter counter inmarin e science.Coulte r electronics sales comp., Toronto. Sigmon,CF. , H.J. Kania and R.J. Beyers (1977). Reductions inbiomas s and diversity resulting from exposure tomercur y in artificial streams. J. Fish. Res.Boar d Can. 34:493-500 . Sonntag,N.C . and T.R. Parsons (1979). Mixing an enclosed, 1300m 3 water column. Effects on the planktonic food web. J. Plankton Res. 1:85-102 .

A4-29 Spijk,S.J . (1975). Bepaling van sporen arseen,cadmium , chroom,koper ,kwik , lood en zink in zeewater d.m.v. atomaire absorptie spectrometrie. Rept. MT-TNO no. CL 75-104, 8pp . Stebbing,A.R.D . (1976). The effects of lowmeta l levels on a clonal hydroid. J. Mar. Biol. Ass.U.K . 56:977-994 . Steemann Nielsen,E . (1952). The use of radioactive carbon (C1 ) formeasur ­ ing organic production in.the sea.J . Cons. Perm. Int.Explor .Me r 18: 117-140. Stockner,J.G .an d N.J.Anti a (1976). Phytoplankton adaptation to environmen­ tal stresses from toxicants,nutrient s and pollutants - awarning . J. Fish. Res.Boar d Can. 33:2089-2096 . Stokes, P.M., T.C. Hutchinson and K. Krauter (1973). Heavy-metal tolerance inalga e isolated from contaminated lakes near Sudbury,Ontario . Can. J. Bot. 51:2155-2168 . Strickland,J.D.H . and T.R. Parsons (1968). A practical handbook of sea water analysis.Bull . Fish. Res.Boar d Can. 167: 1-311. Sunda,W.G . and J.A.M.Lewi s (1978). Effect of complexation by natural organ­ ic ligands on the toxicity of copper to aunicellula r alga,Monochrysi s lutheri. Limnol. Oceanogr. 23:870-876 . Sünda,W.G. ,D.W . Engel and R.M. Thuotte (1978). Effect of chemical specia- tion on toxicity of cadmium to grass shrimp,Palaemonete s pugio: impor­ tance of free cadmium ion.Environm . Sei. Technol. 12:409-413 . Takahashi,M. , G.T.Wallace ,F.A . Whitney and D.W. Menzel (1977). Controlled Ecosystem Pollution Experiment:Effec t ofmercur y on enclosed water columns IManipulatio n of experimental enclosures.Mar . Sei. Comm. 3: 291-312. Taylor,D . (1979). A review of the lethal and sublethal effects of mercury on aquatic life.Residu e Rev. 72:33-70 . Thomas,W.H ., D.L.R . Seibert and M. Takahashi (1977). Controlled Ecosystem Pollution Experiment:Effec t of mercury on enclosed water columns III Phytoplankton population dynamics and production.Mar . Sei. Comm. 3: 331-354. Tjioe,P.S. ,J.J.M .d e Goeij and J.P.W.Houtma n (1973). Automated chemical separations in routine activation analysis.J . Radioanalytical Chem. 16: 153-164.

A4-30 Topping,G. , I.M.Davie s and J.M. Pirie (1981). Processes affecting themove ­ ment and speciation of mercury in themarin e environment. Proc. Symp. on marine enclosed experimental ecosystems,Sidney ,B.C. ,Canada . Springer, New York, inpress . Wollast,R ., G . Billen and F.T. Mackenzie (1975). Behaviour ofmercur y in natural systems and its global cycle. In:Ecologica l toxicology research. Effects ofheav y metal and organohalogen compounds.A.D .McTntyr e and C.F. Mills (eds.), Plenum Press,Ne w York,London ,pp . 145-166. Zeitzschel,B . (1978). Controlled environment experiments in pollution studies. Proc. Oceanology Int. 78,Tech . Sess.B (Biol.Mar . Tech.) pp. 21-32. Zimmerman,J.T.F . and J.W. Rommets (1974). Natural fluorescence as a tracer in the Dutch Wadden Sea and the adjacent North Sea.Neth . J. Sea Res.8 : 117-125.

A4-31 MARINE ECOLOGY - PROGRESS SERIES Vol. 6: 161-174, 1981 Published October 15 Mar. Ecol. Prog. Ser.

Fatean dEffect s of Cadmiumi nMarin e Plankton Communitiesi nExperimenta l Enclosures41

Jan Kuiper

Laboratory for Applied Marine Research, Division of Technology for Society TNO,P .O .Bo x57,178 0A BDe n Helder, The Netherlands

ABSTRACT: Fate and effects of cadmium on the development of North Sea coastal plankton com­ munities, enclosed in plastic bags containing 1.5 m3 natural water, were studied in 2experiments . Both experiments were carried out in summer-autumn and lasted 11 and 7 weeks respectively. Cadmium chloride was added to different bags in single doses of 1, 5 and 50 |ig Cd 1~'; in the first experiment additional doses of 100 and 250 ug Cd I"1wer e added in the middle of the experiment to 2o f6 bags . At the end of the experiments 4-9 % of the added cadmium was recovered in the sediment, the remainder was still in the water phase. Addition of 5 or 50 ug Cd 1~' resulted in slightly higher phytoplankton concentrations in one experiment. The mortality of copepods was increased by addition of 250o r 100 ug Cd l"1; additio n of 50 ug Cd 1~' resulted in a much lower zooplankton biomass. Due to differences in the sensitivity of different species, addition of 50 \ig Cd l"1 caused a change in species composition of the zooplankton community compared to controls. Growth of Pleurobrachia pileus was inhibited at 5 and 1 ptg Cd r\ leading to differences in the numbers of copepods between controls and treated bags. At concentrations of 50 ug Cd l"1 or higher no P.pileus were found. Concentrations of 1-5 \igCó r1 occur locally in polluted waters. The first experiment lasted more than 3 months and showed that nutrient regeneration rates in the bags are large enough to enable the system to persist without artificial nutrient additions.

INTRODUCTION plastic bags suspended in natural water (e. g. Strickj- land and Terhune, 1961; Menzel and Case, 1977;; In nature large and little-understood variations fre­ Davies and Gamble, 1979). This type of research, usinjj quently occur in ecologically important parameters Dutch coastal water plankton communities, was such as population density and species composition. It started in 1974. When it had been shown that tha is therefore difficult to detect long-term effects of method of enclosing a plankton community in a plastic environmental stress in the field. Much of our knowl­ bag could be used for toxicological research (Kuipe-, edge of the influence of individual pollutants on the 1977a, cf. Takahashi et al. 1975), further investigatio i marine ecosystem comes from actual dumping practice was aimed at developing the method and at determir - and from tanker and other disasters. ing the impact of pollutants in low concentrations on Although laboratory experiments are indispensable the development of the enclosed system (Kuipef, and yield useful information, extrapolation of their 1977b, 1981). This paper describes 2 experiments i^i results to field conditions is at present difficult, if not which cadmium chloride was used as model pollutant. impossible. In order to assess the value of experiments Cadmium was chosen because of its increasing impor­ in the laboratory, there is a need for experiments with tance as a pollutant in the field (Abdullah et al., 197$; more complex systems that can be regarded as approx­ Preston, 1973; CEC, 1974; Ketchum et al., 1975). imating field conditions more closely (Ringelberg, Apart from studying fate and effects of cadmium dn 1973; Menzel and Case, 1977). the development of plankton communities in bags, th e To bridge the gap between the laboratory and the first experiment was also used to test the hypothesis aquatic environment, several investigators use large that nutrient regeneration rates in the bags are larç e enough to enable the ecosystem to persist over consii1 - ' This work was carried out under contract No. 227-77-1 erable time without artificial nutrient additions, whi< h ENV N of the E. C. Environmental Research Programme are necessary in much larger, stratified enclosed wat ir

A5-1 columns (Menzel and Case, 1977; Davies and Gamble, for the other analyses (integrated sample 0-2.5 m 1979). depth). During the second experiment samples for phy- toplankton, nutrients, etc. were collected with a non- metallic sampler consisting of 2 chambers that could MATERIALS AND METHODS be opened at any desired depth. To investigate the variation of selected parameters with depth, samples Experimental Handling were always taken at depths of 0.5 and 2.0 m. During the first experiment the ctenophore Pleuro- Construction of the bags and experimental proce­ brachia piîeus developed in considerable numbers. dures have been described in detail by Kuiper (1977a). These numbers were estimated as follows: a Secchi The first experiment started on 17 May and ended on disc (diameter 15 cm) was lowered to a depth of 1.5 m 18 August 1976. The intensive sampling programme in the bags. During a period of 30 min the number of P. described below ended on 15 June. The second experi­ pileus in the water column above the Secchi disc was ment started on 7 September and ended on 25 October recorded every minute. The average number recorded, 1976. multiplied by 47 (= total depth of bag/1.5 times area In each experiment 6 bags were filled simultane­ bag/area disc) gives an estimate to the total number of ously with about 1 400 1 of North Sea water, collected a P. pileus in the bag. The reproducibility of the mea­ few miles off shore. The bags were anchored near a raft surement itself is good (s. d. of a single measurement in the harbour of Den Helder (The Netherlands). Large was estimated to be 10 %), but no information about predators (Ctenophora, fish larvae) were prevented the error of the estimate was obtained. The advantage from entering the bags by filtration of the water of this method is that it gives an estimate of the number through a 2 mm net. of P. pileus without killing them. The bags are too In all experiments single doses of cadmium chloride small to allow an estimate of their number by were added. To this end 100 1 of water was pumped out sampling. of each bag into a PVC container and rapidly mixed At the end of the experiment the sedimented mate­ with 1 1 of a concentrated solution of cadmium chloride rial was collected from the bottom of the bag by in acidified water (0.1 ml 12 M hydrochloric acid). This SCUBA divers using a large, non-metallic injection mixture was at once pumped back into the bag through syringe. a PVC sprinkler, which was slowly lowered into the bag to ensure thorough mixing of the solution with the seawater in the bag. In the first experiment cadmium Analytical Methods chloride was added on Day 3 (the start of an experi­ ment being Day 0) to 4 bags to give initial concentra­ The chlorophyll concentration was measured tions of 1,5 (2 bags) and 50 ug Cd 1~'. Two bags were according to Strickland and Parsons (1968). The sam­ treated identically without addition of cadmium and ples of phytoplankton were preserved with Lugol's served as controls. Because on Day 22 no clear effects iodine (Vollenweider, 1969) and examined with a of the addition of cadmium were expected, an addi­ Zeiss inverted microscope (Utermöhl, 1958). The main tional dose was given on Day 23 to 2 bags to give species were identified, where possible using nomen­ concentrations of 100 and 250 ug Cd l"1 respectively. In clature given by Ingram Hendey (1964) and Drebes the second experiment cadmium chloride was added (1974). on Day 10 to give initial concentrations of 1, 5 (2 bags) Concentration and size distribution of suspended and 50 (ig Cd T1. As in the first experiment 2 bags particulate matter were measured in unpreserved sam­ served as controls. ples with a Coulter Counter, Model TA II fitted with a Samples, except those of zooplankton, were taken population accessory, using a 100 ^m or a 280 urn tube daily, as a rule at 9 a. m. Zooplankton samples were or both (Sheldon and Parsons, 1967, Gamble et al., collected by means of a pipe (length 3 m, diameter 1977). 4 cm) with a ball valve at the end, sampling nearly the Primary production was measured during the second whole water column (0-2.5 m). Each sample (15.7 1) experiment employing Steemann Nielsen's (1952) 14C consisted of 5 lowerings of the pipe into each bag. The method. Samples of 100 ml were added to 1 ml of 14 contents of the sampler were filtered through a 55 (im NaH[ C]0:i solution (ampoules with an activity of net, and the retained material was at once fixed and 3.6 n Ci ml-' were supplied by the International preserved in a 4 % solution of formaldehyde in filtered Agency for l4C determination, Horsholm, Denmark) in seawater. 125 ml light and dark bottles. One light and one dark During the first experiment the (non-metallic) zoo- bottle were suspended at depths of 0.5 and 2.0 m in the plankton sampler was also used to collect the samples bags. After 6 h of incubation (9 a. m. - 3. p. m.) the

A5-2 bottles were taken to the laboratory in a dark box, and where no confidence level is given P < 0.05 wa|s their contents filtered. tested. Each filter was put in a counting vial containing 10 ml of a scintillation solution (Anderson and Zeut- schel, 1970; Pugh, 1973). The vials were counted with a RESULTS Packard Tricarb liquid scintillation counter. The inor­ ganic carbon content of the water was determined by Phytoplankton During the First Experiment titration according to Strickland and Parsons (1968). The concentrations of orthophosphate, ammonia, The water that was used to fill the bags was rela­ nitrate, nitrite and reactive silicate were measured tively clear (Secchi disc visibility 2.6 m) and had i with a Technicon autoanalyzer according to Strickland salinity of 31 %o S. The temperature during the first 60 and Parsons (1968) and Technicon procedures. days was around 14 °C, thereafter températures The zooplankton was counted, identified and meas­ increased to 20 °C. Fig. 1 shows the cadmium concen­ ured by the procedures described by Fransz (1976); trations in the different bags during the experiment nauplii and copepodites of each species were divided into at least 6 size classes and the adults separated by May-August 1976 sex. Subsamples of the 15.7 1sampl e were examined control.bog t 5 with a microscope until at least 150 organisms had 'fjg Cd r', bog 3 been counted. Changes in population densities at the various stages of copepods development - copepods always form the major part of zooplankton biomass in the bags - were used to estimate development and mortality rates of selected species, using multiple regression analysis of abundance of size classes (stages) at the various sampling dates (Fransz, 1976). Production of organic matter by copepods was esti­ mated by multiplying the means between zero and the upper limit of the 95 % confidence interval of develop­ time .Ooys ment rates by the mean density and the weight incre­ :"V^- ment for each time interval (Fransz, 1976). Dry weights of the copepods were derived from regression of dry Fig. 1. Cadmium concentrations in water during first experi-t ment weight on céphalothorax length given by Robertson (1968) and Nassogne (1972). Other parameters measured include water tempera­ Most cadmium remained in the water phase. Even or ture, salinity, Secchi-disc visibility, oxygen concentra­ Day 93 1.4 u.g Cd 1~' was measured in Bag 3 (actua tion, pH and global radiation (Mollgorczynski radia­ initial concentration 1.5 u,g Cd r1). The developmen tion meter). Unfiltered water samples for cadmium of chlorophyll concentrations is shown in Fig. 2 (deptl analysis were preserved with nitric acid (supra pur, integrated samples 0-2.5 m). The initial concentration: samples pH = 2). Cadmium was extracted from the of 1 mg m~3 was relatively low. The phytoplanktor samples and analyzed by atomic absorption spec­ community consisted of diatoms (Biddulphia regia, B trometry following procedures described by Fonds and sinensis, Asterionella japonica, Chaetoceros spp.) Estrof (1973). Cadmium concentrations in the sediment large flagellates (cf. Rhodomonas sp.) and u-flagel- were measured by neutron activation analysis accord­ lates. Of these species the large flagellates and u- ing to Tjoe et al. (1973). flagellates started to reproduce immediately after the bags were filled, producing a chlorophyll maximum or Day 4 or 5. A minimum was reached on Day 7. On Day Statistics 14 a second chlorophyll maximum was produced by diatoms, of which Lauderia borealis was the main Most statistical analyses (Analysis of variance, Stu- species. Only low numbers of u.-flagellates and larger dent's-t, sign-test, Wilcoxon test) were performed on flagellates were found on Day 14. the CDC 6400 computer of IWIS-TNO, The Hague. Considering the first 22 days of the experiment it Computations for the zooplankton analyses were con­ appears that chlorophyll concentrations were higher ducted with the CDC 6600 of the Nuclear Centre at than in the controls after addition of 50 (ig Cd 1~', and Petten (with the help of Dr. H. G. Fransz of the Nether­ on the average slightly higher after addition of 5 \ig lands Institute of Sea Research, Texel). In all cases Cdr'. Analysis of variance gave an estimate for the

A5-3 Chlorophyll,mgm " 20-,

Fig. 2. Chlorophyll concentrations in different bags during first experiment variation coefficient of the chlorophyll measurement chlorapfiylt.mgm -3 from Day 4-22 (5 %). Using this variation coefficient it could be shown (Student's t test) that chlorophyll con­ centrations were higher after addition of 5 ng Cd H (P « 0.02) and 50 ug Cd T1 (P =S 0.05). In a short additional experiment, addition of 5 or 50 ug Cd r1 again led to higher chlorophyll concentrations in the bags (Kuiper, 1980). On Day 23 additional doses of 100 and 250 \ig Cd 1~' were added to one 5 \ig Cd 1~' bag and one control bag respectively. A clear influence of these additions could not be detected. Mineralization in the bags was apparently large enough to supply the phytoplankton with nutrients necessary for growth. It is also clear that later in the experiment (after Day 25) only a few species of small pi- flagellates were able to survive in the bags. The com­ parability with the natural system was by this time low. Fig. 3. Chlorophyll concentrations in different bags during second experiment (average of 0.5 and 2.0 m depth) After nearly 3 months the material of which the bags were made began to distintegrate, especially at the water surface. The first and second phytoplankton spp., Leptocylindrus danicus, Coscinodiscus sp., bloom (1st: flagellates, 2nd: diatoms) depleted the Asterionella japonica) and flagellates (Prorocentrum available nutrients until some nutrient reached growth spp. an n-flagellates). Different diatoms started grow­ rate limiting values. The nutrient which limited ing after the filling of the bags, Chaetoceros sp. being growth of the flagellates is not clear, it was probably the dominant species. These produced a chlorophyll not one of those measured. The growth of Lauderia maximum on Day 4; a minimum was reached on Day borealis was probably limited by lack of silicate (Van 10 (the day of addition of cadmium). After Day 10, Bennekom et al. 1975; Kuiper, 1977a). chlorophyll concentrations stayed relatively low in all bags. A second maximum, caused by growth of dino- flagellates and ^-flagellates {Prorocentrum redfieldi Phytoplankton During the Second Experiment being the dominant species), was reached on different days in the different bags. An influence of the addition The bags were filled on Day O. The salinity of the of cadmium on the development of chlorophyll con­ water was 317«. S, Secchi disc visibility 1.5 m. Tem­ centrations could not be shown. A detailed microscopi­ peratures were around 15 °C throughout the experi­ cal analysis of the phytoplankton could not show any ment. As in the first experiment, the added cadmium significant differences in the species composition remained in the water phase throughout the experi­ between the different bags (test of Wilcoxon). ment. The carbon assimilation was measured from Day 1- Fig. 3 shows the chlorophyll concentrations in the 16. It appeared that most of the differences between bags as a function of time. At the start of the experi­ the measurements on a single depth could be attri­ ment the community consisted of diatoms [Chaetoceros buted to differences in biomass. Therefore the result-

A5-4 ing values were divided by the chlorophyll concentra­ lation was relatively high, but the phytoplankton tions. This relative carbon assimilation at depths of 0.5 biomass did not increase much, indicating that a factor and 2.0 m is presented in Fig. 4. No large differences other than nutrients limited the phytoplankton were found between the bags. Depth, i.e. light regime increase during this period. had a large influence on relative carbon assimilation. The development of the concentrations of the nut­ Interestingly, after Day 10, the relative carbon assimi- rients is shown in Fig. 5. The first phytoplankton peak depleted the silicate and nitrogen compounds. After relstlue carbon assimilation this peak, nutrient concentrations remained below the 1 1 detection limit (Si < 1 \ig at l" , NH4-N < 0.2 \ig at l" , 1 1 N03-N < 0.1 (j,g at l" and N02-N < 0.1 ug at l" ) with mgC (mgchl) the exception of phosphate concentrations which stay around 0.6 fig at P l~'. Nutrients necessary to make phytoplankton growth possible after Day 5 were gen­ erated by mineralization of organic matter and used directly (cf. Podamo, 1974).

September - October 1976

0.8 NO2-N NH.-N

i - 0Â

——»- time, days .V-4-4-è—1*_ *—*- 0 ->0 0 10 20 30 15 20 Fig. 5. Concentrations of nutrients during second experiment tiine(üsvs)

Zooplankton During the First Experiment

The water used to fill the bags contained various zooplankton species; of these, calanoid copepods / 4- formed the major part of the biomass. Temora longicor- nis was the main species, followed by Acartia clausi, , >\'J\ hi Centropages hamatus and Pseudocalanus elongatus. In addition to copepods, larvae of bivalves and worms, 2- A /jgföq and nauplii and cyprids of barnacles were also found. Adult barnacles lived on the bottom of the bag at the %-^/f^' end of the experiment.

0- "1 • 1 • 1111 • 1 Figs. 6 and 7 show the development of Temora Ion- 10 15 20 gicornis and Centropages hamatus during the experi­ time(days) ment. From Day 0-23 the average for duplicate bags is 1 Fig. 4. Carbon assimilation in different bags from 9-15 h presented (Controls 1 and 5; 5 ng Cd l" no's 2 and 4). during second experiment at depths of 0.5 (a) and 2.0 m (b) The 2 aforementioned species, and also Acartia clausi,

A5-5 N r' t f b * A-t «., W 20 30 i.0

Fig. 6. Temora longicornis. Number of individuals in different bags during first experiment, (a) small nauplii; (b) large nauplii; (c) copepodites; (d) adults greatly increased in numbers during the first 40 days of (test of Wilcoxon, P S 0.05). The number of P. pileus the experiment. Although no eggs were counted, they after addition of 1 \ig Cd l"1 was, until Day 42, a factor were definitely produced, since the development time 10 lower than in the control. After addition of 5ng Cd from egg to small nauplius is 1-2 days. These nauplii I-1 only 1 ctenophore was found. In bags with cadmium developed to copepodites and adults in the bags. Dur­ concentrations higher than 5 u,g Cd T', P. pileus was ing the first 4 weeks there appeared to be continuous never detected. Addition of 250 ng Cd. 1~' killed most production of eggs and a subsequent increase of the T. longicornis and all C. hamatus. Addition of 100 ng numbers of small nauplii. Numbers of nauplii of T. Cd r' killed C. hamatus; mortality rates of T. longicor­ longicornis reached a maximum in the controls on Day nis (and also of Acartia clausi) increased. 22, copepodites on Day 30 and adults on Day 24. For C. hamatus maximum numbers of nauplii, copepodites and adults were found on Days 32, 36 and 29 respec­ Zooplankton During the Second Experiment tively. These data indicate that no clear cohorts could be followed throughout their development. After the The water used in the second experiment contained addition of 5 \ig Cd 1~' the total number of copepods (all a zooplankton community in which the dominant species, all stages) was on the average 19 % higher species was Acartia clausi, followed by Centropages than in the controls from Day 7-22. After addition of 50 hamatus, Temora longicornis, Euterpina acutifrons, HgC d r' numbers were similar to those in the controls. larvae of bivalves and worms, nauplii of barnacles, It was not until Day 35 that the cause of these zoeae of Carcinus sp. and Oikopleura sp. Figs. 9 and 10 unexpected observations was revealed. It appeared show the development of nauplii, copepodites and that Pleurobrachia pileus had developed in the con­ adults of A. clausi and of C. hamatus respectively. T. trols; this ctenophore is a predator on copepods (Fraser, longicornis was also present in considerable numbers 1962; Grève, 1970). Fig. 8 shows the number of P. during the experiment. In the controls the calanoid pileus in the bags. It is clear that addition of cadmium copepods developed from nauplii to adults. Around influences the development of P. pileus significantly Day 20 the nauplii reached maximum densities, A.

A5-6 nis and E. acutifrons were relatively more important than in the controls. Fig. 11 shows the situation in all bags on Day 34.

per bog "l" ƒ ^""\ 1 Mov-Auausl 1976 0 o control * 1 yg Cd f*.bog A 5 r • 50 - ...

: — Ur ne .days t -r-V—

Fig. 8. Pleurobrachia pileus. Number of individuals in diffe­ rent bags during first experiment

/ /• j,

v^-- ^v > A/ •1 -J/,/ September - October 1976 ',,/ 1 1 r^r control \ \ \ \ 3 3 I Ug Cd l" big \ \ 2 2 5 " " " 4 4 5 " " " \ \ 6 6 50 " " " s 0

Fig. 7. Centropages hamatus. Number of individuals in diffe­ rent bags during first experiment, (a) nauplii; (b) copepodites; (c) adults. Averages of replicate bags

clausi being the dominant species. In this experiment continuous reproduction also occurred throughout and clear cohorts could not be identified. Addition of 1o r 5 u

A5-7 ' 5 3 2 4 6 bag control control 1/jg Cd SjjgCd 5ugCd SOjjgCd addition A 46 50 63 75 51 15 total number pertiter

0000 /i \ ;£f/ r / 1000 ' S«ct«nb«r - October 1976 \ \ ^H Euterpina 5 1 control \ •"^ Temora P I 1 I ICentropages \ Acortia 100 t 3 3 1 ug Cd 1~' bag 5 " « n i 2 2 ~i<\ S " " i 4 4 50 ' V 6 6 a Fig. 11. Relative distribution of different copepod species present on Day 34 in various bags; second experiment

Fig. 10. Centropages hamatus. Number of individuals in dif­ ferent bags during second experiment, (a) nauplii; (b) copepoditesj (c) adults (U0 September - October 1976 i control 5 control 030 ï lug Cd f' 2 Sug Cd C1 Secondary Production During the Second Experiment i SugCd /"" 0.20 t50pgCd Cl Total secondary production by the calanoid copepods was estimated with the model of Fransz 0.10 (1976). During the first 3 weeks, secondary production &&£-\ "•time, doys increased from 2 to 20 mg dry weight m3 d~' (average of all bags, except Bag 6; 50 ug Cd l"1). Thereafter, secondary production declined to an estimated aver­ Fig. 12. Secondary production in different bags; second ex­ age of 10 mg m~3 d_1. Fig. 12 shows the total biomass as periment a function of time in the different bags as computed from the total numbers of different size classes and different bags as a function of time are also presented length-dry weight relations given by Robertson (1968) in Fig. 12. During the first bloom (Day 0-6) maximum and Nassogne (1972). Addition of 1 and 5 ug Cd l"1 did P/B ratios were found (average 0.21 ± 0.08, N = 11). not influence the biomass development, but after addi­ From Day 10-40 P/B ratios were around 0.11 (stan­ tion of 50 ug Cd r1 biomass was significantly lower (P dard deviation 0.03, N = 47). Addition of 50 ug Cd l-' < 0.01, test of Wilcoxon). Secondary production esti­ did not influence the P/B ratio. During the first experi­ mates and biomass data were used to compute the P/B ment the same analysis was done using zooplankton ratio (production per day divided by the mean data from Day 0-42. During this period the P/B ratio biomass) in the different bags. The P/B rations in the had 2 maxima. The first maximum was around Day 5

A5-8 (P/B = 0.22 ± 0.04, N = 5). The second on Day 21 (P/B (1979), the volume filtered by the copepods per day can = 0.15 ± 0.09, N = 4). A minimum (P/B = 0.07 ± 0.02) be computed from the densities of the copepods. Table was around Day 12; following the second maximum, P/ 1 shows the results of these computations. During the B ratios declined in all bags and from Day 30-40 P/B period with maximum densities of nauplii and copepo­ was 0.07 ± 0.03. In both experiments maximum P/B dites (Day 20-30) the water in the bags was filtered ratios occurred during periods of maximum phyto- totally more than once per day. The estimates of plankton biomass. secondary production and filtration rates, indicate the Daily primary production during the second experi­ large influence of grazing zooplankton in this experi­ ment (mg Cm"' d_1) was estimated to be twice the ment. During the second half of the experiment primary production measured at a depth of 0.5 m or development of phytoplankton biomass appeared to be roughly to be 10 times the chlorophyll concentration prevented by the grazing pressure. assuming an average relative carbon assimilation over the water column of 5 mg C (mg chl a)"1 (6 h)~' (Fig. 4). At the beginning of the experiment, primary produc­ Fate of the Added Cadmium Chloride tion was much larger than secondary production, but during the second half of the experiment they were of Fig. 1 shows cadmium concentrations in the water the same order of magnitude (primary production, 30- during the first experiments; similar results were 50 mg Cm'3 d"'; secondary production 10 mg Cm"3 obtained in the second experiment. Most of the added cadmium remains in the water phase. Table 2 lists the Using filtration rates given by Sonntag and Parsons cadmium concentrations in the sediment on a wet weight basis, and the amount of cadmium present in Table 1. Total volumes filtered by copepods in different bags the sediment at the end. of the 2 experiments. In the during second experiment. Filtration rates: nauplii, 0.5 ml first experiment 7.3 % of the added cadmium was -1 _1 d ; small copepodites, 10m l d ; large copepodites, 25 ml found in the sediment on Day 93; in the second, 4.6 % -1 1 d ; adults, 40 ml d" on Day 48.

altered volume (m m"3

3 0.11 0.05 0.07 0.09 0.11 0.05 Fate of the Added Cadmium 10 0.83 0.70 0.84 0.64 0.73 0.58 17 0.80 0.87 0.51 0.83 0.80 0.60 The cadmium added to the bags remained in the 21 1.30 1.37 1.21 1.29 1.19 0.50 system and accumulated very slowly in the sediment 28 1.44 1.40 1.12 1.26 1.08 0.20 34 0.69 0.63 0.81 0.95 0.59 0.11 due to adsorption and subsequent settling of sus­ pended particles (abiotic particles, phytoplankton cells, dead zooplankton, etc.). Adsorption to walls was negligible. Adsorption of cadmium to particles is much Table 2. Concentrations of cadmium in sediment of different bags at end of experiments less pronounced in sea water than in fresh water. This is probably due to the formation of stable CdCl2 in sea Bag Cd added Cd concentration % of added water (Hahne and Kroontje, 1973; Bryan, 1976, Raspor no. to bag (mg) (mg kg-1) wet Cd in et al., 1977). Preston et al. (1972) state that 18 % of the weight basis sediment total cadmium concentration found in a series of sam­ ples from British coastal waters was bound to the 1st experiment 1 0 1.4 particulate fraction, a high estimate compared with 2 7 28.6 7.3 those of other authors. Eaton (1976) found that on the 3 1.4 4.8 5.6 average less than 0.4 % of the total cadmium was 4 147 440.0 7.3 bound to the praticulate fraction (> 0.45 u). In our 5 350 1110.0 7.3 experiments unfiltered water samples were analyzed, 6 70 362.0 9.2 so that no information on the distribution of cadmium 2nd experiment in the water is available. 1 0 1.2 Ketchum et al. (1975) have also reported very con­ 2 7 18.4 5.0 stant cadmium levels after addition of cadmium to 3 1.4 4.6 4.1 4 7 26.1 4.6 marine micro ecosystems containing sediments. 6 70 lost during analysis Kremling et al. (1978) performed 2 experiments with plankton communities enclosed in 68 m3 bags to which

A5-9 1.3 \ig Cd 1~' was added. Cadmium concentrations in short, lower numbers of copepods were found after the water were nearly constant and at the end of the addition of 50 (xg Cd 1_1. Lower grazing pressure after experiments less than 1 % was found in the sediments. addition of 5 ng Cd 1~' seems improbable, since in the In our experiments the amount of cadmium recovered bags concerned even higher numbers of copepods in the sediment was higher; this is probably due to the were found than in the controls, due to less grazing by longer duration of the experiments and the higher Pleurobrachia pileus. The possibility that addition of productivity of the enclosed plankton community. 5 ng Cd I"1 stimulated phytoplankton growth cannot therefore be excluded, although the working mechan­ isms are unclear. Effects of Cadmium on Phytoplankton Tkachenko et al. (1974) found stimulation of phyto­ plankton growth, measured as carbon assimilation, Addition of cadmium chloride to the enclosed micro after addition of 1-10 \ig Cd r' to natural phytoplank­ ecosystem had the following impact on the phyto­ ton assemblages. Berland et al. (1977) report increased plankton: 1 ng Cd r' did not influence the phytoplank­ growth rates of Skeletonema costatum during the first ton; 5 and 50 ug Cd 1"' led to higher chlorophyll con­ day after addition of 25-100 ug Cdl"1; other inves­ centrations in the first experiment. In the second tigators have also reported growth stimulation after experiment, 5 and 50 (ig Cd r' did not influence the addition of cadmium to diatom cultures (Canterford et phytoplankton. The species composition was not al., 1978) or natural phytoplankton assemblages (Patin changed by addition of 1,5o r 50 \ig Cd 1"'. Between the et al., 1972 Ibragim and Patin, 1975). On Day 23 of the 2 experiments a short (11 d) experiment was per­ first experiment additional doses of 100 and 250 \ig formed, using the same methods, in which higher Cd r1 were added to 2 bags. Because the frequency of chlorophyll concentations were found after addition of sampling was much lower after Day 24 than before, 5 or 50 \ig Cd r1, due to growth of Chaetoceros spp. and because the interactions between the different (Kuiper, 1980). trophic levels became increasingly complicated (see In model ecosystems, higher chlorophyll concentra­ section on Zooplankton), it was not possible to show a tions can be the result of less removal from the water significant influence of the addition of these higher due to less prédation by herbivores, lower sinking cadmium concentrations on the chlorophyll concentra­ rates (e.g. caused by a different species composition of tions. The third maximum in the control and the bags the community) or can be caused by increased growth. to which 1, 5 or 50 ng Cd 1"' was added, occurred on Lower sinking rates seem improbable since the species Day 46, in the bag to which 250 ng Cd 1~' was added on composition was the same in all systems, and since Day 42. The fact that this maximum occurred earlier physical factors (turbulence, light) were the same in after addition of 100 and 250 \ig Cd r' might be a different bags. After addition of 50 pig Cd 1_1 the graz­ result of mineralization of dead zooplankton or ing pressure may have been lower, resulting from reduced grazing. inhibition of the development of the copepods. In the The lowest concentrations influencing the growth of

Table 3. Minimum concentrations exerting effects on marine animals. Based on the sources listed

Species Concentration Effect Source ((igCdl-'ï

Tigriopus japonicus 44 Time to reach F2 generation more D' Agostino and Finney (1974) than doubled Uca pugilùtor 1 Decreased swimming activity of zoea Vernberg et al. (1974) Stage I, reduced salinity and temperature tolerance; depressed respiration of Zoea V Pleuronectes flesus 5 Blood anemia Larsson (1975) Homarus americanus 6 Increased oxygen consumption Thurberg et al. (1977) Morone saxatilis 0.5-5 Depressed oxygen consumption Calabrese et al. (1977) Palaemonetes pugio 44 50% mortality Sunda et al. (1978) Eurypanopeus depressus 10 Decreased development rate Mirkeset al. (1978) Pleuronectus platessa 5 Reduced growth Westernhagen et al. (1978) Mysidopsis bahia 10 Reduced survival; reduced formation Nimmo et al. (1978) of brood pouches Mytilus edulis 50 Reduced development of Trochophora Lehnberg and Theede (1979) Laomedea edulis 3 Irreversible retraction of hydranths Theede et al. (1979) Pseudodiaptomus coronatus 1 Reduced feeding rates Sick and Baptist (1979)

A5-10 fresh water phytoplankton found in the literature vary during the first period the grazing pressure of P. pileus from 2-50 (ig Cd 1_1 (Hutchinson, 1973; Bartlett et al., was more intense in controls than in bags with 5 |ig 1974; Klass et al., 1974; Conway, 1978). Berland et al. Cdl1. (1976) investigated the influence of cadmium chloride The increased mortality rate of copepods in the con­ on 18 marine species. Growth rate-inhibiting concen­ trols due to grazing Pleurobrachia pileus makes trations varied from 5-500 ng Cd r' (mean 85 ag demonstration of a possible increase in mortality rates Cd r'), lethal concentrations for 16 of the 18 species resulting from addition of cadmium more difficult (cf. were equal to or higher than 250 ng Cd 1"'. Li (1978) Gibson and Grice, 1977). Addition of 250 [ig Cd r' and Hollibaugh et al. (1980) recorded growth rate killed most Temora longicornis and all Centropages inhibition after addition of 100 \ig Cd 1'. Tkachenko et hamatus. Addition of 100 |ig Cd 1~' killed C. hamatus al. (1974) found inhibition of several marine phyto- and increased mortality rates of T. longicornis (and planktons at a concentration of 100 |xg Cd 1'. Gener­ also of Acartia clausi). At the beginning of the experi­ ally, inhibition of phytoplankton appears to occur at ment, numbers of copepods, following addition of lower cadmium concentrations in fresh water than in 50 ng Cd r', were comparable to or lower than in the the marine environment. The differences in speciation controls. In the controls a certain grazing pressure of cadmium in fresh and salt water are probably existed because of developing P. pileus. It may there­ responsible for this difference. Due to the much lower fore be concluded that addition of 50 \ig Cd r1 results adsorption in seawater, the amount of cadmium which in increased mortality of copepods or a decreased reaches the cell is lower in seawater than in fresh development rate. water with the same cadmium concentration and cell In the second experiment the effects of cadmium density. were more clear. Addition of 50 ng Cd 1"' resulted in a The differences in the response of the phytoplankton much lower biomass of copepods than in the controls. in the 2 experiments could be due to any of a number of Not all species of copepods were influenced; this led to factors. Firstly, starting conditions differed widely changes in species composition, compared to controls. between experiments (species composition, nutrient In nature such shifts in species composition may exert concentrations, etc.). Secondly, in the first experiment important effects on higher trophic levels via selective cadmium was added before large phytoplankton feeding. blooms had occurred and nutrients were not depleted; In both experiments no significant influence on the in the second experiment cadmium was added just development of copepods after addition of 1 and 5 (ig after a bloom. Under these circumstances dying phyto­ could be shown. Recent literature reveals a decrease of plankton probably supplied large amounts of organic no-effect levels with time. Eisler (1971), reviewing the compounds able to complex the added cadmium. toxicity of cadmium to marine organisms, reports that Härdstedt-Romeo and Gnassia-Barelli (1980) showed some crustacean species were most sensitive, having that organic substances produced by phytoplankton 96h LC 50 values of 320-420 \ig Cd 1"'. According to cells can decrease the amounts of cadmium taken up Pavicic and Järvanpää (1974) the development of by phytoplankton. Mytilis galloprovincialis veligers was inhibited by 80 ^g Cdl"1. Rosenberg and Costlow (1976) reported that 50 ng Cd l"1 decreased the survival and develop­ Effects of Cadmium on the Zooplankton ment rate of some development stages of 2 estuarine crabs; 50 \ig Cd T1 was the lowest concentration exert­ During the first experiment addition of cadmium ing adverse effects on marine animals. Reviews by inhibited the development of Pleurobrachia pileus at Taylor (1977) and Davies (1978) also list very few all concentrations (1-250 ng Cdl~'). After addition of effects from cadmium concentrations lower than 50 ng 1 (ig Cd r' the numbers of this ctenophore were 10 Cd I"1. Taylor (1977) concludes that the range contain­ times lower than in the control; after addition of 5 ng ing 90 % of the literature data in sublethal effects to Cd I*' only 1 P. pileus was found. In the bags with marine organisms was 50-60 000 \ig Cd 1~'. Table 3 1 cadmium concentrations of 50 ug Cd l" or more P. lists additional effects of cadmium concentrations on pileus was never found. Unfortunately, by the time P. marine animals. In fresh water, adverse effects of cad­ pileus was observed, no replicate bags were available, mium seem to occur at lower concentrations (Biesinger because of the addition of 100 and 250 \ig Cd 1"' on Day and Christensen, 1972; Pascoe and Mattey 1977; Tay­ 23. However, the fact that during the first period num­ lor, 1977). The different speciation of cadmium in fresh bers of copepods in the 2 controls were lower than in water is probably also here the key factor for causing the bags to which 5 [ig Cd r' had been added (the this difference. The cadmium concentrations influenc­ numbers were on average even lower in the control ing the development of copepods and Pleurobrachia bag which was later sacrificed), indicates that also pileus in the present report are among the lowest

A5-11 reported in the literature. Other investigators reporting were higher than in the controls. In the second experi­ effects of cadmium at concentrations < 5 ng Cd 1~' ment addition of 1 and 5 ng Cd 1~' did not influence the mostly employed flow-through systems, in which cad­ zooplankton. Addition of 50 ng Cd 1~'resulte d in lower mium was refreshed and probably remained in a non- zooplankton biomass and a different species composi­ complexed form. In the bags part of the cadmium was tion as compared with the controls. Addition of 100 and presumably complexed by organic compounds, and 250 ng Cd I"' increased the mortality rate of the therefore less available for the biota. The importance copepods. of speciation of metals in relation to their toxicity and Although the single dose of cadmium, added to the bioaccumulation is increasingly acknowledged and bags, was probably complexed during the experiment, has also been shown for cadmium. For example, Pre- so that only a limited amount of 'biologically active' mazzi et al. (1978) documented a 5 fold increase of the cadmium was present in the waterphase, concentra­ EC 50 of cadmium to Selenastrum if EDTA was pre­ tions influencing the development of copepods and sent, and Sunda et al. (1978) showed that the free Pleurobrachia pileus are among the lowest reported in cadmium is mainly responsible for toxic effects. literature. These concentrations are comparable to Concentrations of 1-5 ng Cd 1"' are.comparable to those occurring locally in polluted water, indicating those occurring locally in the field; Bryan (1976) gives that in these waters cadmium may already have a 0.04 ng Cd-1 as the mean concentration in the N. E. detrimental influence on the ecosystem. Atlantic Ocean and 0.41 ng Cd 1~' for the North Sea. Acknowledgements. This work was carried out under con­ Boyden et al. (1979) found concentrations around 0.4 tract no. 227-77-1 ENV N of the EEC Environmental Research ng Cd 1~' in a Cornish estuary. Nürnberg and Valenta Programme. Thanks are due to my colleagues W. C. de Koek, (1979) list 53 ng Cd 1~' as average concentration for the M. Dogger-Rutten, G. Hoornsman, J. A.va n Noort-Koeman, P. Roele, B.Schrieken , J. van de Eikhoff, K.va n de Togt, H. van North Sea, being 10 times lower than our figures and het Groenewoud and L. Wittebrood, who helped with the those of Bryan (1976); this probably can be attributed chemical, biological and statistical analysis and with evaluat­ to differences in the analytical method used. Eaton ing the results. J. J. Formsma, D. Pompert and S. Visser are (1976) found a mean of 60 ng Cd 1~' in the Atlantic also thanked for their assistance. Help by other colleagues at 1 Delft is also acknowledged. I further like to thank Drs. G. C. Ocean and 230 ng l" in the Gulf of Maine. In polluted Cadée, H. G. Fransz and W.W .C . Gieskes ofTh e Netherlands coastal waters, such as the Bristol Channel or Scheldt Institute of Sea Research for help and advice. Finally, I thank estuary, concentrations can be as high as 10 ng Cd 1~' the authorities of the Royal Dutch Navy for kind cooperation (Abdullah, 1972); Holmes et al. (1974) even measured in supplying experimental facilities. P. B.Davi s corrected my English manuscript. 78 ng Cd I"1 in an estuary in Texas; maxima of 20-80 ng Cd l"1 were recorded by Kneip (1977) near New York; Wong et al. (1980) found 53 ng Cd T1 in LITERATURE CITED Hong-Kong waters. Abdullah, M. J., Royle, L. G., Morris, A. W. (1972). Heavy metal concentration in coastal waters. Nature, Lond. 235: 158-160 CONCLUSIONS Anderson, G. C, Zeutschel, R. P. (1970). Release of dissolved organic matter by marine phytoplankton in coastal and The cadmium added to the bags remained in the offshore areas of the N. E. Pacific Ocean. Limnol. Oceanogr. 15: 402-407 experimental system and accumulated very slowly 1 Bartlett, L„ Rabe, F.W. , Funk, W. H. (1974). Effects of copper, (< 1 % week" ) into the sediment, which collected on zinc and cadmium on Selanastrum capricornutum. Wat. the bottom of the bags. Res. 8: 179-185 Addition of single doses of 5 and 50 ng Cd 1_1 Bennekom, A. J., van, Gieskes, W. W. C, Tijssen, S.B .(1975) . resulted in higher phytoplankton biomass compared to Eutrophication of Dutch coastal waters.Proc . R.Soc . Lond. B 189: 359-374 controls in the first experiment. In the second experi­ Berland, B.R. , Bonin, D.J. , Guérin-Ancey, O.J. , Kapkov, V.I. , ment, no effects on phytoplankton were detected. In Arlhac, D. P. (1977). Action de métaux lourds à des doses both experiments the species composition of the phyto­ sublethales sur les caractéristiques de la croissance chez plankton was not influenced after addition of 1, 5 or la diatomée Skeletonema costatum. Mar. Biol. 42: 17-30 Berland, B. R., Bonin, D. J., Kapkov, V. I., Maestrini, S. Y., 50 ng Cd l"1. Arlhac, D. P. (1976). Action toxique de quatre métaux During the first experiment, addition of cadmium lourds sur la croissance d'algues unicellulaires marines. inhibited the development of Pleurobrachia pileus at C. r. hebd. Séanc. Acad. Sei., Paris (Serie D) 282: 633-636 all concentrations (1 - 250 ng Cd 1"').A t concentrations Biesinger, K. E., Christensen, G. M. (1972). Effects of various of 50 ng Cd r' or higher P.pileus did not develop in the metals on survival growth, reproduction and metabolism of Daphnia magna. J. Fish. Res. Bd Can. 29: 1691-1700 bags. The presence of P. pileus influenced the Boyden, C. R., Aston, S. R., Thornton, I. (1979). Tidal and development of the copepods in such a way that the seasonal variations of trace elements in two Cornish densities of copepods after addition of 1an d 5 ng Cd I"' estuaries. Estuar. coast, mar. Sei. 9: 303-317

A5-12 Bryan, G. W. (1976). Heavy metal contamination in the sea. and redistribution of Zn and Cd in estuarine system. In: Johnston, R. (ed.) Marine pollution. Academic Press, Environm. Sei. Technol. 8: 255-259 London, New York, San Francisco, pp. 185-302 Hutchinson, T. C. (1973). Comparative studies on the toxicity Calabrese, A., Thurberg, F. P., Gould, E. (1977). Effects of of heavy metals to phytoplankton and their synergistic cadmium, mercury and silver on marine animals, Mar. interactions. Water pollution research in Canada, Univ. Fish. Rev. 39: 5-11 Toronto, Inst. Environ. Sei. Eng. 8: 68-90 Canterford, 'G. C, Buchanan, A. S., Ducker, S. C. (1978). Ibragim, A. M., Patin, S. A. (1975). Effect of mercury, lead, Accumulation of heavy metals by the marine diatom cadmium and copper on primary production and phyto­ Dytilum brightwelli (West) Grunov. Aust. J. mar. Fresh- plankton in some coastal regions of the Mediterranean wat. Res. 29: 613-622 and read seas. Oceanology (Moscow) 15: 589-591 CEC (1974). Problems of the contamination of man and his Ingram Hendey, N. (1964). An introductory account of the environment by mercury and cadmium, Commission of smaller algae of British coastal waters. Part V: Bacil- the European Communities, Luxembourg lariophyceae (diatoms). Her Majesty's Stationary Office, Conway, H. L. (1978). Sorption of arsenic and cadmium and London their effects on growth, micronutrient utilization, and Ketchum, B.H. , Zitko, V., Saward, D. (1975).Aspect s of heavy photosynthetic pigment composition of Asterionella for- metal and organohalogen pollution in aquatic ecosystems. mosa. J, Fish. Res. Bd Can. 35: 286-294 In: Mclntyre, A. D.,Hills , C. F. (eds.) Ecological toxicology D'Agostino, A., Finney, C. (1974). The effect of copper and research. Effects of heavy metals and organohalogen com­ cadmium on the development of Tigriopus japonicus. In: pounds. Plenum Press, New York and London, pp. 75-90 Vernberg, F. J., Vernberg, W. B. (eds.) Pollution and phy­ Klass, E., Rowe, D. W.( Massaro, E. J. (1974). The effect of siology of marine organisms. Academic Press, London, cadmium on population growth of the green alga New York and San Francisco, pp. 445-463 Scenedesmus quadricauda. Bull. Environm. Contam. Tox­ Davies, A. G. (1978), Pollution studies with marine plankton. icol. 12: 442-445 Part II. Heavy metals. Adv. mar. Biol. 15: 381-508 Kneip, T. J. (1977). Effects of cadmium in an aquatic environ­ Davies, J. M., Gamble, J. C. (1979). Experiments with large ment. Metal Bull., Proc. 1stInt . Cadmium Conference, San enclosed ecosystems. Phil. Trans. R. Soc. Lond, B 286: Francisco, pp. 120-124 523-544 Kremling, K., Piuze, J., von Bröckel, K., Wong, C. S. (1978). Drebes, G. (1974). Marines Phytoplankton, Georg Thieme Studies on the pathways and effects of cadmium in con­ Verlag, Stuttgart trolled ecosystem enclosures. Mar. Biol. 48: 1-10 Eaton, A. (1976). Marine geochemistry of cadmium. Mar. Kuiper, J. (1977a). Development of North Sea coastal plank­ Chem. 4: 141-154 ton communities in separate plastic bags under identical Eisler, R. (1971). Cadmium poisoning in Fundulus hétéro­ conditions. Mar. Biol. 44: 97-107 clites (pisces, Cyprinodontidae) and other organism. J. Kuiper, J, (1977b). An experimental approach in studying the Fish. Res. Bd Can. 28: 1225-1234 influence of mercury on a North Sea coastal plankton Fonds, A. W., EstnM, A. J. (1973). Bepaling van sporen lood, community. Helgoländer wiss. Meeresunters. 30: 652-665 koper, zink, cadmium, nikkel, kobalt, mangaan en ijzer in Kuiper, J. (1980). Continued investigations into pelagic mic­ water d. m. v. gelijktijdige extractie en atoomabsorptie- rocosms subject to environmental stress by pollutants. II. spectrofotometrie H20 18: 465-467 Cadmium. Report MT-TNO no. CL 80/31, 72 pp. Fransz, H. G. (1976). The spring development of calanoid Kuiper, J. (1981). Fate and effects of mercury in marine copepod populations in the Dutch coastal waters as plankton communities in experimental enclosures. related to primary production. Proc, 10th Eur. Symp. Mar. Ecotoxicol. Envir. Safety 5: 106-134 Biol. Ostend, Vol. 2: 247-269 Larsson, A. (1975). Some biochemical effects of cadmium on Fraser, J. H. (1962). The role of ctenophores and salps in fish. In: Koeman, J. H., Strik, J.J . T.W .W . (eds.) Sublethal zooplankton production and standing crop. Rapp. P.Réun . effects of toxic chemical on aquatic animals. Elsevier, Cons. perm. int. Explor. Mer 153: 121-123 Amsterdam, pp. 3-13 Gamble, J. C, Davies, J. M., Steele, J. H. (1977). Loch Ewe Lehnberg, W., Theede, H. (1979). Kombinierte Wirkungen Bag experiment, 1974. Bull. mar. Sei. 27: 146-175 von Temperatur, Salzgehalt und Cadmium auf Entwick­ Gibson, V. R., Grice, G. D. (1977). Experimental observations lung, Wachstum und Mortalität der Larven von Mytilus on the effects of copper on copepods and other zooplank­ edulis aus der westlichen Ostsee. Helgoländer wiss. ton: controlled ecosystem pollution experiment. Bull. mar. Meeresunters. 32: 179-199 Sei. 27: 85-91 Li, W. K. W. (1978). Kinetic analysis of interactive effects of Grève, W. (1970). Cultivation experiments on North Sea cadmium and nitrate on growth of Thalassiosira fluviatilis ctenophores. Helgoländer wiss. Meeresunters. 20: (Bacillariophyceae). J. Phycol. 14: 454-460 304-317 Menzel, D. W., Case, J. (1977). Concept and design: Control­ Hahne, H. C. H., Kroontje, W. (1973). Significance of pH and led ecosystem pollution experiment. Bull.mar . Sei. 27:1-7 chloride concentrations on behaviour of heavy metal pol­ Mirkes, D. Z., Vernberg, W. B., DeCoursey, P. J. (1978). lutants: Mercury (II),cadmiu m (II), zinc (II) and lead (II).J . Effects of cadmium and mercury on the behavioral Environ. Qual. 2: 444-450 responses and development of Eurypanopeus depressus Hardstedt-Roméo, M., Gnassia-Barelli, M. (1980). Effect of larvae. Mar. Biol. 47: 143-147 complexation by natural phytoplankton exudates on the Nassogne, A. (1972). Etudes préliminaires sur le rôle du accumulation of cadmium and copper by the Haptophy- zooplankton dans la constitution et le transfer de la ceae Cricosphaera elongata. Mar. Biol. 59: 79-84 matière organique au sein de la chaîne alimentaire Hollibaugh, J. T., Seibert, D. L. R., Thomas, W. H. (1980). A marine en mer Ligure. Thesis, Amsterdam comparison of the acute toxicities of ten heavy metals to Nimmo, D. R., Rigby, R. A., Bahner, L. H., Sheppard, J. M. phytoplankton from Saanich Inlet, B. C. Canada. Estuar. (1978). The acute and chronic effects of cadmium on the coast, mar. Sei. 10: 93-105 estuarine mysid, Mysidopsis béhia. Bull, environm. Con­ Holmes, C. W., Slade, E. A., McLerran, C.J . (1974). Migration tam. Toxicol. 19: 80-85

A5-13 Nürnberg, H. W., Valenta, P. (1979). Studies on the ecotox- Steemann-Nielsen, E. (1952). The use of radioactive carbon icological base lines and speciation of heavy metals in (C14) for measuring organic production in the sea. J. Cons, natural waters and rain. Proc. Int. Conf. management and perm. int. Explor. Mer 18: 117-140 control of heavy metals in the environment, London. CEP Strickland, J. D. H., Parsons, T. R. (1968). A practical hand­ Consultants Ltd., Edinburgh, pp. 472—475 book of sea water analysis. Bull. Fish. Res. Bd Can. 167: Pascoe, D., Mattey, D. L. (1977). Studies on the toxicity of 1-311 cadmium to the threespined stickleback Casterosteus Strickland, J. D. H., Terhune, L. D. B. (1961). The study of in aculeatus L. J. Fish Biol. 11: 207-215 situ marine photosynthesis using a large plastic bag. Lim­ Patin, S.A. , Tkachenko, V. N., Ibragim, A.M. , Fedotova, L.V . nol. Oceanogr. 6: 93-96 (1974). Effect of some metals on primary production in the Sunda, W. G., Engel, D. W., Thuotte, R. M. (1978). Effect of coastal zone of the Caspian Sea. Oceanology (Moscow) 14: chemical speciation on toxicity of cadmium to grass 72-74 shrimp, Palaemonetes pugio: Importance of free cadmium Pavicîc, J., Järvanpää, T. (1974). Cadmium toxicity in adults ion. Environm. Sei. Technol. 12: 409-413 and early larval stages of the mussel Mytilus galloprovin- Takahashi, M., Thomas, W. H., Seibert, D. L. R., Beers, J., cialis Lam. In: Comparative studies of food and environ­ Koeller, P., Parsons, T. R. (1975).Th e replication of biolog­ mental contamination. International Atomic Energy ical events in enclosed water columns. Arch. Hydrobiol. Agency, Vienna, Austria, pp. 179-188 76: 5-23 Podamo, J. (1974). Essai de bilan annual du transfer de l'azote Taylor, D. (1977). A summary of the data on the toxicity of dans le bassin de chasse d'Ostende. I. Utilisation de various materials to aquatic life, Vol. II, cadmium. ICI Ltd. l'azote par le phytoplankton et le phytobenthos. Hy- Report no. BL/A/1778: 1-22 drobiol. Bull. 8: 46-52 Theede, H„ Scholz, N„ Fischer, H. (1979). Temperature and Premazzi, G., Bertone, A. F., Freddy, A., Ravera, O. (1978). salinity effects on the acute toxicity of cadmium to Combined effects of heavy metals and chelating sub­ Laomedea loveni (Hydrozoa). Mar. Ecol. Prog. Ser. 1: stances on Selenastrum cultures. Proc. Research seminar 13-19 EEG and UBA, Berlin. UBA-Berichte 10/78: 169-187 Thurberg, F. P., Calabrese, A., Gould, E., Greig, R. A., Daw­ Preston, A. (1973).Cadmiu m in the marine environment of the son, M. A., Tucker, R. K. (1977). Response of the lobster, United Kingdom. Mar. Poll. Bull. 4: 105-107 Homarus americanus, to sublethal levels of cadmium and Preston, A., Jefferies, D. F., Dutton, J. W. R., Harvey, B. R., mercury. In: Vernberg, F. J., Calabrese, A., Thurnberg, F. Steele, A. K. (1972). British Isles coastal waters: The con­ P., Vernberg, W. B. (eds.) Physiological responses of centrations of selected heavy metals in sea water, sus­ marine biota to pollutants. Academic Press, New York, pp. pended matter and biological indicators, a pilot survey. 185-197 Environ. Pollut. 3: 69-82 Tjioe, P. S., de Goeij, J. J. M., Houtman, J. P. W. (1973). Pugh, P. R. (1973). An evaluation of liquid scintillation count­ Automated chemical separations in routine activation ing technique for use in aquatic primary production analysis. J. radioanalyt. Chem. 16: 153-164 studies. Limnol. Oceanogr. 18: 310-319 Tkachenko, V. N., Mortina, S. V., Lukankina, E. V. The Raspor, B., Valenta, P., Nürnberg, H. W„ Branica, M. (1977). method of toxicological experiments and some results of The chelation of cadmium with NTA in sea water as a toxicological effects of heavy metals on marine and fresh model for the typical behaviour of trace metal chelates in water monocell algae. Trudy vses. nauchno-issled. Inst, natural waters. The science of the total environment 9: morsk. ryb. Khoz. Okeanogr. 100: 63-67 (1974} (cited by 87-109 Leland et al. 1977, Davies 1978) Ringelberg, J. (1973). Parameter dependent (temperature) UtermÖhl, H. (1958).Zu r Vervollkommnung der quantitativen tolerance levels and the influence of the complexity of the Phytoplankton-Methodik. Mitt. int. Verein theor. angew. biological system. Hydrobiol. Bull. (Amsterdam) 7: Limnol. 9: 1-38 106-114 Vernberg, W. B., DeCoursey, P.J. , O'Hara, J. (1974). Multiple Robertson, A. (1968). The continuous plankton recorder: A environmental factor effects on physiology and behaviour method for studying the biomass of calanoid copepods. of the fiddler crab, Uca pugilator. In: Vernberg, F. J., Bull. mar. Ecol. 6: 185-223 Vernberg, W. B.(eds. ) Pollution and physiology of marine Rosenberg, R., Costlow, J. D., (1976). Synergistic effects of organisms. Academic Press, New York, San Francisco, cadmium and salinity development of two estuarine crab London, pp. 381-425 species. Mar. Biol. 38: 291-303 Vollenweider, R. A. (1969). Amanua l on methods for measur­ Sheldon, R. W., Parsons, T. R. (1967). A practical manual on ing primary production in aquatic environments, Black- the use of the Coulter Counter in marine science. Coulter well, Oxford Electronics Sales, Toronto Westernhagen, H. von, Dethlefsen, V., Rosenthal, H., Fürsten­ Sick, L. V., Baptist, G. J. (1979): Cadmium incorporation by berg, G., Klinckman, J. (1978). Fate and effects of cad­ the marine copepod Pseudodiaptomus coronatus. Limnol. mium in an experimental marine ecosystem. Helgoländer Oceanogr. 24: 543-562 wiss. Meeresunters. 31: 471-484 Sonntag, N. C, Parsons, T. R. (1979). Mixing an enclosed, Wong, M. H., Ho, K. C, Kwok, T. T. (1980). Degree of pollu­ 1300 m3 water column: Effects on the planktonic food web. tion of several major streams entering Toloharbou r Hong- J. Plankton Res. 1: 85-102 Kong. Mar. Poll. Bull. 11: 36-40

This paper was submitted to the editor,- it was accepted for printing on June 18, 1981

A5-14 Publication P 81/15 1981-09-29

FATEAN D EFFECTS OF 5-NITROFUROICACID- 2 (NFA)O NA MARIN E PLANKTON INEXPERIMENTA L ENCLOSURES

Jan Kuiper

Laboratory forApplie d Marine Research,Divisio n ofTechnolog y for Society TNO,P.O . Box 57,178 0A B Den Helder,Th e Netherlands

A6-1 ABSTRACT

The fate and effects of a single dose of 0.3 and 1.0 mg 5-nitrofuroic acid-2 (NFA)pe r litre on amarin e plankton community enclosed in large plastic bags (contents 1.5 m3)wer e studied. The plankton communitywa s derived from North Sea coastal waters and themode l ecosystems were anchored in thehar ­ bour of Den Helder,th e Netherlands,wher e theywer e exposed to a natural light and temperature regime. Two experimental unitswer e polluted with NFA, a third served as acontrol .

During fourweek s the development of the phytoplankton, zooplankton and bac­ teriawa s followed, aswer e a set of physico-chemical parameters including nutrients, light,temperature ,etc. .

The nitro-group was removed from theNF Awithi n about one day of the addition of the compound to themode l ecosystems,probabl y as a result of the exposure to light. The intact NFA inhibited the phytoplankton slightly; the remaining molecule produced no detectable effects in the system. The development of the enclosed community was very similar in the different bags.

Key-words: controlle d experimental ecosystem,mesocosm ,marin e pollution, effects,nitrofurans ,phytoplankton , zooplankton,bacteria , degradation.

A6-2 1. INTRODUCTION

There is an obvious need to evaluate the ecological effects of chemicals discharged into the aquatic environment.Man y investigations have been initi­ ated in the fields of toxicology,biodegradabilit y and bioaccumulation to characterize various pollution problems,on e aimbein g thedeterminatio n of standards. Siichstudie s areusuall y performed in the laboratory. The extra­ polation of the results of laboratory experiments to the field is however difficult, if not impossible.

The primary objective of ecotoxicological research should be to predict effects of pollutants in the realworld . The use ofmode l ecosystems,whic h approximate natural ecosystems in anumbe r of important characteristics, could provide amethodolog y to assess thevalu e of laboratory tests.Man y types ofmode l ecosystems have been developed to this end in the past decade (DeKoe k and Kuiper 1981).

In aquatic ecology and ecotoxicology an increasing number of investigators used large,flexibl e plastic enclosures,suspende d innatura l waters to study the relation between the plankton and its environment (Strickland and Terhune 1961,Menze l and Case 1977,Parson s 1978,Davie s and Gamble 1979, Grice et al. 1980). The general aim of these investigations is to bridge the gapbetwee n laboratory and field conditions.Us e of this approach with pollution and dumping problems could help topredic t the influence of pol­ lutants in the field, and thus help the authorities in setting standards to minimize damage to the ecosystem in question (Hueck and Hueck-Van dePla s 1976).

For reasons of cost and ease of experimental handling, enclosures for eco­ toxicological application should be as small aspossibl e (Kuiper etal . 1981). Inou r laboratory plastic enclosures ofmodes t size (contents 1.5 m3, depth 3m )hav e been used to study the influence of stress by pollutants on North Sea coastal plankton communities (Kuiper 1977, 1981a,b). In 1978 a case studywa s performed to investigate the fate and effects of 5-nitro- furoic acid-2 (NFA) in thesemarin e model ecosystems,becaus e the dumping of chemical waste containing nitrofurans into the North Seawa s considered at that time.Nitrofuran s are possible pollutants in the environment,becaus e they are used as feed and food additives and asdrugs . Some nitrofurans are suspected tob emutageni c and carcinogenic agents (Klemencic and Wang 1978, Cohen 1978).

A6-3 2. MATERIALS AND METHODS

The construction of the enclosures and the operation procedures have been described indetai l by Kuiper (1977, 1981a). The experiment started on 11Augus t 1978 (called day 0) and lasted fourweeks . Sevenbag swer e filled simultaneously with approximately 1.5 m3 of natural North Seawater ,collect ­ ed a fewmile s off-shore.Th e bags were anchored near a raft in thehar ­ bour ofDe n Helder,wher e theywer e exposed to anatura l light and tempera­ ture regime.Fou r of thebag s were used in an experiment with chlorinated phenols,whic h is reported elsewhere.

On day 3o f the experiment a single dose of 0.3 and 1.0 mg NFA.l was added to two other bags, using themetho d described by Kuiper (1981a), the remain­ ing bagwa s treated in the sameway ,wit h the exception of the addition of a chemical; this system served as acontrol .

Daily measurements of chlorophyll concentrations,primar y production and total volume of suspended particles (size range 2.5-50.8 ym diameter) as measured with a Coulter counter, showed the development of the phytoplankton. Determinations of theparticl e size distributions in combination withmicro ­ scopicobservation s,gav e information about the species distribution of the phytoplankton. Two or three times awee k zooplankton sampleswer e taken,i n which thebiomas s and the distribution of species, subdivided in different length classes,wer edetermined .Th e development of bacteriawa s monitored by epifluorescence counts.

In addition to the phytoplankton, zooplankton and bacteria, a set ofphysico - chemical parameterswa smeasured , including nitrate,nitrite ,ammonia ,phos ­ phate, silicate,pH , light,temperature , salinity and concentration of NFA inwate r and sediment. Kuiper (1981a)describe d themethod s by which the | different parameters weremeasured . Concentrations of NFAwer emeasure d by 'i High Pressure Liquid Chromatography (HPLC) techniques (Visscher 1980). I I 100m l water sampleswer e preserved with 1m l of 89%phosphori c acid.

A6-4 3. RESULTS

3.1 NFA CONCENTRATIONS INTH E WATER

On day 3 single doses of 0.3 and 1.0 mg NFA.l" were added to the bags. In . theba g towhic h 0.3 mg.l had been added,NF A could not be detected on day 4. Inthe.syste mwit h thehighe r concentration,0. 1 mg NFA.l" was found on day 4,an d 0.07 mg.l on day 5.Durin g the remainder of the experiment NFA could not be detected in the bags.

3.2 DEVELOPMENT OF PHYTOPLANKTON AND NUTRIENTS

At the start of the experiment thewate r had a salinity of 28°/„„ and was relatively clear (Secchi depth 3 m). The water temperature was 18°C (at the end of the experiment temperatures had decreased to 14°C). The phytoplank- ton community at the startwa s formed by large diatoms,o fwhic h various Rhizosolenia specieswer e themos t important (R. setigera,R . robusta, R. delicatula,R . styliformis). R. delicatula and Thallassionema nitzschi- oides started to increase innumbers ,leadin g to a chlorophyll maximum (Figure 1)i n the control on day 6. After this day chlorophyll concentra-

t ° August-September 1978 - 1 £ • control S\ 4 0 3 mg NFA t'1 fii a to mg NFA 1 k ft ft \ \4 P \u • ,^4 ft

jr "Ml .addition ^ time,days i L. i . ... i

Fig. 1Concentration s of chlorophyll as a function of time in the control and theNFA-pollute d bags (average of samples taken at depths of 0.5an d 2.0m) .

tions fell rapidly to aminimu m on day 9, after which concentrations in­ creased again due to growth of Prorocentrummicans , P. scutellum,Nitzschi a longissima and a few u-flagellates. Amaximu m was reached in the control on

A6-5 day 17.A t the end of the experiment the phytoplankton community was formed by u-flagellates and Prorocentrum spp..Additio n of 0.3 and 1.0 mg NFA.l resulted in significantly lower chlorophyll concentrations than in the con­ trol on day 5 (t-test,p <0. 1 and p <0.02 , respectively). Lower concen­ trations of suspended particles asmeasure d with theCoulte r counter also indicated a lower phytoplankton biomass after addition of 0.3 or 1.0 mg NFA.l" .O n day 4 primary production was 16%lowe r than in the control after addition of 0.3 mg NFA.l" ,an d 19%lowe r after addition of 1.0 mg NFA.l (measured from 10-14h at depths of 0.5 and 2.0 m). From day 6 to day 9 chlorophyll concentrations inNFA-treate d bags were higher than in the con­ trol.Afte r day 9 the development of the phytoplankton was the same in all bags. Addition of NFA did not change the species composition of thephyto ­ plankton during the experiment.

Figure 2 shows the concentration of various nutrients in the different bags

August-September 1978 • control

addition^ \

0 10 20 30 0 10 20 30 —*-time,days —~time,days

Fig. 2Concentration s of phosphate, silicate,ammonia ,nitrat e and nitrite in the control and the NFA-polluted bags (average of samples taken at depths of 0.5 and 2.0 m).

during the experiment. The first diatom bloom consumed all silicate,an d nearly all phosphate and nitrogen-containing nutrients.Man y phytoplankton species prefer ammonia tonitrat e as nitrogen-containing compound. The pat­ terns of ammonia and nitrate decrease during the first phytoplankton bloom show that thiswa s also the case in this experiment. The small inhibition

A6-6 of the phytoplankton,directl y after addition of NFA,resulte d ina delayed consumption of phosphate, silicate and ammonium. The nitrite and nitrate concentrations were clearly higher after addition of NFA in comparison with thecontrol .

3.3 DEVELOPMENT OF THE ZOOPLANKTON

At the start of the experiment the zooplankton community was formed by cope- pods. Centropage s hamatus was themai n species,Temor a longicornis,Acarti a clausi,Pseudocalanu s elongatus and Euterpina acutifrons were also found. Apart from these copepods a fewPodo n intermedius, larvae ofworm s and nauplii of barnacles were found.A t the end of the experiment adult worms were found in the sediment on thebotto m of the bags. All calanoid copepods mentioned developed to adults in the bags, although very low numbers of P. elongatus were found. E. acutifrons disappeared from all bags during the experiment.

Figure 3 shows the development of nauplii,copepodite s and adults of Temora longicornis in thedifferen t bags.Althoug h thenumbe r of nauplii were somewhat lower at the end of the experiment,i t is improbable that these differences were caused by NFA. The development of the other species was also similar in the different bags.

mooOr -1 N.m J •-#-• mpoo 1000 :/ 8' nauplii 100 copepodites .J^-V, WfiOO

1000 I-- l£&? Aua"st -September 1978 p^r . control . 100 "OSmgNFA.C o lOmgNFAf1 10 10,000 adults ,'-*Zl\ 1000

100

10,

Fig. 3Developmen t of numbers of nauplii,copepodite s and adults of Temora longicornis in the control and theNFA-pollute d bags.

A6-7 During the experiment the copepod communities reached very high densities (> 400 individuals.1~ (all stages) around day 18).Tabl e Ishow s the total volumes filtered daily by the copepod community, as estimated from filtra­ tion rates for nauplii,copepodite s and adults as givenb y Sonntag and Parsons (1979). After twoweek s the contents of thebag swer e filtered more than once daily.

Table ITota l volume filtered daily by the zooplankton community in the bags. The average of all bags is shown,wit h between brackets one standard deviation (s.d.).

3 -3 -1 Day Volume m .m .d August - September 1978 • control & 0.3 mg NFA.t'1 a 1.0 mg NFA.C1 0 0.03 (0.00) 5 0.12 (0.04) 10 0.31 (0.04) 14 0.90 (0.13) ' \ 19 3.02 (0.38) 24 4.25 (0.60)

28 i addition *'* 3.99 (0.54) 'time, days

Fig. 4 Numbers of bacteria in thewate r of the control and theNFA-pollute d bags during the experiment.

3.4 DEVELOPMENT OF BACTERIA

Figure 4 shows the numbers of bacteria in the different bags; these reached a maximum simultaneously withmaximu m chlorophyll concentrations (day 6-7) and decreased to aminimu m around day 14.Bacteria l numbers increased again simultaneously with increasing phytoplankton activity. Unfortunately, samples from the control taken at day 6 and 7,wer e lost.Th e available data,however , do not suggest any influence of NFA on the development of the numbers of bacteria.

A6-8 4. DISCUSSION AND CONCLUSIONS

4.1 FATE OF NFA INTH EMODE L ECOSYSTEM

Amaximu m of 10%o f the added NFA could be detected in the first days after addition in theba g with thehighes t concentration. Later investigations in the laboratory showed thatNF Awa s unstable under acid conditions. Phosphoric acid had been added to the samples,s o that the lowmeasure d concentrations could be explained by decomposition of NFA in the samples.However , the laboratorium experiments also revealed that NFAwa s unstable in the light; photochemical decomposition had alsobee n found byMcCall a and Reuvers (1970).

The development of the concentrations of nutrients in themode l ecosystems gives evidencefo rth e fate ofNF A in the bags. In all bags nutrient concen­ trations decreased due to consumption by thephytoplankton .Directl y after the addition of NFA nitrite concentrations increased considerably, followed by an increase innitrat e concentrations.Assumin g that thenitro-grou p of NFA is released and reacts withwate r to form nitrite,a maximu m of 6.5 and -1 -1 1.9 umol.l could be formed after addition of 1.0 and 0.3 mg NFA.l , respectively. In the bags 4.9 and 1.5 umol.l weremeasured , respectively, on day 6. Ifw e add thedifferenc e in nitrate concentrations between NFA polluted bags and the control,a closed balance is found. The average increase of nitrate plus nitrite concentrations from day 4-19, in comparison with that in the control,ha s been given inTabl e II. It is clear that directly after addition NFA loses its nitro-group. No information was ob­ tained on the remaining molecule (probably furoic acid). In laboratory experiments rapid biodégradation of this compound has been found. Several

Table IITh e fate of NFA in themode l systems.

Added NFA NO2presen t NO2 measured Average increase of (NO2 +NÛ3 ) inNF A on day 6 concentrations in comparison with _. _. the control mg.l umol.l"! umol.l (umol.l"',betwee n brackets 1 s.d.)

0.3 1.9 1.5 2.60 (1.05) 1.0 6.5 4.9 5.37 (1.18)

A6-9 very sunny days directly followed addition of NFA (global radiation 1800- 2400 J.cm2.d~ ).Availabl e information suggests that the decomposition of NFA in thebag swa s photochemical.

4.2 EFFECTS OF NFA INTH EMODE L ECOSYSTEM

After addition of 0.3 or 1.0 mg NFA.l no effects could be detected on the zooplankton or thebacteri a in the bags. The lower primary production on day 4 resulted in lower phytoplankton biomass,measure d as chlorophyll and total particle volume on day 5.Afte r day 5 the phytoplankton development was slightly delayed in theNFA-treate d bags as compared with the control.

It appeared that the intact NFA inhibited the phytoplankton slightly. McCalla and Reuvers (1970) showed that 100m g NFA.l killed the alga Euglena gracilis;othe r nitrofurans caused "bleaching"o f chloroplasts,bu t NFA did not show this effect. Cohen (1978)ha s reviewed the toxicity of many nitrofurans. He reported that no effects of NFA areknown .H e also stated that nitrofurans lose their antibacterial activity or theirmutageni c or carcinogenic properties after removal of the nitro-group. The results of the present model ecosystem experiment confirm these data:NF A rapidly loses its nitro-group after exposure to the light; from the remaining compound no bio­ logical effects could be detected.

ACKNOWLEDGEMENTS

Thanks are due tom y colleagues W.R. de Jong,W.C .d eKoek ,J.F . de Kreuk, A.O. Hanstveit,G . Hoornsman,J.M . Marquenie,M . Pullens,P . Roele, B. Schrieken,M . van derMeer ,CT . van der Togt,H . van het Groenewoud and H. Visscher,wh o helped with the chemical,biologica l and statistical ana­ lysis,an d with the evaluation of the results.D . Pompert and S.Visse r are also thanked for their assistance. Acknowledgement is also due for help givenb y other colleagues atDelft . I would like to thank the authorities of the Royal Netherlands Navy for their kind cooperation in supplying experimental facilities.Par t of thiswor k was done in the framework of the E.C. Environmental Research Programme, contract no. 227-77-1 ENV N. P.B. Davis corrected the English.

A6-10 5. REFERENCES

Cohen, S.M.: Toxicity and carcinogenicity of nitrofurans. In:Carcinogenesis , vol. 4_:Nitrofurans . Ed. G.T. Bryan,Rave n Press,Ne wYork ,pp . 171-231 (1978). Davies, J.M. and J.C. Gamble:Experiment s with large enclosed ecosystems. Phil. Trans.R . Soc.Lond . B 286:523-544 (1979). Grice,G.D. , R.P.Harris ,M.R . Reeve,J.F . Heinbokel and CO. Davis:Larg e scale enclosed water column ecosystems.A n overview of foodweb I,th e final CEPEX experiment. J. Mar. Biol.Ass .U.K . 60,401-41 4 (1980). Hueck,H.J . and E.H.Hueck-Va n de Plas:A system of tests for potential effects of chemicals in the aqueous environment. In:Principle s and methods for determining ecological criteria onhydrobiocoenoses , R. Amavis and J. Smeets (eds.), Pergamon Press,Oxford ,pp . 467-528 (1976). Klemencic,J.M . and C.Y. Wang:Mutagenicit y of nitrofurans. In:Carcino ­ genesis,vol . A_:Nitrofurans . Ed. G.T. Bryan,Rave n Press,Ne w York, pp. 99-130 (1978). Kock,W.C .d e and J. Kuiper: Possibilities formarin e pollution research at the ecosystem level.Chemospher eJ_0 ,575-60 3 (1981). Kuiper,J. : Development of North Sea coastal plankton communities insepa ­ rate plastic bags under identical conditions.Mar . Biol.44 ,97-10 1 (1977). Kuiper,J. : Fate and effects of mercury inmarin e plankton communities in experimental enclosures.Ecotoxicol .Environm . Safety _5, 106-134 (1981a). Kuiper,J. : Ecotoxicological experiments withmarin e plankton communities inplasti c bags.Proc . Symp. on enclosed marine experimental eco­ systems. G.D. Grice Ed., Sidney,Canada . Springer Verlag, in press (1981b). Kuiper,J. ,Ü.H . Brockmann,H . van het Groenewoud and G. Hoornsman: Influ­ ences of bag dimensions on the development of enclosed plankton communi­ ties during POSER. Inpreparatio n (1981). McCalla,D.R . and A. Reuvers:Actio n of nitrofuran derivatives on the chlo- roplast system of Euglena gracilis:effec t of light.J . Protozool. 17, 129-134 (1970). Menzel,D.W . and J. Case:Concep t and design: controlled ecosystem pollution experiment. Bull.Mar . Sei. 27,1- 7 (1977).

A6-11 Parsons,T.R. : Controlled aquatic ecosystem experiments inocea n ecology research.Mar . Poll. Bull. 9^ 203-205 (1978). Sonntag,N.C . and T.R. Parsons:Mixin g an enclosed 1300m 3 water column: effects on the planktonic food web. J. Plankton Res.J_ ,85-10 2 (1979). Strickland,J.D.H . and L.D.B.Terhune :Th e study of in situmarin e photo­ synthesis using a large plastic bag. Limnol. Oceanogr. 6_,93-9 6 (1961). Visscher,H. : Biodegradatie van 4-chloorfenol en 2,6-dichloorfenol in zee- water. Report MT-TNO no.C L 79/107a, 86 pp. (1980).

A6-12 FATE AND EFFECTS OF 3,4-DICHLOROANILINE (DCA)I NMARIN E PLANKTON COMMUNITIES INEXPERIMENTA L ENCLOSURES

1) 2) JanKuipe r and Arnbjörn O. Hanstveit

Laboratory for applied marine research MT-TNO,P.O . Box 57,De nHelder , The Netherlands 2) Central Laboratory MT-TNO,P.O . Box 217,Delft ,Th e Netherlands

Summary

The fate and effects of single doses of 3,4-dichloroaniline (DCA) on North Sea coastal plankton communities enclosed by large plastic bags (contents 1.5 m3) were studied in two experiments lasting five and six weeks,respec ­ tively. Thebiodégradatio n ofDC A was also studied in laboratory experiments, which were carried out simultaneously, using water from the enclosed model ecosystems.

DCA was not degraded in the laboratory tests and probably also not in the enclosed plankton communities,althoug h concentrations inth ewate r decreased during the experiments. This decrease appeared to be partly caused by dif­ fusion ofDC A through thewall s of the enclosures. _1 After addition of single doses of 2, 10 and 25 (JgDCA. l no effects on the enclosed plankton community could be detected. Addition of 0.1 mg DCA.l had a clear influence on the species composition and thebiomas s development of the phytoplankton, changed the relative species composition of the zoo- plankton and resulted in lower numbers of bacteria. In addition to these effects 1 mg DCA.l limited the phytoplankton growth and resulted in mor­ tality and inhibition of growth of the copepods.

KEYWORDS :experimenta l ecosystem,biodégradation ,pollutio n effects, dichloroaniline,phytoplankton , zooplankton,bacteria .

A7-1 INTRODUCTION

The Oslo Convention for the prevention of marine pollution from dumping of wastes, which was concluded in 1972,ha s led to the development of tests to measure the biological properties (toxicity, persistence and bioaccumulation potential)o f such wastes. (Hueck-van der Plas, 1981;TNO , 1980).

The study of the ecotoxicological properties ofwaste s and chemicals isgen ­ erally carried out in the laboratory. There is,however , little experimental basis for extrapolating the results of laboratory tests to natural ecosys­ tems. Experiments with multi-species systems which more closely simulate natural ecosystems than laboratory systems, might offer amor e representative method for assessing the environmental effects ofwaste s and chemicals (OECD, 1981)

In aquatic ecology an increasing number of investigators are using large, flexible plastic enclosures suspended in natural waters,t o study the plank­ ton and its environment with the general aim of bridging the gap between laboratory and field investigations (Menzel and Steele, 1978; Davies and Gamble, 1979;Gric e and Reeve, 1982).

In 1977-1978 selected organic chemicals were added to North Sea plankton communities enclosed by large plastic bags. These bags were exposed under semi-natural conditions inorde r to investigate the fate of the chemicals and their effects on the enclosed system. A second objective was the direct com­ parison onth ebiodegradabilit y of these compounds in laboratory tests and in the semi-natural environment of the enclosed plankton communities.

In this exercise, 3,4-dichloroaniline (DCA), 2,4-dichlorophenol (DCP), 4-chlorophenol (4CP)an d phenol were used asmode l compounds.Thi s choice was based on differences in the biodegradability of these compounds found in laboratory tests, DCA being the most stable. In this report the results of two experiments inwhic hDC Awa s added to thebag s willb epresented . Results of the other compounds will be presented elsewhere (Kuiper and Hanstveit, 1982).

A7-2 DCA is an interesting pollutant inth e field because it is released byhydro - lytic reactions from methylcarbamate, carbanilate and acylanilide herbicides and fungicides used inagriculture ,propani l and diuron being themos t impor­ tant (Still and Herrett, 1976; Chow and Murphy, 1975; Viswanathan et al., 1978). Chlorinated anilines are also found in the waste water of dye manu­ facturing plants (Games and Hites, 1977).

A7-3 MATERIALS AND METHODS

The construction of the bags (depth 3m , contents 1.5 m3) and operation pro­ cedures used were described by Kuiper (1977, 1981a,b). In the firstexperi ­ ment, which started on 19 August 1977 (called day 0), DCA was added on day 3 in inital concentrations of 2, 10 and 25 M8 DCA.l .Thes e concentrations were chosen because a NOEC (No Observed Effect Concentration) of 5.6|Jg. l was measured ina laboratory test with the reproduction of Daphnia magna asa criterion (Adenia and Vink, 1981). 10|J gDCA. l was added to duplicate bags, 2 and 25 Mg.l to single bags, and one bag served as a control.Th e experi­ ment lasted 35 days. The second experiment started on 19Ma y 1978 and lasted 42 days. DCA was added to twobag s on day 5 in initial concentrations of 0.1 and 1.0 mg.l respectively. Two bags served as controls. DCA (Fluka) was purified by recristallization beforeuse .

At the start of each experiment thebag s were filled simultaneously withap ­ proximately 1.5 m3 natural sea water collected a few miles off shore. The bags were anchored on a sheltered location in the harbour ofDe nHelder ,Th e Netherlands.

During the experiment the development of the phytoplankton, zooplankton and bacteria was measured, as well as a set of physico-chemical parameters in­ cluding nutrients (phosphate, ammonia, nitrate, nitrite, silicate), pH, light, temperature and concentrations of DCA in the water and sediment. To follow the development of phytoplankton, chlorophyll concentrations were measured as well as the primary production,particl e volume distribution and species composition. Zooplankton organisms were counted, identified and measured by the procedures described by Fransz (1976). Nauplii and copepodi- tes of each species were divided into at least six size classes and the adults separated by sex. Subsamples were examined with amicroscop e until at least 150 organisms had been counted. Changes in thepopulatio n densities of the various stages of the copepod development - copepods always form the major part of zooplankton biomass in the systems - were used to estimate development and mortality rates of selected species using multiple regression analysis of abundance of size classes (stages) at the various sampling data (Fransz, 1976). The generation time is then the reciprocal of the development rate from nauplius to adult. Production of organic matter by copepods was estimated by multiplying the means between zero and the upper limit of the

A7-4 95% confidence interval of development rates by the mean density and the weight increment for each time interval between sampling dates (Fransz 1976). Dry weights of the copepods were derived from regression of dry weight on céphalothorax length given by Robertson (1968)an d Nassogne (1972). From the daily copepod production and theirbiomas s estimates,th e production per unit biomass (P/B ratio)wa s calculated.

Zooplankton samples were collected at least twice a week, taking integrated samples of 15.7 1 from nearly the whole water column (0-2.5 m).Th e total volume filtered per day by the copepod community was estimated from filtra­ tion rates per development class given by Sonntag and Parsons (1979). All other samples were taken daily, as a rule at 9 a.m. at depths of 0.5 and 2.0 m.

The various parameters were measured using the methods described by Kuiper (1981a,b), with the following exceptions. The incubation period for the car­ bon assimilation measurements was shortened from 6 to 4hour s (10.00 - 14.00 h, Savidge 1978). The extraction of chlorophyll was improved by using a Braun MSK cell homogenisator (Derenbach, 1969). During the first experiment numbers of colony forming bacteria were followed by plate counts (medium 2216 E, Oppenheimer and Zobell, 1952), during the second experiment the total number of bacteria was followed by epifluorescence counts (Daley and Hobbie, 1975). Inth e second experiment the concentration ofAT P was determined as described by TN0 (1980).

The concentrations of DCA during the first experiment were measured by gas chromatography (TN0,1980) . During the second experiment highpressur e liquid chromatography (HPLC)wa s used. Water samples were preserved by addition of 85% phosphoric acid to a final concentration of 1%. Before theHPL C analysis the sample was centrifuged. Methanol/H20/H3P04 (65:35:1)wa s used as carrier fluid; injection was performed with a 300 |jlinjectio n loop and aValc o in­ jectionvalve .A 250x 5m m stainless steel column filled with 4 g Lichrosorb RP 18, 10 pm (Merck), a Tracor 950 pump giving a flow rate of 1.5 ml.min" , and as detector, a Du Pont 837 UV spectrophotometer (\= 243 nm) was used. The detection limitwa s 0.05 mg.l

A7-5 A die away test, described byTN O (1980),wa suse d forth estud y ofth ebio - degradation inth elaboratory .

Water was collected from thebag s of the second experiment onda y5 and3 2 and brought toth elaborator y thesam e day,set s ofsi x30 0m lconica l flasks of polycarbonate (Nalgene) forth e first series and of glass forth esecon d series, contained 150m lo fnatura l seawater.To each seto f flasks 0.1mg. l or 1mg.l " ofDC Awa sadded .T othre e ofth eflask s 0.375pCi ,4. 5mCi/mmo l

14 of [ C]-DCAi n 1N H2S04 solution (Radiochemical Research Centre,Amersham ) was added, and oneo f these flasks was sterilized by theadditio n of1. 5m l of 10%solutio n of HgCl2 to serve as blank. Two flasks without added[ 14C] were used for the measurement of the ATP-concentration, using the method described above.

The flasks were incubated ina Ne w Brunswick Psychrotherm shaking apparatus at about 12°C (thetemperatur e of thewate r at the start of the secondba g experiment) inth edark .

Samples of five ml were taken, acidified with 0.1 ml 2N H 2S04,an dflushe d 14 with airt o remove the C02. Onem lo fth eacidifie d samplewa sadde d to1 5 ml of Lumagel scintillation liquid (Lumac B.V.)an d counted in a Packard TriCarb. Model 2650 scintillation counter with automatic background and quench correction.

In order to measure thepossibl e assimilation of [14C]-DCAi nth ebacteria , the rest of the acidified sample was filtered through membrane filters (Sartorius 0.2 (J™por e diameter) which were washed with 5 ml filtered sea water andadde d to1 5m lLumage l andcounte d asdescribe d above.

To study possible diffusion ofDC Athroug h theplasti c material ofth ebags , a laboratory experimentwa sperforme d asfollows : Glass tubes with a hole of 1 cm2, which was covered by the plastic bag material, were filled with 28.3m l of seawater . In thiswa yth esurfac et o volume ratiowa sth esam e asi nth efiel d experiment. 1mg. l unlabelled DCA and 0.85 |jCi[14C]-DCA was added,th econten twa ssterilize d byth eaddi ­ tion of HgCl., and the tube was placed in 265 ml sea water in a conical flask. The flask was incubated on the shaking-machine at 15°C, and at intervals samples were taken and counted as described above. At theen do f the experiment theactivit y inth eplasti c samplewa sdetermined .

A7-6 RESULTS

CONCENTRATIONO FDC AI NTH EWATE R

Fig. 1 shows theconcentratio n ofDC Ai nth ewate r inth ebag s during both experiments.

30 DCA.pg.P

August-September 1977

20 - ûi \ \

o —o ^ j S ;+•+.,->_>•*-—,-,—„

1.0 OCA, mg.l' t v May-June 1978 V/<\ OS

addition time, days lo-o o—o-o— o ( 20 30 40

Fig. 1.Concentration s ofDC Ai nth ewate r ofth ebag s during bothexperi ­ ments.* an d* DC Ai nsedimen t during2n dexperiment ,afte r addition of 1.0an d0. 1m gDCA.l- 1, respectively.

In thefirs t experiment a gradual decrease inth econcentratio nwa sobserve d after addition of 10an d2 5 (Jg.l toabou t half theinitia l concentration after4 weeks .I nth esecon d experiment this gradual decreasewa sals oobser ­ ved inth eba gi nwhic h 0.1 mg.l DCAwa sadded . Inth eba ginitiall y con­ taining 1 mg.l ,a relatively rapid decrease ofth econcentratio n wasob ­ served during thefirs t couple of days. After 21 days theconcentratio nre ­ mained constant at about 0.5 mg.l .Analysi s of theDC A concentration in

A7-7 sludge collected in the bags during the experiment did not yield concentra­ tions significantly different from those in the water.N o DCA was detected in samples of theba g material at the end of the experiment.

In the parallel laboratory degradation experiments which were performed in polycarbonate flasks, a decrease in theDC A concentration was found,identi ­ cal under both sterile and non-sterile conditions as demonstrated for 1 rag.l in Fig. 2. On the other hand no such loss was observed in glass flasks. This indicates that biological degradation did not occur in these laboratory experiments.

1.0

DCA,mg.l 0.5 °-8. - \

—»- time,days J L 20 30 UO

Fig. 2.DC A concentrations ina laboratory die-away test in polycarbonate (0, 0)an d in glass flasks (A).I n thepolycarbonat e flasks the con­ centrations in sterile (0)an d non-sterile (0)flask s are presented.

The amount ofDC A absorbed to the biomass remained nearly constant during the test at about 2-5%o f the initial concentration.

The laboratory diffusion experiments showed diffusion of DCA through the plastic of the bags. In 64day s 30%o f theDC A with an initial concentration of 1.0 mg.l in the water phase had diffused through theplasti cmaterial , only about 1% was absorbed to themateria l at the end of the experiments. At the end the concentration in thewate r phase inside was about 0.6 mg.l

DEVELOPMENT OF THE PHYTOPLANKTON DURING THE FIRST EXPERIMENT

The water that was used to fill the bags had a salinity of 30 °/00 and was relatively clear (Secchi disc visibility 2.0 m). Water temperatures decreased during the experiment from 18-14°C. The phytoplankton community at the start of the first experiment consisted of diatoms (of which Skeletonema costatum

A7-8 andLeptocylindru s danicus dominated),dinoflagellate s (Prorocentrum micans, Ceratium fusus) and various species of |j-flagellate s (o.a. Phaeocystis poucheti). The developeraent of phytoplankton biomass as measured by the chlorophyllconcentration s isshow ni nFig .3 .Afte rfillin gth ebags ,th e

August- September 1977 • control * 2fig DCA.I'' o » 10 » 25 \et+r**ç&

•time, day s

20 30

Fig.3 .Chlorophyl lconcentration s (averageo fsample stake na t0. 5 and2. 0m depth)i nth edifferen tbag sdurin gth efirs texperiment . chlorophyll concentrations increased to a maximum on day 5. This increase coincided with an increase inpopulation s of Leptocylindrus danicus, Phaeo­ cystis poucheti andothe r|J-flagellates .Fro mda y5-1 0 chlorophyllconcentra ­ tions decreased. After day 10a sligh t increase,du e togrowt ho f(J-flagel ­ lates, occurred. At the end of the experiment the phytoplankton community consistedo f y-flagellatesan da fe w Nitzschia longissima. Theadditio no f2-2 5(Jg. l DCAdi dno tsignificantl yinfluence dth edevelop ­ mento fth ebiomas so fth ephytoplankto no rth especie scomposition .

Theconcentratio no fparticulat ematte rwa smeasure dwit hth eCoulte rcounte r (particle diameter range 2.5-50 pm)i nth e different bags; thepatter nwa s the samea s that of the chlorophyll concentrations.Th edistributio n ofth e suspended particles over the different size categories corresponded totha t expected frommicroscopica lobservations .

Thecarbo nassimilatio nwa smeasure da tdepth so f0. 5 and2. 0m .Th epattern s of the carbon assimilation corresponded generally to those shownb ychloro ­ phyllan dparticulat ematter .T oexclud eth einfluenc eo fbiomass ,th ecarbo n

A7-9 assimilation per unit chlorophyll was calculated. This relative carbon assi­ milation is shown inFig . 4. In this experiment depth (i.e. light) appeared

rel. carbon assimilation mgC.Ch}'1.tmg chtf'

August-September 1977 m control 20 + 2fjg DCA I'I riVt* "'O

10

;^*^8*«!==s=t b

20 30 —»- time, days

Fig. 4.Th e relative carbon assimilation at depths of 0.5 (a)an d 2.0 m (b) in themode l ecosystems at different DCA concentrations during the first experiment. to be the most important factor regulating the relative carbon assimilation of thephytoplankto n community.

Fig. 5 shows the concentrations of selected nutrients as a function of time. Because no significant differences were found between the different bags, the results of all bags were averaged. Apparently the first phytoplankton peak consumed the greater part of silicate- and nitrogen containing nutrients. Phosphate concentrations decreased from 1.4 to 0.5 pmol.l .I t seems proba­ ble that the limiting factor of the first phytoplankton bloom must be looked for among silicate and the nitrogen containing compounds (v.Bennekor ae t al., 1975). During the remainder of the experiment nutrient concentrations re­ mained low. Only ammonia and nitrite concentrations increased again after day 20, indicating that another factor such as zooplankton grazing or another nutrient limited phytoplankton biomass after day20 .

A7-10 P N02. °. NHi.fjmol.l~' °f fjmol.t'' fjmot1 \ N03.fjmoU~' - \ Si ,/jmoir' u i, -0. 5

August -September 1977 7.5 V ° NHt 0 l * NOy Û VI • Si Ü PO •^^ ^ u 0.6 • N07

7.0

07.

0.5 7 - 0.2

/. 0~- 00 0.0 70 20 30 »- time,days

Fig. 5.Concentration s ofselecte d nutrients inth emode l ecosystems during the first experiment.Fo reac h sampling dayth eaverag e value ofal l samples ispresented .

DEVELOPMENT OFTH EPHYTOPLANKTO N DURING THESECON D EXPERIMENT

The water atth estar t ofth eexperimen t wasturbi d (Secchi disc visibility 0.8 m)an dha da salinity of2 6 /oo.Th etemperatur e ofth ewate r increased during thefirs t 2 weeks from 12t o17°C , andthe n decreased to15° Ctoward s the endo fth eexperiment . Inth ewee k before theexperimen t started chloro­ phyll concentrations inth etida l inlet toth eWadde n Sea,fro m whichth e water to fill theenclosure s wastaken , haddecrease d from approximately5 0 -3 -3 mg.m to 15mg.r a .Thi s decrease continued inth ebag s (Fig.6 )an dfro m day 5 when theDC Awa sadded , until day10 ,extremel y lowchlorophyl l concen- _3 trations were measured (0.01-0.05 mg.m ).Th ephytoplankto n bloom preceding this minimum wasgenerate d byp-flagellate s (partly P. poucheti) whilst dia­ toms (Nitzschia longissima, Leptocylindrus danicus, Skeletonema costatum, Nelosira sulcata, Rhizosolenia setigera) werepresen t insmal lnumbers .Th e

A7-11

_i Chlorophyll,mg.m

20 30 UO addition • time,day s

Fig. 6. Chlorophyll concentrations (average samples 0.5 and 2.0 m depth) in the model ecosystems during the second experiment. presence of a few Scenedesmus sp. showed that part of the water came from fresh water sources.

After day 10 chlorophyll concentrations in the controls increased due to growth of |j-flagellates, again partly P. poucheti, and Skeletonema costatum and a maximum chlorophyll concentration was reached around day 15. In the meantime other |J-flagellates had started to increase their numbers in the controls and from day 18 chlorophyll increased to amaximu m of approximately _3 30mg. m around day40 . In the system to which 0.1 mg DCA-l had been added,th e increase of chlo­ rophyll after day 10 started some days later than in the controls.Th e spe­ cies generating this growthwer e mainly |J-flagellates and also N. longissima, reaching a maximum 3-4 days later than in the controls.Durin g the remainder of the experiment the development of chlorophyll concentrations in this en­ closure was different to that inth e controls;afte r day 30 chlorophyll lev­ els were lower than in the controls, although by that time the species com­ position was thesame .

A7-12 In the system to which 1 mg DCA.l had been added the maximum chlorophyll concentration was reached 2-3 days later than in the controls, and was much higher. The species generating this growth peak around day 18 was mainly JV. longis- sima; in addition L. danicus and a few |J-flagellates were found. After day 18 chlorophyll concentrations decreased, a new maximum was found on day 28. After day 28 chlorophyll values remained very low as compared with the con­ trols. The species forming the phytoplankton community in the last weeks of the experiment in this DCA contaminated system were large flagellates (cf. Rhodomonas spp.); different from the small |J-flagellatesdominatin g the con­ trols.

The concentration of particulate matter in the different bags (Coulter coun­ ter, size range 2.5-50 Mm) showed in general the same pattern as found with chlorophyll concentrations. Comparison of the development of chlorophyll and suspended particulate matter after addition of 1m g DCA.l revealed that the peak around day 18di d not differ inphytoplankto n biomass from that found in the controls around day 15. The amount of chlorophyll per unit particulate matter was higher than in the controls. This difference was probably caused by the different species generating this peak (|j-flagellatesan d S. costatvm in the controls, and N. longissima in the contaminated system).

The differences in the species composition between the bags were clearly shown by the particle size distributions. Fig. 7 shows as an example the particle distribution in the systems on day 40. After addition of 1 mg DCA.l larger particles were more important than in the controls.

The pattern of the primary productivity measurements again followed the pat­ tern of phytoplankton biomass. Fig. 8 shows the relative carbon assimilation at a depth of 0.5 m in the different bags. The relative carbon assimilation is lower after addition of 0.1 mg DCA.l from day 10-26, and after addition of 1mg. l from day 10-21. This was also found at a depth of 2.0 m. The ATP-concentrations in the bags during the second experiment followed a pattern roughly similar to that shownb y chlorophyll,bu t thevariation s were much lower. There was a slight difference between the ATP-concentrations in the bags with 0.1 and 1.0 mg.l DCA. In the bags containing DCA adela y of the increase beginning on day 10 of biomass as compared to the control bags was found. This also compares with the effects on the chlorophyll concentra-

A7-13 —^-cumulative ret.vol. % 20 1*0 60 80 100

control ^^ 2 5 - I* fjm diameter I 1 i - 6.L .. control E§3 Si-101 .,

0.1mg DCA.!'' M Wl-16 0 „ HÜ (6.0 -32.0 .,

1 mg DCA. f' EZ2 320-508 ,. m î May-J une 1978, day U0

Fig. 7.Distributio n of thephytoplankto n overvariou s size categories in the different bags on day 40 of the second experiment.

20 ret. carbon assimilation mgC. 4/?"'. Img.chll'l May-June 1978 • control

o n + 0.1 mg DCA.I'1 a 1.0

addition •- timenays *•

20 30 iO

Fig. 8. The relative carbon assimilation at a depth of 0.5 m in the controls and DCA contaminated model ecosystems during the second experiment. tions and on the particle counts.Th e ratio chlorophyll/ATP was also calcula­ ted and is presented in Fig. 9. In this figure large differences between the controls and the contaminated systems are apparent.

A7-1A 30 Chlorophyll/ATP May-June1978 ° control 20 • /Vi + 0.1mg DCA!' * 1.0 ,.

hhL*-.i=>'J^£ 20 30 kO -»- Urne,days

Fig. 9. Development of the chlorophyll/ATP ratio in the different model eco­ systems during the second experiment.

Fig. 10 shows the concentrations of some selected nutrients during the ex­ periment. The low initial concentrations of nitrate decreased directly after filling the bags.

Mineralization of organic matter caused the ammonia concentrations to in­ crease. The growth peak of diatoms from day 10-17 resulted in a decrease of ammonia and silicate concentrations, until silicate probably limited diatom growth, and flagellates became dominant in the bags. The delay of thephyto - plankton growth in the systems towhic h DCAwa s added, is clearly shownb ya delay in the consumption of silicate and ammonia from day 10-15. Although phosphate concentrations were very low from day 10-17 (< 0.1 (Jraol.l ), phosphate apparently does not limitphytoplankto n growth.

DEVELOPMENT OF THE ZOOPLANKTON DURING THE FIRST EXPERIMENT

Calanoid copepods dominated the zooplankton at the start of the experiment, Acartia clausi being the most important species. In addition to calanoid copepods, Euterpina acutifrons, Podon intennedius, nauplii ofbarnacle san d larvae of worms were found in very low numbers. The species A. clausi, Cen- tropages ha.ma.tus, Pseudocalanus elongatus and Temora longicornis developed from egg to adult in the enclosures.Naupli i of these species reached maximum numbers in different periods of the experiment.A t the end of the experiment T. longicornis had become the dominant species in all enclosures.Th edevel ­ opment of T. longicornis is shown in Fig. 11. The addition of DCA did not

A7-15 30 iO time, day s

Fig. 10.Concentration s of selected nutrients inth e differentmode l eco­ systems during the second experiment.

A7-16 1003 Nauplii ii.l-l

100_ " Copepodites '• n.l-l

August - September 1977 4 4 control 3 3 2 yg DCA 1~' 2 2 10 " " " 1 1 10 " " " 5 5 25 " " " limG(davs)

Fig.11 .Developmen to f Temora longicornis inth edifferen tbag sdurin gth e firstexperiment . result in significantdifference sbetwee n thesystems .Th egeneratio ntimes , averaged for allbags , found forth edifferen tcopepo dspecie swere :4 0day s for T. longicornis, 33day sfo r A. clausi, 30day sfo rC .hamatu san d4 1day s for P. elongatus. These generation times are similar to those found inth e opense a (cf.Raymont ,1976 ;Parson se tal. ,1977) .

A7-17 In Fig. 12 some parameters characterizing the total calanoid copepod commu­ nity are presented. Fig. 12a shows the development of the total biomass. After filling the bags, copepod biomass increased more than ten-fold during the first two weeks, thereafter copepod biomass decreased and remained ata -3 level of 200-300 mg dry weight.m .N o differences were found which can be ascribed to the addition of DCA. Since copepods formed the major part'of the zooplankton, secondary production canb e estimated by the production of these copepods. This secondary production is shown in Fig. 12b. Due to the high variance of the development rates on which the production estimates are based, the differences between the bags are not significant. To eliminate the influence ofbiomas s on the total production, theproductio n per unit of bio­ mass was also computed. This P/B ratio decreased from approximately .40 at the start of the experiment to .05 at the end of the experiment.

August- September 1977 m control + 2fjgDCAI'J ° • 70 » 25

time,day s

Fig. 12.Developmen t ofbiomas s (dryweight , 12a)an d production (12b)o f the calanoid copepods in the different bags during the first experiment.

A7-18 DEVELOPMENTO FTH EZOOPLANKTO NDURIN GTH ESECON DEXPERIMEN T

At the start of the experiment T. longicornis dominated the zooplankton, followed byC .hamatus ,A . clausi and Paracalanus parvus. Fig. 13show sth e development of the totalbiomas s of these copepods inth ebags .Afte raddi ­ tiono f 0.1 mgDCA. l thebiomas sdevelope da si nth econtrols .Additio no f 1 mg DCA.l resulted in a high mortality of all species during thefirs t weeks. After this period numbers remained very low,bu t did not decrease further.Th edevelopmen to fth ecopepod sappeare dt ob einhibited .

Mgy-June1978 1200 o control • + 0.1 mg DCA I" a 1.0 ., ,. „ 800

30 U0 Urns, days Fig.13 .Developmen to fbiomas so fcalanoi dcopepod s (dryweight )i nth e differentbag sdurin gth esecon dexperiment .

Fig.1 4show sth edevelopmen to f T. longicornis inth edifferen tbags . In the controls nauplii,partl yproduce d inth ebags ,develope d toadults . Thiswa s also found for theothe r speciesmentioned ,bu t T. longicornus re­ mainedth emai nspecie sthroughou tth eexperiment .Additio no f0. 1m gDCA. l resulted insmal ldifference swit h thecontrols .Jus t after addition of0. 1 mg.l numberso f T. longicornis, A. clausi andC . hamatus werelowe rtha ni n thecontrols .Als oa tth een do fth eexperimen tsmal ldifference swer eobser ­ ved,bu ti ti sno tcertai ni fthes edifference sca nb eattribute d toth eDCA . Additiono f 1m gDCA. l resulted ina hig hmortalit y of T. longicornis and also of the other copepod species (adults ofC .hamatu san dA . clausi were absentfro mthi senclosur eafte r4 weeks) . In both contaminated enclosures the relative importance of T. longicornis increased, because the mortality of the other species was higher. In the controlsa naverag eo f67 %o ftota lcopepo dbiomas sconsiste do f T. longicor-

A7-19 Nauplii 100-, n.l-1

15 20 30 35 40 45

Fig.14 .Developmen to f Temora longicornis inth edifferen tbag sdurin gth e secondexperiment . nisafte rda y5 ,afte rth eadditio no f0. 1m gDCA.l" 1 76%,an dafte radditio n of 1mg. l 85%.Thes e observations showtha tDC A also changedth erelativ e speciescompositio no fth ecopepo dcommunity .

A7-20 Fig. 15 shows the secondary production and the P/B ratios in the different enclosures in the second experiment. In the controls and with 0.1 mg DCA.l -3 -1 added, secondary production was high (>5 0 mg.m ..d )durin g a considerable period (day 8-25), resulting in a very high standing stock of copepods -3 (> 1000 mg.m ,Fig . 13).Th e development of the P/B ratios in the controls show a similar pattern to that in the first experiment in all enclosures.A s could be expected from the biomass development, secondary production was strongly inhibited by the addition of 1 mg DCA.l .However , the P/B ratio appeared to show a recovery after day25 .

100 secondary production o mg.m'l.d'l / '°, May-June1978 o control 80 \t M + Olmg DCA.l' 60 a 1.0 .. „ „

iO V 20

'A A— A—a, —a — a— û— ù Tù

05 r

OA

0.3

0.2 / «S 01 Nïl\

20 30 U0 -*- time,days

Fig. 15.Developmen t ofproductio n (15a,dr yweight )an d P/B ratios (15b)o f the calanoid copepods inth e different bags during the second ex­ periment.

During the period with high population densities and maximum production (day 10-33), the contents of the enclosures were estimated to be filtered more than once aday .

A7-21 DEVELOPMENT OF THE BACTERIA

During the first weeks of the first experiment the number of colony forming _i bacteria in the different bags increased from 0.8 x 104 to 5 x 104.ml Numbers stayed at this level throughout the experiment. Differences between the bags due to the addition ofDC A could not be detected. Fig. 16 shows the number of bacteria counted by epifluorescence microscopy during the second experiment. The large amount of detritus and the decreasing phytoplankton peak at the start of the experiment offered a suitable substrate,an d numbers of bacteria increased during the first days. During this period numbers of bacteria were lower in the contaminated enclosures. A clear dose-effect relation was not found.

bacteria, Nx105.mt~1 50

May-Jane1978 40 t> control • + 0.1mg DCAl' * 1.0 „ 30

20 \AH\ ^/H JA, 10

30 iO 20 »- time, days

Fig. 16.Developmen t of the number ofbacteri a in themode l ecosystems during the second experiment.

A7-22 DISCUSSION

FATE OF THE DCA

The course of the decrease inDC A concentration in thewaterphas e inth e bags did not indicate the occurrence ofbiodégradation . No biodégradation was observed in the laboratory tests with water from the bags. In these tests the results in sterilized controls were identical to those with the test samples, even in the plastic flasks. This lastobserva ­ tion also indicates that no primary biodégradation occurred, as then differ­ ent absorption rates should have been observed in the sterile and non-sterile flasks.

The diffusion of DCA through the plastic materials, as demonstrated in sterile laboratory systems gave results which could explain the decrease in concentration in the bag-experiments, although the fact that the concentra- -l _i tion of DCA remained constant at 0.5 mg.l when 1mg. l was added to the bag remains unexplained as the laboratory diffusion experiments indicate that a constant decrease in the concentration might be expected. Janicke and Hilge (1980) reported that in performing a "confirmatory test" with active sludge the decrease in the concentration of DCA could for the greater part be ex­ plained by absorption ofDC A toplasti c parts of the testapparatus .

The results obtained in the experiments confirm thewel l known persistency of DCA (Still and Herrett, 1976; El-Dib and Aly, 1976; Gledhill, 1975;Janick e and Hilge, 1980). In soil and pure cultures an extensive metabolism has sometimes been observed (Still and Herrett, 1976; Bordeleau and Bartha, 1972; Corke et al., 1979; Prasad, 1970). By enrichment and co-metabolism it has been possible to isolate a bacterium which is able to degradeDC A completely (You and Bartha, 1982). The slow mineralization of DCA cannot be entirely explained by the inherent recalcitrance of the compound. It has been suggested that polyme­ rization and binding reactions lead to low availability of the com­ pound (Youan d Bartha, 1982), but also that DCA is apoo r inducer for anilide oxidase (Reber et al., 1979).

Contrary towha t was tob e expected from the literature data, it could not be shown in the experiments recorded here,tha t binding ofDC A to the biomass in the water plays an important role in the elimination of DCA from the water

A7-23 column. Analysis of collected sludge did not give concentrations ofDC A sig­ nificantly higher than those in the water, although the possibility of a failed extraction can not be excluded (You and Bartha, 1982). The determina­ tion of the absorbtion of 14C-labelled DCA in the biomass in the laboratory experiments showed a relatively minor part of the total DCA absorbed to the biomass.

EFFECTS OFDC A ON THEENCLOSE D ECOSYSTEM

The results of both experiments recorded here confirm the earlier finding that plankton communities in different bags develop inth e sameway , if they are exposed to approximately the same environmental conditions (Takahashi et al., 1975; Kuiper, 1977, 1981a,b, 1982). These results allow larger differ­ ences between the systems tob e attributed to the addition ofDCA .

After addition of 2, 10 or 25 |JgDCA. l no effects could be found on the enclosed ecosystem. Addition of 0.1 mg DCA.l resulted in a slight inhibi­ tion of the phytoplankton and in lower numbers of bacteria than in the con­ trols. Addition of 1 mg.l resulted in a high mortality of the copepod populations, lower numbers of bacteria, and especially at the end of the experiment in low phytoplankton concentrations. The species composition of the phytoplankton and zooplankton was changed after addition of 0.1 and 1m g DCA.l . Clear effects were found on the chlorophyll/ATP ratio. The phyto­ plankton in the contaminated systems shows a chlorophyll a/ATP ratiomaximu m on day 17,whil e the controls show an increasing ratio after day21 .

The use of the chlorophyll/ATP ratio as an indicator of the physiological state of thephytoplankto n was reviewed byKar l (1980). Generally high ratios (>3 )ar e an indication of anutrien t limitation, although it is difficult to identify the exact nature of this limitation. In this experiment thephyto ­ plankton growth in the controls after day 21 isprobabl y nutrient limited as can also be expected from the low nutrient concentration in this period (Thomas and Dodson, 1968;Epple y and Thomas, 1969;va n Bennekom et al, 1975).

Data on the toxicity ofDC A for aquatic organisms are scarce (Adema and Vink, 1981). Janicke and Hilge (1980) summarize results of toxicity tests with bacteria (Pseudomonas putida), algae (Scenedesmus quadricauda) anda protozo a (Uronema parduczi) with a number of chlorinated anilines. The no-observed

A7-24 effect concentration (NOEC) for DCA was 3.4 mg.l for the alga, 1.6 mg.l for the protozoa, while the bacterium was much less sensitive than the other species (NOEC = 19mg. l ).McLees e et al. (1979), working with 2,6-dichloro- aniline and 3,5-dichloroaniline, found LC50 values (96hours )o f 3.6 and 2.5 mg.l , respectively, using Crangon septemspinosa. Caldwell et al. (1979) studied the effects of propanil, a herbicide from which DCA is formed by hydrolysis, on Cancer magister, and found that 0.6 mg.l resulted in a decrease of the survival of the zoeae. These values are in accordance with LC50 values of DCA found inou r laboratory with daphnids,plaic e and Chaeto- gammarus, which ranged from 0.1-1 mg DCA.l .Som e other test animals were less sensitive (Adema and Vink, 1981). The EC50 values obtained with three different algae ranged from 0.45-3.2 mg DCA.l ,th e lowestvalu e attributed to the marine algae Phaeodactylum tricornutum (Adema and Vink, 1981). How­ ever, the no-observed effect concentration (NOEC)fo r the inhibition of the reproduction of Daphnia magna was 5.6 (JgDCA. l (Adema, 1978), and Hooftman and Vink (1980) found a NOEC of 3 M8-1 with reproduction of the marine polychaete Ophrgotrocha diadema as a criterion. These no-effect levels area factor 18-1000 lower than the LC50 (21 d) for the same animals, and also lower than the lowest concentration of DCA ofwhic h effects were found in the enclosed plankton communities. The extreme effects on the reproduction in the laboratory tests with animals, could notb e found in the field test. In interpreting these results, itmus tb e kept inmin d that the extremely low NOEC values were obtained in chronic tests where the DCA-concentration was kept constant by refreshing the medium. This can lead to amuc h higher dose per organism than in a single addition situation as in these enclosure ex­ periments, especially when the substance such as DCA has a tendency to bind to dissolved organic matter present in the test system (Still and Herrett, 1976)o r to undergo polymerization reactions itself (You and Bartha, 1982). Comparing the effects obtained with marine algae in the laboratory and with phytoplankton inth e field test,th e laboratory test is less sensitive.

INTERACTIONS BETWEEN PHYTOPLANKTON AND ZOOPLANKTON

During the second week of the second experiment estimated secondary produc- -3 -1 tion in the controls (20-35 mg Cm .d ,usin g a conversion factor of 0.45 to convert dry weight of copepods into carbon, Nassogne, 1972) was of the same order of magnitude as or even exceded, the daily primary production

A7-25 during this period (estimated as twice the carbon assimilation at adept h of 0.5 m from 10-14 h, by assuming that carbon assimilation increased linearly with light intensity and that daily production at a single depth is three times the production at that depth measured from 10-14 h (Kuiper et al., 1982)). Later in the experiment secondary production declined from 50% to 4% of daily primary production, indicating that the copepods did not grow well on thephytoplankto npresen t at the end of the experiment.

During the first experiment the daily secondary production declined from around 20% of estimated daily primary production on about day 10, to 5% at the end of the experiment. The declining P/B ratios of the copepods during both experiments also point in the direction of a declining efficiency of the conversion of phytoplankton into copepod biomass. Apart from the fact that larger development stages have lowe P/B ratios, this lower efficiency is probably caused by the quality of the phytoplankton present as food.A t the end of the experiments the phytoplankton community in the controls is formed by p-flagellates. In the system to which 1 mg DCA.l was added larger flagellates dominate thephytoplankto n at the end, and inthi s enclosure P/B ratios were higher than in the controls during the last week of theexperi ­ ment.

The observed succession of the phytoplankton towards |j-flagellates in the controls is, however, probably largely caused by selective grazing of the copepods. Estimated filtration rates indicated that the contents of the bags were filtered 1-3 times per day during a considerable period. During the second experiment more phytoplankton was found in the control in which less zooplanktonwa s present.Bot h observations are an indication of the large in­ fluence of zooplankton on the development of the phytoplankton. Itwa s found that if grazing was less (after addition of 1m g DCA.l )large r cells domi­ nated the phytoplankton. This was also found in other experiments (Kuiper, 1981a,Kuipe r and Hanstveit, 1982)an d by other investigators employing large enclosures (Takahashi et al., 1975; Grice and Menzel, 1978; Steele and Gamble, 1982). Thomas and Seibert (1977) found that when copepods were grazed by ctenophores later on inthei r experiment,large r cellsbecam e rela­ tively more important again. Ryther and Sanders (1980) varied the copepod population density in an enclosed plankton experiment and demonstrated that lower zooplankton densities led to the dominance of larger phytoplankton species.

A7-26 The large number of copepods canals o have a growth stimulating influence on the phytoplankton because they mineralize organic matter (cf. Cooper, 1973, Lehman, 1980). The increasing ammonia concentrations in thewate r during the first experiment point in this direction. Smith (1978) gives an ammonia regeneration rate by small zooplankton of 0.1 (Jmol. (mg dry weight) .h During the first experiment standing stocks of copepods varied from 40 to 400 -3 -3 rag.m .Durin g the second experiment more than 1000mg. m was present. 1000 mg zooplankton per m3 would result in a remineralization of 2.4 pmol ammonia.1 .d ,usin g the value of Smith (1978). Production of ammonia (and phosphate) by the copepods may explain the development of the concentrations of these nutrients. Unfortunately no estimates are at the moment available for the relative importance ofbacteri a and zooplankton in the mineralization processes in the enclosed system.

ACKNOWLEDGEMENTS

Part of this work'was financed by the North Sea Directorate of the Ministry of Transport and Public Works and by the Commission of theEuropea nCommuni ­ ties under contract 227-77-ENV(N) in the framework of the 2nd Environmental Research Programme. Thanks are due to our colleagues G. Hoornsman, W.R. de Jong,W.C . deKoek ,J.F . de Kreuk,J.M . Marquenie,P . Roele,B . Schrieken,M . van derMeer ,H . Oldersma,M . Pullens,T . Putman,K . van der Togt,H . van het Groenewoud, H. Visscher and L. Wittebrood, who helped with the chemical, biological and statistical analysis and with the evaluation of the results. D. Pompert and S.Visse r are also thanked for their assistance. Wewoul d like to thank the authorities of the Royal Dutch Navy for their kind cooperation in supplying experimental facilities.P.B . Davis corrected theEnglish .

A7-27 REFERENCES

. Adema,D.M.M. , (1978). Daphnia magna as a test animal in acute and chronic toxicity tests. Hydrobiologia 59,125-134 .

. Adema,D.M.M . and G.J. Vink (1981). A comparative study of the toxicity of 1,1,2-trichloroethane,dieldrin , pentachlorophenol and 3,4-dichloroaniline for marine and freshwater organisms. Chemosphere 10,533-554 .

. Bennekom,A.J . van,W.W.C . Gieskes and S.B. Tijssen (1975) Eutrophication ofDutc h coastalwaters . Proc. R. Soc. Lond. B189:359-374 .

. Bordeleau,L.M . and R. Bartha (1972). Biochemical transformations of herbicide derived anilines in culture medium and in soil. Can. J. Microbiol. 18,1857-1864 .

. Caldwell,R.S. ,D.A . Buchanan,D.A . Armstrong,M.H .Mallo n and R.E. Milleman (1979). Toxicities of the herbicides 2,4-D, DEF, propanil and triflurelin to theDunganes e crab, Cancer magister. Arch. Environm. Contam. Toxicol.8 ,383-396 .

. Chow,A.Y.K . and S.D. Murphy (1975). Production of methemoglobin forming metabolite of 3,4-dichloroaniline by liver invitro . Bull. Environm. Contam. Toxicol. 13,9-13 .

. Cooper,D.C . (1973) Enhancement of net primary productivity by herbivore grazing in aquatic laboratorymicrocosms . Limnol. Oceanogr. 18,31-37 .

A7-28 Corke,CT. ,N.J . Bunce, A.L. Beaumont,an d R.L. Merrick (1979). Diazonium cations as intermediates in microbial transformation of chloroanilines to chlorinated biphenyls,az o compounds,an d triazenes. J. Agric. Food Chem. 27,644-646 .

Daley,R.J . and J.E. Hobbie (1975). Direct counts of aquatic bacteria by a modified epifluorescence technique. Limnol. Oceanogr. 20,875-882 .

Davies,J.M . and J.C. Gamble (1979). Experiments with large enclosed ecosystems. Phil. Trans.R . Soc. Lond. B286,523-544 .

Derenbach,J.B . (1969). ZurHomogenisatio n des Phytoplanktons für die Chlorophyllbestimmung. Kieler Meeresforsch. 25,166-171 .

El-Dib,M.A . and Aly,O.A . (1976). Persistence of some phenylamide-pesticides in the aquatic environment 111. Biological degradation. Water Res. 10,1055-1059 .

Eppley,R.W . and W.H. Thomas (1969). Comparison of half saturation constants for growth and nitrate uptake bymarin e phytoplankton. J. Phycol.5 ,373-379 .

Fransz,H.G . (1976). The spring development of calanoid copepod populations in the Dutch coastal waters as related toprimar y production. Proc. 10thEur . Symp.Mar .Biol. ,Ostend . vol.2 ,247-269 .

Games, L.M. and R.H. Hites, (1977). Composition, treatment efficiency and environmental significance of dye manufacturing plant effluents. Anal. Chem. 49,1433-1440 .

A7-29 Gledhill,W.E . (1975). Biodegradation of 3.4,4'-trichlorocarbanilide, TCC®, in sewage and activated sludge. Water Res.9 ,649-654 .

Grice, G.D. and D.W. Menzel (1978). Controlled ecosystem pollution experiment:Effec t ofmercur y on enclosed water columns.VII I Summary of results. Mar. Sei. Comm. 4, 23-31.

Grice,G.D . and M.R. Reeve (1982). Marine mesocosms, biological and chemical research in experimental ecosystems. Springer,Ne w York,43 0pp .

Hooftman,R.N . and G.J. Vink (1980). The determination of toxic effects of pollutants with the marine poly- chaetewor m Ophryotrocha diadema. Ecotox.Environm . Safety 4,252-262 .

Hueck-van der Plas,E.H . (ed.) (1981). Disposal of chemical waste in the marine environment. Implications of the international dumping conventions. Proceedings of the international symposium. Chemosphere 10,505-716 .

Janicke,W . and G. Hilge (1980). Messung der Bio-elimination von Chloranilinen. GWF-Wasser/Abwasser, 121,131-135 .

Karl,D.M . (1980). Cellular nucleotide measurements and applications inmicrobia l ecology. Microbiol. Rev. 44:739-796 .

Kuiper,J. , (1977). Development of North Sea coastal plankton communties in separate plastic bags under identical conditions. Mar. Biol. 44,97-107 .

A7-30 Kuiper,J . (1981a). Fate and effects of mercury in marine plankton communities in experi­ mental enclosures. Ecotox.Environm . Safety,5 ,106-134 .

Kuiper,J . (1981b). Fate and effects of cadmium in marine plankton communities in experi­ mental enclosures. Mar. Ecol. Prog. Ser. 6, 161-174.

Kuiper,J . (1982) Ecotoxicological experiments with marine plankton communities in plastic bags.In :G.D . Grice andM.R .Reev e (eds),Marin e Mesocosms, Springer,Ne w York,pp . 181-193.

Kuiper,J . and A.O.Hanstvei t (1982) Fate and effects of 4-chlorophenol (4CP) and 2,4-dichlorophenol in marine plankton communities in experimental enclosures, (submitted toAquati cToxicology) .

Kuiper,J. , H. van het Groenewoud and G.Hoornsma n (1982). Diurnal variation of some plankton parameters in an enclosed marine community. Neth. J. Sea Res. (in press).

Lehman,J.T . (1980). Release and cycling of nutrients between planctonic algae and herbi­ vores. Limnol. Oceanogr. 25,620-632 .

Menzel,D.W . and J.H. Steele (1978). The application of plastic enclosures to the study of pelagic marine biota. Rapp. P.V. Réun. Cons. int.explor .Mer . 173,7-12 .

McLeese,D.W. , V. Zitko and M.R. Peterson (1979). Structure-lethality relationships for phenols, anilines and other aro­ matic compounds in shrimp and clams. Chemosphere,8 , 53-57.

A7-31 Nassogne,A . (1972). Etudes préliminaires sur la rôle du zooplankton dans la constitution et le transfer de la matière organique au sein de la chaine alimentaire marine enMe r Ligure. Thesis, Amsterdam.

OECD (1981). OECD Guidelines for testing chemicals. OECD, Paris, Summary of consi­ derations in the report from the OECD expert group on ecotoxicology, pp. 1-10.

Oppenheimer, C.H. and CE. Zobell (1952). The growth and viability of sixty-three species of marine bacteria as influenced by hydrostatic pressure. J. Mar.Res . 11,10-18 .

Parsons,T.R. , M. Takahashi and B.Hargrav e (1977). Biological océanographieprocesses . Pergamon Press,Oxford , 332pp .

Prasad, I. (1970). Mutagenic effects of the herbicide 3,4-dichloropropioanilide and its degradationproducts . Can. J. Microbiol. 16,369-372 .

Raymont,J.E.G . (1976). Plankton and productivity of theoceans . Pergamon Press,Oxford ,66 0pp .

Reber,H. ,V . Helm and N.K. Karanth (1979). Comparative studies on the metabolism of analine and chloroanilines by Pseudomonas multivoriuis strainA n1 . Eur. J. Appl. Microbiol.Biotechnol . 7,181-189 .

Robertson,A . (1968). The continuous plankton recorder: a method for studying the biomass of calanoid copepods. Bull. Mar.Ecol . 6: 185-223.

A7-32 Ryther,J.H . and J.G. Sanders (1980). Experimental evidence of zooplankton control of the species composition and size distribution ofmarin e phytoplankton. Mar. Ecol. Prog. Ser. 3,279-283 .

Savidge, G. (1978). Variations in the progress of 14C uptake as a source of error in estimation ofprimar y production. Mar. Biol.49 ,295-301 .

Smith, S.L. (1978). Nutrient regeneration by zooplankton during a red tide off Peru with notes onbiomas s and species composition of zooplankton. Mar. Biol.49 ,125-132 .

Sonntag,N.C . and T.R. Parsons, (1979). Mixing an enclosed, 1300 m3 water column: effects on the planktonic foodweb . J. PlanktonRes . 1,85-102 .

Steele,J.H . and J.C. Gamble (1982) Predator control inenclosures . In: G.D. Grice and M.R. Reeve (eds.), Marine Mesocosms, Springer, New York,pp .227-238 .

Still,G.G . andR.A . Herrett (1976). Methylcarbamates, carbanilates and acylanilides. In: P.C. Kearney and D.D. Kaufmann (eds.): Herbicides chemistry,degra ­ dation and mode of action. Vol.2 . Marcel Dekker,Ne wYork .pp .610-664 .

Takahashi,M. , W.H. Thomas,D.L.R . Seibert,J . Beers,P . Koeller and T.R. Parsons (1975). The replication ofbiologica l events in enclosed water colums. Arch. Hydrobiol. 76,5-23 .

Thomas,W.H . and A.N.Dodson , (1968). Effects of phosphate concentration on cell division rates and yield of a tropical oceanic diatom. Biol. Bull. 134,199-208 .

A7-33 Thomas,W.H . andD.L.R . Seibert (1977). Effects of copper on the dominance and diversity of algae: controlled ecosystem pollution experiment. Bull. Mar. Sei. 27,23-33 .

TNO, (1980). Degradability, ecotoxicity and bioaccumulation. The determination of. possible effects of chemicals and wastes on the aquatic environment. Part II, 281 pp. Government Publishing Office, The Hague, The Netherlands.

You, I.S., and R. Bartha (1982). Metabolism of 3,4-dichloroaniline by Pseudomonas putida. J. Agric. Food Chem. 30,274-277 .

Viswanathan,R. , I. Scheunert,J . Kohli,W .Klei n and F. Korte (1978). Long-term studies on the fate of 3,4-dichloroaniline 14C in a plant- soil system under outdoor conditions. J. Environ. Sei. Health,B13 ,243-259 .

A7-34 FATE AND EFFECTS OF 4-CHLOROPHENOL AND 2,4-DICHLOROPHENOL INMARIN E PLANKTON COMMUNITIES INEXPERIMENTA L ENCLOSURES

1) 2) JanKuipe r and Arnbjörn0 . Hanstveit

' Laboratory forApplie d Marine Research MT-TNO,P.O . Box 57,De nHelder , The Netherlands. 21 ' Central Laboratory MT-TNO,P.O . Box 217,Delft ,Th e Netherlands

Summary

The fate and effects of 4-chlorophenol (4CP) and 2,4-dichlorophenol (DCP) added to North Sea coastal plankton communities enclosed by large plastic bags, were studied in three experiments of 4 to 6week s duration.Th e biodé­ gradation of the compounds was studied in laboratory experiments using water from the enclosed ecosystems. _l 4CP and DCP,adde d at initial concentrations of 0.1 - 1.0 mg.l ,disappeare d from thewate r in the enclosures in5 to2 3day s 4CP generally being the less persistent. Degradation rates were generally comparable to those found in laboratory tests with the same water. 4CPwa s removed bybiodégradation ,DC P probably by a combination ofbiodégradation ,photodegradatio n and/or chemical degradation. Results indicated that biodégradation rates could be limited by lack of inorganic nutrients, leading to much lower degradation rates than would be expected from routine laboratory tests. Faster degradation after repeated addition of 4CP showed adaption of the bacterial community. _l Addition of 0.3 mg 4CP or DCP.l inhibited the phytoplankton growth rate _l slightly. 1 mg 4CP or DCP.l inhibited the phytoplankton, changed the species composition and also influenced the zooplankton. In two of the three _i experiments 1 mg DCP.l resulted in a temporary lowering of bacterial num­ bers following addition. In one experiment inhibitory effects were found after 4CP and DCP had disap­ peared from thewater ,pointin g to the formation of amor e toxic intermediate

A8-1 during the degradation of these compounds. The laboratory tests also indi­ cated the formation of relatively stable intermediates.

The concentrations causing effects in the different bag experiments were quite similar. This indicates that, although the development of the plankton communities during the different experiments was different, the concentra­ tions resulting inecologica l effects are quite reproducible.

KEYWORDS : experimental ecosystem,biodégradation ,pollutio n effects, chlorinated phenols,phytoplankton , zooplankton,bacteria , North Sea.

A8-2 INTRODUCTION

A series of experiments to study the fate of model pollutants and theiref ­ fects on structure and function of Dutch coastal plankton communities in plastic enclosures has been carried out (Kuiper 1982). In the initial expe­ riments heavy metals were used as model pollutants (Kuiper 1981a,b). The general aim of these experiments has been to develop ametho d which could be used in aquatic ecotoxicology to bridge the apparent gap between conditions in laboratory toxicity tests and the field situation. The aim of later research, partly reported here, was to investigate if themetho d canb e used to study the influence and fate of organic compounds in enclosed natural plankton communities.

The following compounds were used: 3,4-dichloroaniline (DCA), 2,4-dichloro- phenol (DCP), 4-chlorophenol (4CP) and phenol. The choice of compounds was based on differences in the biodegradability of these compounds found in laboratory tests. The results with DCA, themos t recalcitrant compound,wer e reported elsewhere (Kuiper and Hanstveit 1982). 4CP and DCP were tested in three experiments. The present paper presents the results of one experiment in detail. Results of two other enclosure experiments with 4CP andDC P will be used for comparison.

Chlorinated phenols are potential pollutants in the field because of their importance as intermediates in the chemical industry (Mansour et al. 1975, Holcombe et al. 1982). They are also intermediates in the degradation of phenoxyacetic acid, some carbamates and other biocides such as pentachloro- phenol (Mansour et al. 1975, Weber and Ernst 1978, Buikema et al. 1979, Alexander 1979, Baker et al. 1980). They are found in marine waters in _l concentrations ofu p to 9 pg.l (Buikema et al. 1979).

A further objective of the experiments reported here was to compare the biodegradability of the compounds in the bags with that in simultaneous laboratory tests. Results of the laboratory tests were reported by De Kreuk and Hanstveit (1981)an d Hanstveit (1982).

A8-3 MATERIALAN D METHODS

The construction of thebag s and the operation procedures have been described in detail by Kuiper (1977, 1981a,b). The first experiment, reported here in detail, started on 19 March 1978 (called day 0) and lasted 42 days. 4CP and DCP were added in single doses on day 5 in initial concentrations of 0.1 and _l 1.0 mg.l . Two bags served as controls. Both 4CP and DCP were purified by recristallization before use. To study possible adaptation of the bacteria a second dose of 4CP in the same concentrations was added to the appropriate bags on day 32. The second experiment started on 11 August 1978 and lasted 28 days; the third experiment started on 25 August 1980 and also lasted 28 days. In the latter experiments 4CP and DCP were added in concentrations _l of 0.3 and 1.0 mg.l on day 3 (2nd experiment) or day 2 (3rd experiment). During the 3rd experiment 4CPwa s added three times to one of the enclosures.

At the start of each experiment the bags were simultaneously filled with 1.5 m3 of natural seawater collected a few miles off shore. The bags were anchored near a raft in the harbour of Den Helder, The Netherlands. During the experiment the development of thephytoplankton , zooplankton and bacteria was measured, aswa s a set ofphysico-chemica l parameters including nutrients (phosphate, ammonia, nitrate, nitrite, silicate), pH, light,temperatur e and concentrations of 4CP and DCP in water and sediment. To follow the develop­ ment of the phytoplankton chlorophyll concentrations, primary productivity, particle volume distribution and species composition were measured. Zooplank­ ton organisms were counted, and their biomass and production estimated by theprocedure s described by Kuiper and Hanstveit (1982).

Zooplankton samples of 15.7 1 were collected at least twice a week, taking mixed samples from nearly the whole water column (0-2.5 m).Othe r samples were taken daily, as a rule at 9a.m . at depths of 0.5 and 2.0 m. The methods by which the different parameters were measured are described by Kuiper (1981a,b) and Kuiper and Hanstveit (1982). The concentrations of 4CP and DCP were measured by HPLC (TNO 1980, Kuiper and Hanstveit 1982) in samples which were preserved by addition of phosphoric acid. The eluents used were methanol/H20/H3P04 (500/500/10 for 4CP; 670/330/10 for DCP).Th e laboratory degradation tests were performed with water from the different enclosures as described by Kuiper and Hanstveit (1982). General conclusions from these tests will be used here for comparison with the model ecosystem data.

A8-4 To study possible absorption and/or diffusion of 14C-labeled 4CP or DCP through the plastic foil model experiments of the type described for DCA by Kuiper and Hanstveit (1982)wer e carried out.

RESULTS

Fate of the added compounds. Figure 1 shows the concentrations of 4CP and DCP in the water during the first experiment.Fo r the initial concentration of 0.1 mg 4CP.1 degradation

/\ May-June 1978 1.0 iCP \ DCP \ « iCP V • DCP fmg.r') *

0.5

0.1 •"le*-.. -time,days •o—o-o-o-_ . • -., J 1 ,•--•-•. rzs0 s » ! X._L 10 15 20 25 30 35 iO i5

Fig. 1.Concentration s of 4CP and DCP in thewate r of thebag s during the first experiment.

_l took 16 days; a concentration of 1 mg 4CP.1 was degraded in 19 days. The second dose of 4CP on day 32 disappeared from the water in 11 days,bu t the pattern of the degradation curve (Fig. 1)wa s not the same as that of the first dose. DCP disappeared from thewate r in 8 to 23 days following roughly linear patterns. In both other experiments 4CP and DCP concentrations de­ creased ina similar way to that seen in the first experiment,a lagphase and a strong decrease in the case of 4CP, a nearly linear decrease forDCP . The rates of decrease differed between the experiments (TableI) .

A8-5 Table1 .Numbe ro fday safte rwhic h4C Po rDC Pha ddisappeare d fromth ewate r ofth eenclosure .

4CP DCP lowdos e highdos e lowdos e highdos e experiment

1,1s t dose 16 19 8 23 2nd ti 11 11 - -

2,1s t dose 5 5 10 10

3,1s t dose 8 8 15 18 2nd M - 4 - - 3rd H - 2 - -

4CPan dDC P concentrations weremeasure d inth e sediment collected insedi ­ ment traps. The concentrations found were the same as those inth ewater , indicatingtha tadsorptio nt oparticulat ematte rwa sno timportant . The diffusion experiments inth e laboratory showed thatafte r 64day s2 %o f theadde d 4CPan d 8%o fth eDC Pha d diffused throughth efoil .0.02 %o fth e added 4CP and 0.09%o f theDC Pwa sadsorbe dt oth efoil .Thes edat aindicat e that absorption and diffusion dono tpla y an important role inth eobserve d decreaseo fconcentration si nth ebags .

Developmento fth ephytoplankton . -firs texperimen t-

Inth ewee kbefor e the 1stexperimen tstarted ,chlorophyl lconcentration si n the Marsdiep, the tidal inlet from which the water to fill the bags was -3 _3 taken, had decreased from approximately 50 mg.m to 15 mg.m .Figur e 2 shows that this decrease continued in the enclosures.Fro m day5-10 ,jus t afterth eadditio no fth ecompounds ,extremel ylo wchlorophyl lconcentration s _3 weremeasure d (<0.05mg. m ).Th ephytoplankto nbloo mprecedin gthi sminimu m was generated by p-flagellates;diatom swer epresen ti nsmal lnumber s(JVitz - schia longissima, LeptocyJindrus danicus, Skeletonema costatum, Melosira sul­ cata, Rhizosolenia setigera). After day 10 chlorophyll concentrations in­ creased due to growth of [j-flagellates (partly Phaeocystis poucheti), and

A8-6 30 r chlorophyll, mg.m May-June 1978 • control o il 4 0.1 mg UCP I' » 1.0

20

10 A I**

time,day s M L 20 30 UO addition Fig.2a .Chlorophyl l concentrations inth econtrol san d4C Pcontaminate d bags duringth efirs texperiment .

30r- chlorophyll, mg.m'

I May-June 1978 • control

* 01 mg DCPI''' x 1.0 20

\

S~°^

\ 'V* ii *4" ••-• . •time,days _Li*-x If L 20 30 UO addition Fig.2b .Chlorophyl l concentrations inth econtrol san dDC Pcontaminate d bags duringth efirs texperiment .

A8-7 Skeletonema costatum. Thereafter the development in all bags was no longer the same. In the controls and thebag s which received 0.1 mg 4CP orDCP. l , maximum chlorophyll concentrations were reached around day 15, after which concentrations again decreased. Inth emeantim e other p-flagellates had star­ ted to increase in numbers, and after day 18 chlorophyll concentrations in- _3 creased to amaximu m of around 30mg. m around day40 . _i After addition of 1m g 4CP.1 deviating chlorophyll concentrations appeared after day 18 (Fig. 2a),afte r addition of 1m g DCP.l" after day 25 (Fig.2b); the increase of chlorophyll seen in the other bags was inhibited 2-3 weeks after the addition of the compounds. A maximum was found around day 35,bu t chlorophyll concentrations remained low until the end of the experiment. Large flagellates were more important in the bags to which 1 mg 4CP or DCP.l hadbee n added than inth e controls.

The concentration of particulate matter in the different bags as measured with a Coulter counter (size range 2.5-50|j mdiameter) , generally showed the same pattern as that shown by chlorophyll concentrations, although only 5 days after the addition of 1m g 4CP.1 the amount ofparticulat e matter was lower than inth e controls from day 10-17. During this period the differences in chlorophyll concentrations between this bag and the controls were small.

The differences in the species composition between the bags were shown clearly by the particle size distributions. Figure 3 shows the particle size spectra in the systems on day 40 as anexample .

»- cumulative vol.,% 20 U3 60 80 100 -1 H

^^ diameter 2.5 - 1*0 urn

I I „ CO - 64 urn ^ .. 6i - 10.1 urn E3 101 -160um H 16.0 -32.0um •H .. 32.0 - 508um

Fig. 3. Particle size distribution inth e different bags on day 40 of the first experiment.

A8-8 Carbon assimilation by the phytoplankton was measured at depths of 0.5 and 2.0 m. The patterns in time correspond in general to those shownb y chloro­ phyll and particulate matter. Depth and biomass were the main factors in­ fluencing the carbon assimilation in the systems. Carbon assimilation was higher at lower depth and higher biomass. To exclude the influence of bio­ mass, results were divided by the chlorophyll concentration; the resultant relative carbon assimilation at a depth of 0.5 m in the controls and the 4CP treated bags is shown in Figure 4.Thi s relative carbon assimilation was in- _l hibited after addition of 1rag 4CP. 1 between day 16an d24 .

20 . relative carbon assimilation rngC. 4 /)~ ' . May-June 1978 1 (mg chl)' decontrols A 0.7 rng^CP.r1 A 1.0 .,

10 -

time, days

Fig. 4.Th e relative carbon assimilation at a depth of 0.5 m in the controls and 4CP contaminated bags during the first experiment.

_l The results after addition of 1m g DCP.l were remarkable in that the rela­ tive carbon assimilation showed the same pattern as in the controls indica­ ting that no inhibition took place. The similar carbon assimilation in the controls and the systems treated with DCP did not, however, result in the same phytoplankton increase. Apparently the doubling rate of phytoplankton cells, but not the assimilation rate, was inhibited after addition of 1 mg DCP.l" (cf.Kuipe r 1981a).

A8-9 Figure 5 shows the average concentrations of various nutrients during the first experiment in all bags (only minor differences were found between the different enclosures). Concentrations ofphosphat e and nitrate decreased

Si + May-June 1978 N03o'

30 UO »- time (days)

Fig. 5. Concentrations of selected nutrients in thebag s during the first experiment.Th e average concentration over all bags for each sampling date isgiven . immediately after the start of the experiment,ammoni a concentrations increa­ sed for the first two weeks as a result ofmineralizatio n of organic matter. The growth peak of diatoms around day 15 resulted in a decrease of ammonia and silicate concentration. Concentrations of nutrients remained low for the rest of the experiment (unfortunately ammonia determinations for the last weeks of the experiment were unsuccessful).

- second experiment - _1 During the second experiment addition of 0.3 or 1.0 mg 4CP orDCP. l resul­ ted in ados e related delay of the firstphytoplankto n peak (Figure 6), which was formed mainly by Thallassionema nitzschioides and Rhizosolenia delica- tula. Addition of 4CP and DCP also resulted in lower concentrations of sus­ pended particulate matter (size range 2.5-50 |Jmdiameter) , lower carbon as­ similation rates, and a delay in the consumption of nitrate, ammonia, sili­ cate and phosphate. After the chlorophyll minimum on day 9, concentrations again increased due to growth of Prorocentmm micans, P. scutellum, Nitzschia longissima and low numbers of p-flagellates.A maximum was reached in the control on day 17. In the bag which had received a dose of 1 mg _l 4CP.1 a maximum of the same species was found on day 13, in the other

A8-10 August- September 1978 + control o 0.3mgDCP.r' • 1.0 .. .. Û 0.3mg iCP.f' * 10 „ ..

20 30 -»- time (days)

Fig. 6. Chlorophyll concentrations in the differentbag s during the second experiment.

contaminated systems the phytoplankton development was similar to that in the control. At the end of the experiment the phytoplankton consisted mainly of p-flagellates and low numbers of Prorocentrum spp.N o differences were found in the species composition between the different bags.

- third experiment -

Figure 7 shows the chlorophyll concentrations in the controls and the con­ taminated bags during the third experiment. Soon after filling chlorophyll concentrations increased rapidly, reaching a maximum on day 4. This bloom was generated mainly by Leptocylindrus danicus and L. minimus and other diatoms, although P. poucheti also increased in numbers during these first days. This phytoplankton bloom collapsed and minimum chlorophyll values were reached on day 9 inth e controls.Subsequentl y numbers of(j-flagellate s andlarge rflagellate s (Dinophysis acuminata, Prorocentrum micans, Exuviaella balthica) increased, and a second chlorophyll maximum in the controls was reached around day 18.A t the end of the experiment the phytoplankton com­ munity consisted of |J-flagellates and some dinoflagellates. _i Addition of 0.3 mg 4CP.1 on day 2 resulted in slightly lower chlorophyll concentrations than in the controls on following days, later in theexperi ­ ment chlorophyll concentrations developed similarly to those inth econtrols .

A8-11 Chlorophyll, mg.m August-September 1980 o» Controls A 0.3 mg iCP.r' 20 * 1.0 . + 0.3 . DCP.„ x 1.0 ..

10 -

20 30 -*• time (days)

Fig. 7. Chlorophyll concentrations in the different bags during the third experiment.

_1 The inhibitory effect of 1rag 4CP. 1 was more pronounced than that of 0.3 mg _i 4CP.1 . _i After the 2nd addition of 1 mg 4CP.1 on day 16 chlorophyll concentrations decreased more rapidly than in the controls. The 3rd addition on day 24 did not result in a further decrease, but the _i species composition in the enclosure treated with 1 mg 4CP.1 differed markedly from that in the controls after day 20. In thepollute d bag larger flagellates (P. micans among others) were relatively more numerous than in the controls.Coulte r counts confirmed the microscopic observations.

DCP did not influence the development of the chlorophyll concentrations directly after the addition (Fig. 7),althoug h after day 14difference s were detected. In theba g treated with 1m gDCP. l abloo m ofExuviae2! a balthica and |j-flagellates occurred, which was much higher than in the controls.Als o _l after addition of 0.3 mg DCP.l E. balthica was more important than in the controls during this phase of the experiment. As in the first experiment larger cells were relatively more important after _l addition of 1m g DCP.l compared with thecontrols .

A8-12 Developmento fth ezooplankto n

Calanoidcopepod sdominate dth ezooplankto ni nal lthre eexperiments . At the start of the first experiment Temora longicornis was themos timpor ­ tant species,whils t Centropages hamatus, Acartia clausi, Paracalanus parvus and Euterpina acutifrons were also found. Inth e controls all the calanoid species developed from nauplii, partly produced in the bags, to adults. T. longicornis remainedth edominan tspecies .

Figure 8 shows the development of the biomass of the calanoid copepodsi n thedifferen tenclosure sdurin gth efirs texperiment .Afte radditio no f

Biomass

1000

May-June 1978 S controls o0.1mgiCP.!~' * 1.0 .. 500 + 01 „ OCP . ' 1.0 .. ., .,

30 UO time (days)

Fig.8 .Developmen to fbiomas so fcalanoi dcopepod s(dr yweight )i nth edif ­ ferentbag sdurin gth efirs texperiment .

0.1 mg 4CPo rDCP. l copepods developmentwa s the samea si nth econtrols . Addition of 1m g 4CP or DCP.l obviously limited the development of the copepods.Figur e 9 shows as anexampl eth edevelopmen to fth enauplii ,cope - podites and adults of A. clausi in the controls and the 4CP contaminated _l bags.Additio n of 1m g 4CP.1 influenced all species.U p toda y1 7number s ofnaupli iwer e lower than inth e controls,afte rda y 17number so fnaupli i of A. clausi and also of T. longicornis and P. parvus increasedan dremaine d higher than in the controls throughout the rest of theexperiment .Number s _i of all stages ofC .hamatu swer ealway slowe rafte radditio no f1 m g4CP. 1 thani nth econtrols .

A8-13 100.000 nauplii, N.m -3

10.000 - /

1000 -

Hay-June 1978 o• controls 100 Û 0.3mgiCP.(-1 * 1.0 . I addition

10.000

1000

*fa

10 20 30 40 —»• time (days)

Fig.9 .Developmen to fAcarti aciaus ii nth econtrol san d4C Pcontaminate d bagsdurin gth efirs texperiment .

_i Addition of 1m gDCP. l first inhibited the increase of thecopepods ,and , later inth eexperiment ,resulte d inhighe rnumber s ofnaupli i of T. longi- cornis and P. parvus compared with the controls.Number s of copepodites of A. clausi andP . parvus also showed adifferen tpatter n intime ,compare d totha ti nth econtrols .

A8-14 Total secundary production by the copepods mentioned above,wa s estimated with the model of Fransz (1976). The results are shown inFigur e 10.Pro - _l ductioni smarkedl yreduce dafte radditio no f1 m g4C Po rDCP. l

700 May-June 1978 secondary production, o control mg.m'3 of'' * control A 01 mg i CP./" ' * 80 1 mg iCP.r' * 0.1 mg DCP x 1 mg DCP

60 y /

40

20 /\ *^ /

time, days a "

10 20 30 U0

Fig.10 .Developmen to fproductio n (dryweight )o fcalanoi dcopepod si nth e differentbag sdurin gth efirs texperiment .

The strong effects of the addition of4C P seen inth efirs texperimen twer e not seen in the second experiment. In this case the total biomass ofth e copepodswa s on theaverag eonl y 20%lowe rtha n inth econtrol safte raddi - _l tiono f 1m g 4CP.1 fromda y7-20 . Thedevelopmen to f C. hamatus appeared to be delayed. Inth eothe r systems the development of thezooplankto nwa s similart otha ti nth econtrol . _l During the third experimentadditio n of0. 3m g4C Po rDCP. l againresulte d in a similar development of the zooplanktont otha t inth e controls.Afte r addition of 1m g 4CP.1 thedevelopmen to fcopepod s (A. clausi, C. hamatus) wasdelayed ,an dlate ri nth eexperimen thighe rnumber so fnaupli iwer efoun d in this system as comparedwit h thecontrol s (Figure 11).Additio no f 1m g DCP.l resulted ina marke d inhibition ofth e copepod development asshow n inFigur e12 .Evidentl ysecondar yproductio nwa sals olimited .

A8-15 100.000 Y Nauplii V N.m-3 ^C" - 10,000

August-September 1980 ' 1000 • o controls A 0.3mg bCP.r1 A 1.0 „ _1_ 10 20 30 —>• time,days

Fig.11 .Number so fnaupli io f Acartia clausi inth econtrol san d4C Pconta ­ minatedbag sdurin gth ethir dexperiment .

Biomass, mg.m"3 1000 A

August-September 1980 o• controls + 0.3 mg DCP.r' x 1.0 ., 500

time (daysl

Fig.12 .Biomas so fcalanoi dcopepod s (dryweight )i nth econtrol san dDC P contaminatedbag sdurin gth ethir dexperiment .

Duringth ethir dexperimen tth ecladocera n Podon intermedins developed inth e bags reachingmaximu mnumber s inth e controls around day 17 (44pe rlitre) . Afteradditio no f1 m g4C Po rDCP. l numberso fP . intermedius remainedmuc h

A8-16 lower (a maximum of 8 per litre was found on day 16 after addition of DCP, after addition of 4CP a maximum of 13 P. intermedius per litre was found, also on day16) .

Development of the decomposers

Figure 13 shows the numbers ofbacteri a in the controls and the DCP contami­ nated enclosures during the first experiment. The large amount of detritus

Bacteria, Nx105.mt~'

May-June1978 a»controls + 0.1 mg DCP. I ' x 1.0 ..

30 iO time (days)

Fig. 13.Numbe r ofbacteri a in the controls and DCP contaminated bags during the first experiment. and the decreasing phytoplankton bloom at the start of the experiment formed a suitable substrate, and bacterial bioraass increased during the first week of the experiment reaching a maximum in the controls around day 5. Addition of 4CP did not result in large differences with the controls. Bacterial numbers were lower after addition of 0.1 or 1.0 mg DCP.l , although a clear dose-effect relationship isno t apparent.

During the second experiment no effects on the bacteria could be detected after the addition of the chlorinated phenols. Bacterial numbers showed the _l same pattern in time in all bags with maxima at day 6 (28 x 105.ml ) and day 24 (16x 10s.ml") .

A8-17 During the third experiment (Figure 14)number s ofbacteri a increased in the controls to amaximu m on day 9. From day 15 to the end of the experiment

bacteria,N y 10. ml'

August-September 1980 o control * H * 0.3mg DCP.r' « 1.0

>*^{=£?Ns

—*~ time,days j I 20 30

Fig. 14.Number s ofbacteri a in the controls and DCP contaminated bags during the third experiment.

.1 -l numbers were around 15 x 1.05.ml .Afte r addition of 4CP and 0.3 mg DCP.l a similar development to that in the controls was observed. After addition _l of 1m gDCP. l lowernumber s were found from day 4-10.

DISCUSSION

- Fate of the added compounds -

In agreement with laboratory experiments (Lee and Ryan 1979, Baker and May- field 1980, DiGironimo et al. 1979) 4CP generally disappeared faster than DCP from thewate r in themode l ecosystems. In the first experiment,however , _l 0.1 mg DCP.l disappeared faster than4CP .

Various processes can cause the decrease of 4CP- and DCP-concentration in the water. The decrease of 4CP concentrations in the second and third experiment, and also after day 20 in the first, shows kinetics which strongly indicate that the decrease was caused by biodégradation. The laboratory diffusion experiments showed that absorption and diffusion through the plastic foil of the enclosures was of minor importance. The initial slow decrease of 4CP during the first experiment cannot therefore be explained by diffusion.

A8-18 Adsorption on settling particles is also improbable, because concentrations of 4CP (and DCP) in the sediment were the same as those in the water.(Photo) chemical degradation can occur (Buikema et al. 1979), although biodégradation during this first phase cannot also be excluded. A possible mechanism for such slow degradation of 4CP could be the delay of rapid adaptation of the bacteria to 4CP by the high concentrations of easily degradable,natura l or­ ganic substrate in the first weeks of the experiment, favouring the develop­ ment of other bacteria. Other workers have also found that the presence of readily assimilated substrates may inhibit the formation andmaintananc e ofa microbial flora capable of decomposing themor e resistent compounds (Chou and Bohonos 1979,Atla s 1981).

The effects of adaptation of thebacteria l flora on the degradation rate was studied by repeated addition of 4CP during the first and third experiments, and by repeated addition of 4CP to samples from the bags in the laboratory tests. 4CP concentrations in the bags decreased immediately after the second additions, indicating the presence of an adapted flora.Result s of the labo­ ratory testspointe d in the same direction.

The degradation curves of DCP in the enclosures were in all three experiments roughly of a first order nature, and could be explained by photodegradation (Buikema et al. 1979, Crosby 1976) and/or chemical degradation (Baker and Mayfield 1980). In the simultaneous laboratory die-away tests, using water from the bags, degradation curves were obtained that indicated biodégradation of DCP (i.e. fast degradation after a lag phase, Hanstveit 1982). The lag phase was shorter whenwate r was taken from theba g later in the experiments. This last observation indicates thatbiodégradatio n ofDC P also played a role in the model ecosystems. A similar mechanism as given for 4CP could explain the slowbiodégradatio n ofDC P in themode l ecosystem.

Although the comparison of the results of the laboratory and semi-field ex­ periments was complicated because different analytical methods were applied (C14-method or specific chemical analysis of 4CP and DCP),th e results of both tests were generally similar.

During the first experiment, however, more than 10 days elapsed after the second addition of 4CP,befor e itha d disappeared from thewater , following a linear pattern. This was contrary to the expected fast exponential decrease, which was found in a laboratory experiment using the same water and carried out simultaneously (De Kreuk and Hanstveit 1981). A similar large difference

A8-19 between the two methods, coinciding with the same linear pattern, was found after addition of phenol in another enclosure experiment (DeKoe k and Kuiper 1981). These linear decreases in concentrations indicate that a factor, other than 4CP or phenol,limite d the growth ofbacteri a degrading these compounds. The very lowphosphat e (<0. 1 [M) or nitrate (<0. 1 |JM)concentration s during the second phase of the experiments probably limited the growth rate of the bacteria (De Kreuk and Hanstveit 1981). In the simultaneous laboratory tests during the third experiment addition of nutrients increased the degradation rates.

Limitation of biodégradation rates by low concentrations of inorganic nu­ trients has been reported by many authors (Atlas and Bartha 1972,Floodgat e 1979, Lee and Ryan 1979, Olivieri et al. 1978). In natural oligotrophic waters, like most marine waters, bacteria capable of degrading a compound, have to compete with other bacteria and thephytoplankto n for these inorganic nutrients. Goldman et al. (1979) showed that in oligotrophic oceanic waters the phytoplankton can grow at maximum rates, indicating the ability of the phytoplankton to compete with bacteria at very low nutrient concentrations. Davis et al. (1977)foun d that thebiodégradatio n rate ofvariou s oils depen­ ded mainly on the nutrient status of the seawate r sample used in theexperi ­ ment. De Kreuk and Hanstveit (1981) showed that in winter degradation rates of 4CP were higher than in summer when nutrient concentrations were low, using water samples from a station 30 km off the Dutch coast in the North

Sea. The finding that the degradation rate of a compound can be limited by4 low concentrations ofminera l nutrients is important for the extrapolation of results of laboratory biodégradation tests toth e field. Laboratory tests are often performed in the dark, so that competition between bacteria and algae for nutrients is generally absent. Moreover, nutrients are often added in unrealistic large amounts to laboratory cultures (OECD 1981). For this reason, laboratory tests could easily overestimate the degradation rate in oligotrophic environments, such as most seas and oceans during a large part of theyear .

Another interesting observation during the first experiment was the strong effect of 4CP and DCP on the phytoplankton after a fewweeks ,a t a time that these compounds had disappeared from thewater . The formation of amor e toxic intermediate during the degradation of these phenols is probable (cf. Stern 1979). The water in thebag s which had received repeated additions of 4CP was tinged yellow. The simultaneous laboratory experiments also indicated that

A8-20 intermediates, not easily degraded, were performed. De Kreuk and Hanstveit (1981) showed that the degradation of 4CP, measured as 14C loss from the die-away culture,wa s not complete,bu t stopped at about 50%degradation .

Generally, degradation via the ortho-cleavage of the aromatic ring results in complete mineralization (Evans et al. 1971, Loos 1976). The degradation of 4CP can, however, also procède via themeta-pathway ,whic h forms chlorinated catechols and 2-hydroxy-5-chloromuconicacid semi aldehyde as important end products (Knackmuss 1981,Jank e and Fritsche 1979). The yellow colour found in both the laboratory and the enclosure experiments is characteristic for the last mentioned compound. Unfortunately, no sampleswer e available at the moment that this hypothesis was suggested. Chromatography of samples from separate laboratory experiments also pointed to 2-hydroxy-5-chloromuconicacid semi aldehyde as a metabolite (Visscher,unpublishe d results from this labo­ ratory).

- Effects of the added compounds on the ecosystem -

As in former experiments plankton communities in duplicate bags developed similarly (Kuiper 1977, 1982), so that differences between the contaminated bags and the controls could be attributed to the added compounds.

Initial high mortality and inhibition of the zooplankton resulted from the addition of 1 mg 4CP.1 in the first experiment. Addition of 1 mg DCP.l also inhibited the development of the copepods, but the effects were less severe than after addition of 1m g 4CP.1 .Durin g the third experiment 1m g _i 4CP or DCP.l limited the development of copepods and also of the cladocer P. intermedius, DCP giving the stronger effects. In the second experiment, in which 4CP and DCP were most rapidly degraded, copepod development was only slightly delayed after addition of 1m g 4CP.1 _l After addition of 0.1 mg 4CP orDCP. l in the first experiment no effects on phytoplankton or zooplankton could be detected. Directly after the addition _l of 0.3 mg 4CP or DCP.l in the second experiment the phytoplankton was slightly inhibited. During the third experiment addition of 0.3 mg 4CP.1 again caused a temporary lowering of the phytoplankton biomass as compared _1 with that in the controls, after addition of 0.3 mg DCP.l no effects on _1 phytoplankton biomass could be detected.After addition of 1m g 4CP or DCP.l the phytoplankton was influenced in all experiments. In the first experiment the phytoplankton was inhibited a few weeks after addition of 1 mg 4CP or DCP.l ,a t a time that the compounds could no longer be traced in thewater .

A8-21 From available information it can be concluded that around 0.3 mg 4CP or _l DCP.l the first effects can be detected within the enclosed plankton system. In both experiments in which the development of copepods was influenced by the additions,th e species composition of thephytoplankto n differed from the controls in such a way that larger species were found in the contaminated enclosures, having lower copepod populations. Although species dependent inhibition effects cannot be excluded, it seems more probable that size selective feeding of the copepods caused this change in the phytoplankton species composition. Such food chain effects were also found inpreviou s ex­ periments (Kuiper 1981a,Kuipe r and Hanstveit 1982).

Other effects of the added compounds on interactions between trophic levels could be observed during the second and third experiments.Durin g the second experiment temporarily lower numbers of copepods were found simultaneously with higher phytoplankton densities after addition of 1 mg DCP.l (Fig. 6, around day 13).Th e higher phytoplankton biomass found after addition of 1m g DCP.l in the 3rd week of the third experiment (Fig. 7) was probably also the result of a lower grazing pressure (Fig. 11),an d not a direct growth stimulation by theDC P addition. Table II summarizes the effects of 4CP and DCP in the different experiments. In spite of large differences between the experiments (different species composition, different physico-chemical conditions (temperature, light, nutrients, etc.), it appears that concentrations causing noticeable effects in the different experiments were very similar. Themagnitud e of the measured effects differed between the experiments. It appeared that when degradation was fast, as during the second experiment, the effects were less severe, illustrating the classical rule that both dose and persistance are important factors determining the toxicity of achemical .

Buikema et al. (1979) gave an excellent review of recent literature on the toxicity of phenolics. They state that effects of 4CP and DCP onalga e have _l been found at concentrations of 10mg. l .Th e concentrations of 4CP and DCP _l affecting animals ranged from 2 - 14mg. l _l Telford (1974)foun d that 0.1 mg DCP.l increased thebloo d glucose level in three species of cray fish.McLees e et al. (1979)repor t aLC5 0 (96h )o f 4.6 _i mg 4CP.1 to Crangon septemspinosa. Kobayashi et al. (1979)foun d aLC5 0t o _i _i goldfish of 9 mg 4CP.1 and 7.8 mg DCP.l . Erickson and Freeman (1978) _l found that concentrations of 1-8 mg 4CP.1 stimulated phytoplankton growth,

A8-22 higher concentrations inhibited the algae. Holcombe et al. (1982) reported _l that 0.46 mg DCP.l reduced the survival of fathead minnows in a 28d test ; _l the 96h LC50 was 8.2 mg.l . In our laboratory LC50 (48h )t oDaphni a magna _i of 4CP and DCP were 6.5 and 5.1 mg.l , respectively. The LC50 (21days )o f the same chemicals to D. magna were 3.6 and 2.3 mg.l ,respectively . The concentrations promoting effects in the model plankton systems are com­ parable and atmos t a factor 10 lower than results reported by otherauthors .

Acknowledgements

Thanks are due to our colleagues W.R. De Jong, W.C. De Koek,J.F . De Kreuk, G. Hoornsman, J.M. Marquenie, H. Oldersma, A.H.L. Pullens, P. Roele, B. Schrieken,R.A.M . VanArdenne ,M . Van derMeer ,C.Th . Van der Togt,H . Van het Groenewoud and H. Visscher who helped with the chemical,biologica l and statistical analysis and with the evaluation of the results. D. Pompert and S. Visser are thanked for their assistence. We would like to thank the authorities of the Royal Dutch Navy for their kind cooperation in supplying experimental facilities. Part of thiswor k was financed under contract 227-77-ENV(N)b y the Commission of the European Communities in the framework of the 2nd Environmental Research Programme. Another part (laboratory tests and 1980 enclosure experi­ ment) was financed by the North Sea Directorate of the Dutch Ministry of Transport and Public Works.P.B . Davis corrected theEnglish .

A8-23 Ol Q o + + 00 1—! + + + + TJ o\ 04 C 1-4 + 15 Q l-l CO Ol i . + U » « + 4-1 4-1 U U 01 U r-l O 4->•* CU CO a 4J U X 01 O 00 01 01 4J u a 44 4-1 -C 0 4-1 4J d 01 S o Xi t» o •H 4-> u O u 4J d S Si M d 00 •r4 O o 4-) CO 4-> 14 d O M 00 O» o 04 d 14 4-1 B « d 00 M O r-4 o d d U 04 4-1 rH o n O M co r-4 CO 4*1 •H 4-1 a, 01 >> teo U U 0 •H Si i-i 01 01 4-> U 04 04 4-1 44 01 O o 44 J>S. 04 44 o 10 W 04 co O N j3

A8-24 REFERENCES

Alexander,M . (1979) Role of cometabolism. In: Proceedings of the Workshop microbial degradation of pollutants in marine environments,eds .A.W . Bourquin and P.H. Pritchard. EPA,Gul f Breeze,Florida ,pp . 67-75. Atlas,R.M . and R. Bartha (1972) Degradation and mineralization of petroleum in seawater: limitation by nitrogen and phosphorus. Biotech.Bioeng. 14:309-318 . Atlas,R . (1981) Microbial degradation of petroleum hydrocarbons: an environmental perspective. Microbial Rev. 54:180-209 . Baker,M.D. ,C.I .Mayfiel d andW.E .Innis s (1980) Degradation of chlorophenols in soil, sediment and water at low tem­ perature. WaterRes . 14:1765-1771 . Baker,M.D . and C.I.Mayfiel d (1980) Microbial and nonbiological decomposition of chlorophenols and phenol in soil. Water,Air ,Soi l Poll. 13,411-424 . Buikema,A.L. ,M.J . McGinniss anJ . Cairns (1979) Phenolics inaquati c ecosystems:a selected review of recent literature. Mar. Envir. Res.2 : 87-181. Chou,T.W . andN . Bohonos (1979) Diauxic and cometabolicphenomen a inbiodégradatio nevaluations . In: Proceedings of the Workshop microbial degradation of pollutants in marine environments, eds. A.W. Bourquin and P.H. Pritchard. EPA, Gulf Breeze,Florida ,pp . 76-87. Crosby,D.G . (1976) Herbicide photodecomposition. In: P.C. Kearney and D.D. Kaufmann (eds.): Herbicides, Chemistry, de­ gradation and mode of action (2nd edition). Marcel Dekker,Ne wYor k and Basel.Vol .2 ,pp .835-890 . Davis, S.J., CF. Gibbs and K.B. Pugh (1977) Quantitative studies onmarin e biodégradationo f oil.III . Comparison of different crude oil residues and effects of seawater source. Environ. Pollut. 13:203-215 . DiGeronimo,M.J. ,R.S .Boethlin g and M. Alexander (1979) Effect of chemical structure and concentration onmicrobia l degradation inmode l ecosystems. In: Proceedings of the Workshop microbial degradation of pollutants in marine environments,eds .A.W . Bourquin and P.H. Pritchard. EPA, Gulf Breeze,Florida ,pp .154-166 .

A8-25 Erickson, S.J. and A.E. Freeman (1978) Toxicity screening of fifteen chlorinated and brominated compounds using four species ofmarin e phytoplankton. In: Proc. 2nd Conf. Water chlorination, eds. R.L. Jolley, H. Gorchev and D.H. Hamilton, Gatlingburg, Tennessee, Ann Arbor Science Publ. pp. 307-310. Evans,W.C. , B.S.W. Smith,P . Moss andH.N . Fernley (1971) Bacterial metabolism of 4-chlorophenoxy-acetate. Biochem. J. 122,509-517 . Floodgate,G.D . (1979) Nutrient limitation. In: Proceedings of the Workshop microbial degradation of pollutants in marine environments, eds. A.W. Bourquin and P.H. Pritchard. EPA, Gulf Breeze,Florida ,pp .107-118 . Fransz,H.G . (1976) The spring development of calanoid copepod populations in the Dutch coastal waters as related toprimar y production. Proc. 10thEur .Symp.Mar.Biol . Ostend. 2:247-269 . Goldman,J.C. , J.J. McCarthy and D.G. Peavey (1979) Growth rate influence on the chemical composition of phytoplankton in oceanicwaters . Nature 279:210-215 . Hanstveit,A.O . (1982) Vergelijking van de biologische afbraak in zeewater van 4-chloorfenol en 2,4-dichloorfenol inmode l ecosystemen en laboratoriumtoetsen. Rapport MD N&E 82/1. Central Laboratory TNO, Delft, The Netherlands (in dutch). Holcombe,G.W. , G.L. Phipps and J.T. Fiandt. (1982) Effects of phenol, 2,4-dimethylphenol, 2,4-dichlorophenol, and penta- chlorophenol onembryo ,larval ,an d early juvenile fathhead minnows (Pimephales promelas). Arch. Environ. Contam. Toxicol. 11:73-78 . Janke,D . and W. Fritsche (1979) Dechlorierung von 4-Chlorphenol nach extradioler Ringspaltung durch Pseudomonas putida. Z. Allg.Microbiol . 19:139-141 . Knackmuss,H.J . (1981) Degradation ofhalogenate d and sulfonated hydrocarbons. In: T. Leisinger et al. (eds.). Microbial degradation of xenobiotics and recalcitrant compounds. Academic Press,Londo npp . 190-212 Kobayashi,K . H. Akitake and K. Manabe (1979) Relation between toxicity and accumulation of various chlorophenols in goldfish. Bull. Jap. Sei. Fish. 45:173-175 . Kock,W.Chr . de,an d J. Kuiper (1981) Possibilities formarin e pollution research at the ecosystem level. Chemosphere 10:575-603 .

A8-26 Kreuk,J.F . de,an d A.O.Hanstvei t (1981) Determination of the biodegradability of the organic fraction ofchemi ­ calwastes . Chemosphere 10:561-573 . Kuiper,J . (1977) Development of North Sea coastal plankton communities in separate plastic bags under identical conditions. Mar. Biol. 44:97-107 . Kuiper,J . (1981a) Fate and effects of mercury on marine plankton communities in experi­ mental enclosures. Ecotox.Environm . Safety 5: 106-134. Kuiper,J . (1981b) Fate and effects of cadmium on marine plankton communities in experi­ mental enclosures. Mar. Ecol. Prog. Ser. 6: 161-174. Kuiper,J . (1982) Ecotoxicological experiments withmarin e plankton communities in plastic bags. In: G.D. Grice and M.R.Reev e (eds.), Marine Mesocosms,Springe r verlag, New York.pp . 181-193. Kuiper,J . andA.O .Hanstvei t (1982) Fate and effects of 3,4-dichloroaniline (DCA) in marine plankton com­ munities in experimental enclosures. Submitted toAquati c Toxicology. Lee, R.F. and C.Rya n (1979) Microbial degradation of organochlorine compounds in estuarine waters and sediments. In: Proceedings of the Workshop microbial degradation of pollutants in marine environments, eds. A.W. Bourquin and P.H. Pritchard. EPA, Gulf Breeze,Florida ,pp .443-449 . Loos,M.A . (1976) Phenoxyalkanoicacids . In: P.C. Kearney and D.D. Kaufmann (eds.), Herbicides (2nded.) . Marcel Dekker,Ne w York.Vol . 1,pp . 1-128. Mansour,M. ,H . Perlar andF . Korte (1975) Beitrage zur ökologische Chemie. Chemosphere 4:235-240 . McLeese,D.W. , V. Zitko and M.R. Peterson (1979) Structure-lethality relationships for phenols, anilines and other aromatic compounds in shrimp and clams. Chemosphere 8: 53-57. OECD (1981) OECD Guidelines for testing of chemicals.Par t 1. OECD,Paris . Olivieri,R. , A. Robertiello and L. Degen (1978) Enhancement microbial degradation of oil pollutants using lipophylic fertilizers. Mar. Pol.Bull . 9:217-220 .

A8-27 Stern (1979) Need forbiodégradatio n studies by the EPA. In: Proc. of the workshop microbial degradation ofpollutant s inmarin e environments, eds. A.W. Bourquin and P.H. Pritchard, EPA, Gulfbreeze, Florida,pp . 11-14. Telford,M . (1974) Blood glucose in crayfish. II.Variation s induced by artificial stress. Comp.Biochem . Physiol.48A :555-560 . TNO (1980) Degradability, ecotoxicology and bioaccumulation. The determination of possible effects of chemicals and wastes on the aquatic environment, partII . Government publishing office,Th e Hague,Th e Netherlands. 281pp . Weber,K . and W. Ernst (1978) Levels and pattern of chlorophenols in water of the Weser estuary and the GermanBight . Chemosphere 11:873-879 .

A8-28 Publication P81/5 3 1981-12-01

INFLUENCES OF BAGDIMENSION S ON THE DEVELOPMENT OF ENCLOSED PLANKTON COMMUNITIES DURING POSER

JanKuiper * Uwe Brockmann** Henkva nhe t Groenewoud* Gerard Hoornsman* KlausHammer* *

Laboratory forApplie d Marine Research,Divisio n of Technology for Society TNO,P.O . Box 57, 1780A B Den Helder,Th eNetherlands .

Institut für organische u. Biochemie der Universität Hamburg, MartinLuthe r King Platz 6, 2000Hambur g 13,FRG .

A9-1 ABSTRACT

During POSER (=Plankto n Observations in Simultaneous Enclosures inRos - fjorden)natura l plankton communities were studied, enclosed by large plastic bags,whic h were anchored in situ.Enclosure s of different dimensions were 3 used,rangin g indept h from 3 to 40m , and containing 1.5-30m ofwater . Thus itwa s possible to study the effects of the dimensions of the enclosure on the development of theplankto n inside.

The development of the phytoplankton in large enclosures was very similar to that in small enclosures during POSER, ifw e compare themea n values in small bags with those of the total mixed layer of the large bags. The large and the small enclosures had been filled simultaneously. Numbers ofbacteri a in small bags increased faster than those in the large enclosures,probabl y because of closer contact with the sediment.Copepo d populations suffered high mortality in particular in the small bags. This mortality may havebee n influenced by the extremely cold weather prevailing during the experiment.

It appeared that fluctuations innatura l factors (temperature,light , nutrients, etc.)wer emuc h more important than the dimensions of theen ­ closure for theplankto n development.

The application of enclosures to study the effects of pollutants on natural plankton ecosystems isdiscussed , and it is concluded that optimum dimen­ sions depend on the aim of the experiment ,th enumbe r of trophic levels enclosed, the population density on these levels and on the species present on these levels.Thi smean s that,i fmarin e phyto and zooplankton interactions 3 are studied in eutrophicwaters ,enclosure s of 1-2 m are sufficiently large, and for practical reasons should be preferred over larger ones. Inmor e 3 oligotrophicwaters ,enclosure s with avolum e of approximately 10m are to be preferred, so that larger zooplankton samples canb e taken.

A9-2 1. INTRODUCTION

To circumvent themethodologica l difficulties of usual experimental field work'an dsom e limitations of laboratory investigations,mode l ecosystems were used in experimental aquatic ecology to study aspects of ecosystem dynamics under semi-natural conditions (Vervelde and Ringelberg 1977, Menzel and Gase 1977,Menze l and Steele 1978).

The aim ofmode l ecosystem experiments was either of a rather fundamental character i.e. to test hypotheses or to study mechanisms which act in the system (Strickland and Terhune 1962,Parson s et al. 1977,Brockman n etal . 1977, 1979,Parson s 1978, Sonntag and Parsons 1979,Gric e et al. 1980), or related tomor e practical applications.I n the latter case the aims differed as tomeasur e and to explain the response of the ecosystem to stress by xenobiotic pollutants (Menzel and Case 1977,Reev e et al. 1976,Kuipe r 1977b, 1981, Davies and Gamble 1979,Gric e andMenze l 1978,Marshal l and Mellinger 1980), perturbation by raise in temperature (Donze 1978)o r eutrophication (Schelske and Stoermer 1972), to study the fate of chemicals in the eco­ system (Topping and Windom 1977,Wad e and Quinn 1980,Le e et al. 1978)o r to validate predictions based on theory or laboratory tests (Hueck et al. 1978).

In themarin e environment natural plankton communities enclosed by large, flexible, translucent plastic bags have frequently been used as physical models for thenatura l ecosystem (for reviews seeKinn e 1976,Reev e et al. 1976, Menzel and Steele 1978,Davie s and Gamble 1979,1981 ,Kuipe r 1981). Important aspects of these experiments are the replicability of experimental units in an experiment and the extent towhic h themode l is representative for the real system.

Unfortunately, replicability and representativity seem tob e inversely inter­ related properties of experimental design (Gamble and Davies 1981). Amono - species algal culture in the laboratory isno t representative for most mechanisms acting innatur e (e.g. competition, grazing, etc.),bu t the possibilities for replication arever y large.O n the other hand it can be stated that the ocean is the only representative for itself, though it is not replicable.

A9-3 Two factors are ofprim e significance for experimental design: the duration of the experiment and thedimension s of the enclosure,bot h factors being related via the generation time of enclosed organisms and the rate of fouling of thewalls .Optima l duration of (ecotoxicological) experiments was dis­ cussed by Grice et al. 1980,Kuipe r 1981,an d others.Concernin g the size, and excluding very short-term experiments (1-2 days), enclosures of widely 3 different dimensions have been used (0.3- 16.000m )t o study natural plankton. In a field of applied research, such as ecotoxicology, enclosures as small as possible are preferred in terms of convenience of experimental handling, of possibilities of replication,an d costs (Davies and Gamble 1979).

Experiments with marine plankton communities in large (Takahashi et al. 1975, 3 3 68m )an d relatively small enclosures (Brockmann et al. 1977,3- 4 m ; 3 Kuiper 1977a, 1;5 m )showe d that the development of bacteria, phytoplankton and zooplankton replicated sufficiently for periods up to 4-8 weeks so as todetec t effects of pollutant by comparisonwit h non-polluted controls. Possibilities for replication decrease,however ,wit h increasing enclosure size (Gamble and Davies 1981). It isno t clear which aspects of the represen- tativity are lost,whe n smaller enclosures are used in comparison with large ones.

One of the aims of POSERwa s to intercompare the development of plankton ecosystems in enclosures of different dimensions.Th e depth of the enclosures used varied from 3 to 40m , thus they included only apar t of ormor e than 3 the euphotic zone.Th e contents of the enclosures varied from 1.5 to 30m . So as to study possible differences in the reaction topollutan t stress of the communities enclosed, in two experiments mercury (II)chlorid ewa s added to some of thebags . Effects of these mercury additions are reported separately (Kuiper et al. 1981). No relevant differences were detected between the effects of mercury on plankton communities in enclosures of different dimensions. In this paper a comparison will be presented between thedevelopmen t of the plankton in unpolluted large and that in unpolluted small enclosures.

A9-4 2.MATERIAI S AND METHODS

The general set-up of POSER was given by Brockmann et al. (1981a). Two experiments were performed inwhic h relatively small bags (depth 3m , 3 3 volume 1.5 m )an d large ones (depth 40m ,volum e 30m )wer e filled simultaneously. On 6Marc h 1979 the first experiment was started (POSERI) . Six smallbag s were simultaneously filled with aVanto nFlex-i-line r pump (Kuiper 1981b), the inflow hose of the pump being displaced from a depth of 20m to0. 5 m during pumping so as to correct for zooplanktonpatchiness . At the same day four large bags were filled, themetho d being used asde ­ scribed by Brockmann et al. (1981b). Mercury was added to some of the bags on 8March . In thebag s and the fjord the developments of the phytoplankton, zooplankton and thebacteri a weremonitored , aswel l as a set of physico-chemical para­ meters influencing thedevelopmen t of thebiot a (nutrients,pH , temperature, salinity, incident light). Sampling methods and analytical methods werede ­ scribed by Brockmann et al. (1981b)an d Kuiper et al. (1981). A detailed account of physical,chemica l and biological variations in the fjord was givenb y Brockmann et al. (1980).

During the first week of the experiment weather conditions were very bad and on 12-13Marc h strong currents were present in the fjord. These currents caused a total exchange ofwate rmasse s in theBtfrtf yBigh t (Brockmann et al. 1980). As a result,direc t comparisons between the plankton development in thebag s and those in the fjord (a second aim)wer e difficult and theexperi ­ ment was stopped.

A second experiment was started on 16Marc h (POSER 2). This time 7 small bags 3 3 (1.5m )an d two large bags (length 20m , volume 15m )wer e filled. Some bags received either additional nutrients ormercury , orboth ; two small sys­ tems and one large onewer emonitore d without any additions (cf.Kuipe r etal . 1981). The same parameters were measured as during the first experiment. During POSER 34 enclosures were launched. Results of some of themwil l be used here for comparisonwit h results obtained from the experiments described. Details about the various enclosure experiments are summarized inTabl e I.

A9-5 Table 1 Starting days, dimensions and duration of various enclosure experiments during POSER referred to in thispaper .

bag dimensions starting experiment code depth contents day duration (m) (m3) (days)

POSER1 11 3 1.5 6 March 8 12 3 1.5 u 8 18 40 30 M 8 19 40 30 M 8

POSER 2 21 3 1.5 16Marc h 20 22 3 1.5 u 20 24') 3 1.5 it 20 25') 3 1.5 ti 20 28 20 15 n 10

Norwegian C 15 11 6 March 29 control

German U 40 30 3 March 12 controls V 40 30 4 " 9 NC 10 7.5 8 " 27 B6 40 30 19 " 18 CC 40 30 21 " 16 DD 40 30 22 " 10

1) All enclosures in thisTabl e refer tonatura l planktonpopu ­ lations asharveste d during the experimental period. Additional nutrientswer e added tobag s 24an d 25,whic h was not so in the other experiments cited.

A9-6 3. RESULTS

PHYTOPLANKTON AND NUTRIENTS

During POSER 1phytoplankto n concentrations in all bags decreased in the same way as in the fjord; thiswa s probably due to the settling of cells. Concentrations ofreactiv e silicate, in two small and two large bags (U and V), are shown inFig . 1.Thes e concentrations remained constant during this period. In the large enclosures, the stratificationremaine d undisturbed; the small bags were not stratified. In the different bags the other nutrients showed the same patternwit h time.Th e phytoplanktonwa s probably not active and no important differences could be detected between thedifferen t enclosures.

123 1U3

depth. m 3 bag U ::i::;:::z::i:: !rm<3.0 S 30-1.0 ZZZ'UO [jmot StI 3S 'WMWtZMMs

bag V HID*3.0 5 3.0-iO tZ2">0 fjrrtot Si I

depth, m

0 1 bag 12 2 : ^/^ 3 -I ^ 73 93 11.3 ZZZ3 2.9-30 fimol Si I''

Fig. 1 Development of concentrations of reactive silicate in two large and two small enclosures during the first part of POSER.

A9-7 The development of average chlorophyll concentrations during POSER 2 in two small enclosures (Nos.2 1 and 22)an d in one large enclosure (No. 28) is shown inFig . 2.Fo r the large bag theaverag e concentration ispresente d in the

POSER 2 „ large control 28 10- 10m) , 10- 3m) chlorophyll,mg m • a small .. 21 » „ 22 ° CC 10-10m) • CC 10- 3 m)

30 -\\

\ —.^ / IS 20 ^- time, days

Fig. 2 Development of chlorophyll concentrations in enclosures during POSER 2.

upper 10m ; in this layer thebul k of primary production occured (Jahnke pers. comm.). After the bags had been filled, chlorophyll concentrations decreased owing to the settling of phytoplankton cells. After the third day in the small and the fourth day in the large bag, chlorophyll concentrations increased owing to growth of diatoms,o fwhic h Thallassiosira nordenskioldiiwa s the dominant species.Chaetocero s debilis and Ch.boreali s also increased in num­ bers in this period. In all bags maximawer e reached around the 10th-llth day (27March) , afterwhic h chlorophyll concentrations decreased again.Th e aver­ age concentrations inal l bags were about the same .

InFig . 2 the average chlorophyll concentration inba g CC is also indicated. Bag CCwa s filled some days later than the others.Fo r this enclosure the mean values of theuppe r depths corresponding to the smallbag s are also presented so as to compare thedevelopmen t at different depths.I n the large bags the chlorophyll concentrations near the surface showed a development differing very much from that in the smallbag s or from the average of the upper 10m of the large bags.

A9-8 Inba g BB amaximu m could also been found on the 10th day; thismaximu m was generated by the same diatoms as in the other enclosures.

Two small bags, also filled on 16March ,wer e spiked with nutrients (Nos.24 and 25,8 uga t N0..1 , 1.5 ugat P0..1 and 5pga t Si.l" ,se eKuipe r etal . 1981 for details). Fig. 3 shows the development of chlorophyll concentrations in these bags and inon e large bag (C),fille d with nutrient-rich water on 6 March. In thenutrient-spike d bags abloo m occured of the same diatom species as in the other bags,wit h amaximu m inbot h bags around the 11thday . Itwa s remarkable that abloo m of the same specieswa s found inba g C,althoug h itwa s filled much earlier with different water.

„

3 POSER2 chto rophylt,mg m o small bag (2U) 125) [• o targe bag C fO-Wmj ifi •\ A- \ \0 / \ / \ 41 °\ 1 time,days v^; v rr. i 5 10 IS 20

Fig. 3 Chlorophyll concentrations in two nutrient-spiked small bags and one nutrient-rich large enclosure during POSER 2.

Fig. 4 shows the development of silicate and phosphate concentrations in two small (Nos.2 1 and 22)an d two large bags (No. 28 andCC) . The growth of diatoms caused adecreas e of silicate concentrations.Thi s de­ crease was faster in the small bags than in the large bags,bu t itwa s never observed inba g CC neither at its surface norwithi n theuppe r 10m . The phosphate concentrations inba g CCwith which the experiment had been started later,were lower at the surface and increased during the experiment. Inparticula r in the small bags, phosphate concentrations showed stronger fluctuation thanwithi n bagCC .

A9-9 Si.umotP \

POSER2 o large control 28 I O-10 m I OSO 28 10- 3ml 21 21

»-»•- Vf*»» \/\ « .. (0- 3m) y o

O.SO

Vô< -rime ,da/ s

F-ig. 4 Reactive silicate and phosphate concentrations in two small and two large enclosures during POSER 2.

ret. carbon assimilation,mg C (mg chO.' lihrsl POSER 2 70 awithout nutrients depth0.5 m • „ „ „ 20m & with nutrients „ OS m A .. .. „ 2 0m

so

i.o

- time, days

0. 16 March

Fig. 5 Primary production per mgof chlorophyll in the small bags, average of two bags with and without nutrients added during POSER 2.

A9-10 Fig. 5 shows the primary production perm g chlorophyll in the small bags on different depths during POSER 2. Inhibition of primary productivity by high light intensities near the surface,whic h could be adisadvantag e of small bags,wa s not found. Rates of primary production inbag s 21 and 22 are clearly limited by lownutrien t concentrations.

ZOOPLANKTON

During POSER 1th e initial conditions in the small and large bags were very similar, calanoid copepods being themos t important.Durin g the experiment, numbers of copepods seemed to remain constant in the unpolluted bags, but the experiment was too short to allow conclusions on possible differences between thebags .

At the start of the second experiment the zooplankton community resembled that of the preceeding period. Nauplii of Calanus finmarchicus were themos t important innumbers .Apar t fromC . finmarchicus,Acarti a clausi, Centropages hamatus,Pseudocälariu selongatus ,Oithon a similis and an unidentified harpacticoid copepod were found.Naupli iwer e most abundant. Fig. 6 shows the total number of copepods in two small (21 and 22)an d one large (28)ba g during the experiment.

15.000 number of copepods per m^

5.000

0

16March ' rime, days

Fig. S Total number of copepods (all stages) in two small and one large enclosure during POSER 2.

A9-11 Table 2 shows the species composition and thenumbe r of organisms in the same bags on selected days of the experiment. In all bags thenumber s of copepods decreased; in the large bag total numbers decreased with approximately 20% perweek , themortalit y in the smallbag swa s even larger (> 50%pe r week). Table 2 shows that in the large enclosures,declinin g numbers were primärly caused by mortality of nauplii of C. finmarchicus. In the small bags all species suffered highmortality , although the numbers of 0. similis were too low toallo w conclusions.

Table 2 Numbers of copepods in large and small bags on selected days during POSER 2.Th e bagswer e filled on 16Marc h 1979. (Numbers per m3, n =nauplii ,c = copepodites,a = adults).

bag large (28) small (21 and 22)

^^day 1 12 1 12 19 species ^^-^^

T. longicornis n 400 250 663/ 612 153/255 -/- c 10 50 102/ 51 -/102 -/- a 30 10 -/- -/- -/-

A. clausi n 1050 800 816/ 510 561/459 153/153 c 50 - 204/ 153 -/- -/- - a 10 - -/- -/- -/-

C. hamatus n 650 400 1224/1122 204/255 153/ 51 c 110 240 204/ 51 51/102 -/ 51 a - - -/- -/- -/-

C. finmarchicus n 9700 5450 5714/4847 612/357 102/ 51 c 140 790 -/ 51 51/- -/- a - - -/- -/- -/-

0. similis n 150 - -/ 51 -/ 51 -/- c 100 50 51/ 102 -/- 51/- a - 10 -/- -/- -/- others 10 40 205/205 51/- -/ 51

total 12410 8130 9233/7805 1683/1683 765/357

A9-12 In the fjord,number s remained more or less constant. Development of the copepod nauplii into larger stages could notb e detected in the fjord because thewate r masses exchanged during the experimental period. In the large en­ closure No. 28number s of (small)copepodite s of C. finmarchicus increased 3 gradually from 140 to 790m during 12days . Apparently the larger nauplii did develop very slowly into small copepodites.

BACTERIA

During POSER 1tota l numbers of bacteria (epifluorescence counts) increased in the small bags from 2.5 to 5x 10 .ml in six days (seeKuipe r et al.1981) . Too few samples are available from the simultaneously filled large bags to allow any conclusions. Inba g U,however ,number s of colony forming units (CFU) increased from 600 to 2000.ml during the firstwee k of enclosures; after thiswee k numbers decreased again.A similar increase in thenumbe r of CFUwa s observed inba gV , filled in the same period (Hentzschel pers.comm.). Inba g C the numbers of CFU increased by a factor 10durin g the first ten days of the experiment, the increase in the total numbers as estimated with the epifluorescence technique being somewhat smaller. It seems that during the first period of POSER bacteria develop rather similarly in the different enclosures. Fig. 7 shows the number of bacteria in small and large enclosures (Nos.21 , 22 and 28)durin g POSER 2. In the small bags thenumbe r of bacteria increased, but in the large bag,number s were more or less constant. In other large bags available for comparison (BB and CC),number s of CFU showed a slight increase with time,tw o clear maxima occuring around 24an d 31 March inbot h enclosures (Hentzschel pers. comm.). Comparison with the epifluorescence counts is difficult,however ,becaus e achangin g proportion of the total number found may be colony forming.

Fig. 7 numberof bacteria, /' Number of bacteria in two small bags and one large enclosure during POSER 2.

POSER2 AW large control 28 small control 21 .. 22

• time.days

A9-13 4.DISCUSSIO N AND CONCLUSIONS

Menzel and Steele (1978) argued that the optimum size of an enclosure depends on thenumbe r of trophic levels.The y stated that "the most likely acceptable 3 volume to avoid adverse complications (like fouling) is 30m ".Th e choice of this volume was not further argumented. If carnivorous zooplankton is 3 . included it is assumed that aminimu m of 100m is necessary. Case (1978) indicated that the objective of the experiment is also an impor­ tant factor indimensionin g an enclosure. Wewil l discuss the development of theplankto n community on thevariou s trophic levels in relation toen ­ closure size.

PHYT0PLANKT0N

In the experiments reported here no principal differences were found between the development of thephytoplankto n in large and small bags when comparing themea n values of small bags and those of the totalmixe d layer of the large bags. This similarity was probably caused by the remineralization in the small bagswher e thebiomas s could not leave the region near the surface and by diffusion and remineralization from the lower depths in the large enclosures. The differences of phytoplankton development in the corresponding depths (Fig. 2)ar e caused by sedimentation of elements bound toparticle s out of the surface regioni nth edee p enclosures (Fig.4) .Thi s interpretation is also supported by thenutrien t measurements (see below). During POSER 2a n increase indiatom s occurred in all bags, and the species composition of these blooms was very similar. In the fjord and in thebag s to which nutrients had been added, the same species were found. The species composition was expected tob e one of the first parameters which would be influenced by a stress exerted by containment in abag .

Another indication for the absence of a dominant bag influence was that thediato mbloo m around 27Marc h occurred innearl y allbag s filled with natural plankton, irrespective of the day of filling.Eve n inba g C, filled on 6March ,a diatom bloom of Th.nordenskioldii ,Ch .boreali s and Ch. debilis could be observed around 27Marc h (3week s after filling). These findings indicate that natural factors like light,nutrients , temperature are regulating the development of the enclosed phytoplankton. The development inbag s ranging 3 from 1.5-30m appears tob e quitenatural . The factors initiating the bloom

A9-14 probably can be found in the raising water temperature and the increasing light conditions (see Brockmann et al. 1981c this volume).

In the experiments reported here,surfac e inhibition of primary production was not found. The primary productivity on depths of 0, 0.5, 2an d 3m was very similar. Jahnke measured primary production in the fjord at depths of 0, 3, 10an d 20m . He neither observed surface inhibition;h e found an average relative carson assimilation of 2.37 mgC.(mg chl) .( 6 hrs) (N= 32,s. d = 1.33) at depths of 0 and 3m . Surface inhibition could limit application of enclosures having low depths,althoug h in these cases a perspex cover, absorbing most of the light <30 0 nm,woul d largely prevent this inhibition (Brockmann et al. 1974).

During the first part of POSER the phytoplankton was probably inactive,an d nutrient concentrations remained constant in small and large bags. During POSER 2nutrien t concentrations in the large bags increased or at least decreased more slowly than in the small bags (Fig. 4).Thi s can firstly be attributed to lower rates of nutrient comsumptionb y the phytoplankton, owing to the inhibition of primary production at greater depths by lack of light and secondly tonutrien t diffusion coming up to themixe d layer from nutrient-rich layer below 10m depth (Kattner et al. 1981). This latter assumption is supported by the strong increase of silicate compared to phosphate in enclosure CC caused by the late phase dominating dinoflagel- lates (Jahnke et al.1981 )utilizin g phosphate but no silicate. In experiments with the explicit aim to study the light dependence of several processes,dee p enclosures are necessary, but for ecotoxicological experiments this difference in the rate of nutrient consumption does not seem tob e relevant.Th e effect of this difference on the phytoplankton could be thatmaxim a ofbloom s are reached earlier in small enclosures than in the large ones.

ZOOPLANKTON

Calanoid copepods formed themos t important part of the zooplankton during POSER. C. finmarchicus dominated the community, a thing which isver y usual in thenorther n North Sea and thewater s of theNorwegia n current (Furnes 1976, Mathews 1978). Nauplii were most abundant,whic h shows that the spring increase had started, since C. finmarchicus hibernates in the copepodite stage V in thesewater s (Mathews 1978).

A9-15 During POSER numbers of nauplii did not increase and, if at all,the y developed very slowly into larger nauplii.Apparentl y the extraordinary low temperatures (Kolsted 1973) limited the further development of copepods, although appropriate food was available. Colebrook (1979)foun d a similar situation of anorma l phytoplankton spring bloom and adelaye d development of copepods in thever y cold waters off Greenland and the Grand Banks region in the northernAtlantic .

During the short POSER 1experiment ,number s of copepods seemed to remain constant in thebag s during the first six days. In thebag s and the fjord the same species and stages were found. During the second experiment numbers of copepods declined in all bags. In the large bag the rate of decrease (20% per week) seems tob e toohig h as regards for natural conditions, since predatorswer e not observed. Themortalit y in the smallbag swa s certainly related to the size of the enclosure.

In the small bags, all species decreased innumbers . This isremarkable , since in experiments with enclosed plankton from Dutch coastal waters (Kuiper 1977a,b , 1981b) or fromHeligolan d (Brockmann et al. 1977), partly the same species were present as in the B«Sr«JyBigh t and in those experiments they developed at arat e comparable to that in the open North Sea (Kuiper 1977b). This indicates that thehig h mortality observed during POSER is not a general property of therelativel y smallvolume ,an d the reasons must probably be found in the extreme temperatures of the upper water layers (-1 -2 C).Kraus e (pers.comm. )foun d maximum densities of C. finmarchicus and P. elongatus in the B«Srj5yBigh t at depths of 40m . The figures of Grice et al. (1977)o n thepatch y distribution of zooplankton in large bags,wit h nearly all thenauplii near thebotto m ,als o point to the possibility that the organisms wanted to swim downwards. In the small bags this downward migration may be stimulated by thewate r in the upper layers thatwa srelativel y less shaded and colder than that in the deeper, large 3 bags. Inmuc h largerbag s as used in theCEPE X programme (100-1300m ) sometimes unexplained decreases innumber s of copepods were also found (Davies and Gamble 1979,Reev e and Walter 1976,Parson s et al. 1978).

Another drawback of the small bags in BtfrtfyBigh t wasrelate d to the low density of the zooplankton. Removal by sampling of more than 5% of the zoo- plankton per week had tob e avoided, and therefore thenumber s of organisms in the samples from the small bags were low.Lo w numbers increase the error of

A9-16 themeasurements , and therefore make discrimination between treatments more difficult.

BACTERIA

During POSER 2 the numbers of bacteria increased more strongly in the small bags than in the larger ones. This seemed also the case during the first experiment.Thi s may be caused by the fact that in the smallbag s dead phytoplankton does not settle toa great depth,bu t remains in close con­ tact with theuppe r water layer.Therefore ,i n the small bags the bacteria in thewate r may react more directly to changes in the phytoplankton than in the large bags. This immediate reaction of bacteria in thewate r column may speed up the regeneration of nutrients,whic h are directly available for phytoplankton growth. In experiments weremos t nutrients are regen­ erated by bacteria therat e of succession of phytoplankton may therefore be faster in small enclosures than in a larger one. In ecotoxicological experiments this does not seema disadvantage.

ADVANTAGES AND LIMITATIONS OF DIFFERENT SIZES

Kuiper et al. (1981)discusse d the effects ofmercur y added to enclosures during POSER. They showed that the results were very similar in large and small bags.Eve n smaller enclosures were used to study the effects of pollutants on natural plankton;Lacaz e (1974)an d Horstmann (1972)obtaine d useful results,whe n they studied phytoplankton reactions after perturbation in 600 1an d 300 1enclosures , respectively.

Marshall and Mellinger (1980) studied the effects of cadmium on thedevelop - 3 ment of freshwater plankton enclosed in 8 1an d 150m over threeweeks ,in ­ cluding zooplankton inbot h cases.Th e stated that the effects of enclosing onLak eMichiga n plankton after threeweek s (atoptima l depths)wer e rela­ tively small. Barica et al. (1980), alsoworkin g in a freshwater system,use d enclosures 3 of 9an d 230m .The y showed that in enclosures of both sizes for several months the development of the phytoplanktonwa s similar to that in the lake.

A9-17 Invariou s experiments the development of enclosed marine communities was compared to that in the external environment to study the extent of represen­ tation of the experimental design (Davies et al. 1975,Takahashi etal .1975 , Gamble et al. 1977). Available information suggests that the development of a plankton community inside aplasti c bag isa t least qualitatively similar to that of the free community during periods up to four weeks (Menzel and Steele 1978). It is also clear that important differences exist between a free and an enclosed community (in the latter: less turbulence,n o advection of new species,onl y small scale patchiness) which must have an important influence on a longer time scale.Eve n in thever y large enclosures 3 employed by Lund (1972, 16.000m )th eplankto n community diverged from the free community after severalmonths .

As regards to the phytoplankton Menzel (1977)reporte d that similar patterns 3 of successionwer e obtained inA 1 flasks and 68m enclosures, suggesting that small scale laboratory experiments with natural phytoplankton assemblages canb e usedwit h some confidence topredic t events at the phytoplankton level inmuc h larger systems.Experiment s with continuous cultures when natural assemblages of phytoplanktonwer e used pointed in the same direction (DeNoyelles et al. 1980). An essential advantage of larger enclosures, i.e. thepossibilit y of studying the interactions between trophic levels,dis ­ appears,however ,whe n thesemuc h smaller enclosures are employed. Again it depends on the aim of the experiment which experimental set-up can be chosen.

Apart from drawbacks,suc h as limited possibilities for sampling sparse populations (carnivorous zooplankton), limited extent of representation for 3 deep, stratified waters,etc . smallbag s (1.5m )als o present advantages over large enclosures.Whe n we use very large enclosures it is very difficult to start with identical plankton communities (Case 1978,Parson s et al. 1977, Takahashi et al. 1977), which is essential if the development invariou s bags has tob e compared.

Further the plankton distribution inver y large enclosures isver y patchy (Takahashi et al. 1975, Grice et al. 1977,Davie s and Gamble 1979). This patchy distribution makes the interpretation of data difficult,unles s very many samples are taken.

A9-18 The most serious drawback of using bags inplankto n research is themuc h lower turbulence and consequent lower mixing inside aba g as compared with the turbulence innatura l waters (Verduin 1969,Takahash i and Whitney 1977, Steele et al. 1977,Menze l and Steele 1978,Davie s and Gamble 1979). This lower turbulence influences,fo r instance, the sinking rates of phytoplankton and therefore influences thedevelopmen t of the total system (Steele and Henderson 1976). One consequence of the lowmixin g rate is that inver y large bags addition of nutrients isnecessar y tokee p primary production going. Nutrients which aremineralize d in the sediment of thebag s do not return to -theuppe r layers,wher e most of theprimar y production occurs.

Artificial addition of nutrients is adrawbac k since it renders the system more unnatural and since effects of added chemicals on the mineralization process cannot be studied on the enclosed system as awhole .

Sometimes the contents of aba g are artificially mixed (Parsons et al. 1977a, Strickland and Terhune 1961,Brockman n et al.1974 ,Sonnta g and Parsons 1979), but themixin g of very large containers appears tob e a problem (Sonntag and Parsons 1979). In small enclosures mixing is easier and partly the result of sampling activities.

Another drawback of large and deep enclosures is that they aremuc h more vulnerable to changes in the salinity of the surrounding water (Case 1978, Grice et al. 1977,Kremlin g et al. 1978,POSE R results) limiting general application inman y marinewaters. -

Grice and Menzel (1978)state d that results of themercur y experiments with 3 3 1300m enclosures showed that the limitations of the smaller 68m bags had been largely overcome; they pointed to the ready development of zoo- plankton in thesever y large containers,i n contrast with declining popu­ lations in earlier experiments,whic hwer e caused by thepresenc e of preda­ tors. However, in the foodweb experiment (Grice et al. 1980)herbivor e 3 predators in the 1300m enclosures again dominated the system for a long time. Finally large enclosures make theprovisio n of many replicates,whic h is desirable inecotoxicologica l experiments,mor e difficult and costly (Davies and Gamble 1979). When smaller bags are used, experimental handling and inter­ pretation of the data is easier and costs are lower.

A9-19 The optimum dimensions of enclosures inplankto n research are related to the aim of the experiment, the number of trophic levels,th e population density at these levels and the species present at these levels.Fo r ecotoxicological experiments with plankton communities, excluding fish and other larger 3 .. carnivores, enclosures with a capacity of 1- 2m appear tob e sufficiently large in eutrophic waters likeDutc h coastal waters,wit h their relatively high zooplankton densities. Inmor e oligotrophic waters, like those inBtfrf£ y Bight, larger bags are preferred to allow sufficiently large zooplankton 3 samples. Enclosures containing about 10m seem tob e large enough in these

ACKNOWLEDGEMENTS

Wewoul d like to thankMarijk e van der Meer,Maria n Pullens and Piet Roele, for analyses they have done.W e also thank Einar Dahl for his provision of data on the development of theplankto n in enclosure NC. We acknowledge the assistance of the captain and crew of the RV Viktor Hensen during our stay in the Rosfjord. We thank theNetherland s Ministry of ForeignAffair s and the Norwegian Authorities for their help inprovidin g experimental facili­ ties in theRosfjord . Part of these investigations was supported by the Deutsche Forschungsgemeinschaftvi a the Sonderforschungsbereich 94"Meeres ­ forschung"Hamburg .Mr . D. v. Assen corrected the English text.

A9-20 5. LITERATURE CITED

1.Barica ,J. , Kling,H . and Gibson,J . (1980). Experimental manipulation of algal bloom composition by nitrogen addition. Can. J. Fish. Aquat. Sei. 37:1175-1183 .

2. Brockmann,U.H. , Eberlein,K. , Junge,H.D. , Trageser,H . and Trahms,K.J . (1974). Einfache folientanks zur Planktonuntersuchung in situ,Mar . Biol. 24i 163-166.

3. Brockmann,Ü.H. ,Eberlein ,K. , Hosumbek,P. , Trageser,H. , Maier-Reimer,E. ,Schöne ,H.K . and Junge,H.D . (1977). The develop­ ment of a natural plankton population ina n outdoor tank with nutrient-poor seawate r 1.Phytoplankto n succession,Mar . Biol.43 : 1-17.

4. Brockmann,U.H. , Eberlein,K. , Junge,H.D. ,Maier-Reimer ,E . and Siebers,D . (1979). The development of anatura l plankton population in an outdoor tankwit h nutrient-poor water II.Change s in dissolved carbohydrates and amino acids.Mar . Ecol. Prog. Ser. 1:283-291 .

5. Brockmann et al. (1980). Geilo paper.

6. Brockmann et al. (1981a). Concept of POSER. Thisvolume .

7. Brockmann et al. (1981b). General methods and technical design of POSER. This volume.

8. Brockmann et al. (1981c).Phytoplankto n paper,thi svolume .

9. Case,J.N . (1978). The engineering aspects of capturing a marine environment,CEPE X and others,Rapp .P . V. Réun.Cons . int.Explor . Mer. 173: 49-58.

10. Colebrook,J.M . (1979). Continuous plankton records: seasonal cycles of phytoplankton and copepods in theNort hAtlanti c Ocean and the North Sea.Mar . Biol._5J_ :23-32 .

11. Davies,J.M. , Gamble,J.C . and Steele,J.H . (1975). Preliminary studies with a large plastic enclosure. In:Estuarin e Research, L.E. Cronin (ed.),Vol . I,Academi c Press,Ne wYork ,pp .251-264 .

12. Davies,J.M . and Gamble,J.C . (1979). Experiments with large enclosed ecosystems. Phil.Trans .R . Soc. Lond. B. 286:523-544 .

13.DeNoyelles ,F. , Knoechel,R. , Reinke,D. , Treanor,D . andAltenhofen ,C . (1980). Continuous culturing of natural phytoplankton comunities in the experimental lake area:Effect s of enclosure,i n situ incuba­ tion, light,phosphoru s and cadmium. Can. J. Fish.Aquat . Sei. 37:424-433 .

14. Donze,M . (1978). Measurements of the effect of heating on survival and growth of natural plankton populations.Verh . Internat. Verein. Limnol. 20: 1822-1826.

A9-21 15.Fumes ,G . (1976).Th eNorwegia nCoasta lCurren tProject .Som e preliminary resultsfro ma synopti cexperimen t inth eNorwegia n coastalcurren t (SEX75) .Repor t 1/76,Univ .Bergen ,3 4pp. , 73figures .

16.Gamble ,J.C. ,Davies ,J.M .an dSteele ,J.H . (1977).Loc hEw eBa g Experiments,1974 .Bull .Mar .Sei .27 :146-175 .

17.Gamble ,J.C .an dDavies ,J.M . (1981).Th eapplicatio nof•enclosure s toth estud yo fmarin epelagi csystems .In :G.D .Gric ean d M.R.Reev e (eds.)Marin eMesocosms ,Springe ri npress .

18.Grice ,G.D. ,Reeve ,M.R. ,Koeller ,P .an dMenzel ,D.W . (1977).Th e useo flarg evolume ,transparent ,enclose d sea-surfacewate r columnsi nth estud yo fstres so nplankto necosystems .Helgolände r wiss.Meeresunters .30 :118-133 .

19.Grice ,G.D. ,Harris ,R.P. ,Reeve ,M.R. ,Heinbokel ,J.F .an dDavis ,CO . (1980).Large-scal eenclose dwate rcolum necosystems .A novervie w offoodwe bI ,th efina lCEPE Xexperiment . J.mar .biol .Ass .U.K .60 :401-414 .

20.Grice ,G.D .an dMenzel ,D.W . (1978).Controlle decosyste mpollutio n experiment:effec to fmercury .VIII .Summar yo fresults . Mar.Sei .Cotma .4 :23-31 .

21.Horstmann ,U . (1972).Übe rde nEinflus svo nhäusliche mAbwasse rau f dasPlankto ni nde rKiele rBucht .Kiele rMeeresforsch .28 :178-198 .

22.Hueck ,H.J. ,Adema ,D.M.M. ,Koc kde ,W.Chr .an dKuiper ,J . (1978). Experienceswit hth evalidatio no fecotoxicologica ltests . Proc.Researc hSemina rEE Gu .Umweltbudesamt ,Berli n1977 . UBABericht ej_0 :159-167 .

23.Kinne ,0 . MarineEcology ,Vol .Ill :Cultivatio npar tI ,Wiley , pp.69-77 .

24.Kolstedt ,S . (1973).Vurderin ga vfjordresipiente ri Farsun dkommune . Rept.Nors kInstitut tfo rVannforskning ,Osl ono .0-139/70 ,pp .74 .

25.Kremling ,K. ,Piuze ,J. ,Bröcke lvon ,K .an dWong ,CS . (1978). Studieso nth epathway san deffect so fcadmiu mi nControlle dEco ­ systemEnclosures .Mar .Biol .48^ :1-10 .

26.Kuiper ,J . (1977a).Developmen to fNort hSe acoasta lplankto n communitiesi nseparat eplasti cbag sunde ridentica lconditions . Mar.Biol .44 :97-107 .

27.Kuiper ,J . (1977b).A nexperimenta lapproac hi nstudyin gth einfluenc e ofmercur yo na Nort hSe acoasta lplankto ncommunity .Helgolände r wiss.Meeresunters .30. :652-665 .

28.Kuiper ,J . (1981a).Ecotoxicologica l experimentswit hmarin eplankto n communities inplasti cbags .In :CD . Gricean dM.R .Reev e(eds. ) MarineMesocosms ,Springe r (inpress) .

A9-22 29. Kuiper,J . (1981b). Fate and effects of mercury inmarin e plankton communities in experimental enclosures. Ecotox. Environm. Safety 5: 106-134.

30. Kuiper,J. , Groenewoud van het,H. , Hoornsman,G. , Roele,P . (1981). Effect of mercury on enclosed plankton communities in the Rosfjord during POSER (this volume).

31. Lee,R.F. , Gardner,W.S. ,Anderson ,J.W. ,Blaylock ,J.W . and Barwell-Clarke,J . (1978). Fate of polycyclic aromatic hydro­ carbons in controlled ecosystem enclosures. Environm. Sei. Techn.j_2 :832-838 .

32. Lund,J.W.G . (1972). Preliminary observations on theus e of large experimental tubes in lakes.Verh . Intern.Verein . Limnol 18:71-77 .

33. Lacaze,J.C . (1974). Ecotoxicology of crude oils and the use of experi­ mental marine ecosystems.Mar . Poll.Bull ._5 : 153-156.

34. Marshall,J.S . and Mellinger,D.L . (1980). Dynamics of cadmium stressed plankton communities. Can.J . Fish.Aquat . Sei. 37:404-414 .

35. Matthews,J.B.L. , Hestad,L . and Bakke,J.L.W . (1978). Ecological studies inKorsfjorden ,Wester n Norway. The generations and stocks of Calanus hyperboreus and Calanus finmarchicus in1971-1974 , Oceanologica acta _1_:277-284 .

36. Menzel,D.W . and Case,J . (1977). Concept and design: Controlled eco­ systems experiment. Bull.Mar . Sei. 2_7: 1-7.

37. Menzel,D.W . and Steele,J.H . (1978). The application of plastic enclosures to the study of pelagic marine biota.Rapp . P.V. Rëun. Cons. int.Explor .Mer . 173:7-12 .

38. Menzel,D.W . (1977). Summary of experimental results:Controlle d ecosystempollutio n experiment. Bull.Mar . Sei. 27:142-145 .

39. Parsons,T.R. , Bröckelvon ,K. , Koeller,P. , Reeve,M.R . and Holm-Hansen,0 . (1977). The distribution of organic carbon in a marine planktonic foodweb following nutrient enrichment. J. exp.mar . Biol.Ecol . 26:235-247 .

40. Parsons,T.R. , Harrison,P.J . and Waters,R . (1978). An experimental simulation of changes indiato m and flagellate blooms.J . exp.mar . Biol. Ecol.32 :285-294 .

41. Parsons,T.R . (1978). Controlled ecosystems experiments.Rapp . P.V. Réun. Cons. int.Explor .Mer . 173:5-6 .

42. Reeve,M.R . and Walter,M.A . (1976). A large scale experiment on the growth and prédation potential of Ctenophore populations. In:Coelenterat e ecology and behaviour,G.O . Mackie (ed.) Plenum Press,Ne w York and London,pp . 187-199.

A9-23 43. Reeve,M.R. ,Grice ,G.D. , Gibson,V.R. ,Walter ,M.A. ,Darcy , K.an d Ikeda,T . (1976). A controlled environmental population experiment (CEPEX)an d its usefulness in the study of largermarin e zooplankton under toxic stress. In:Effect s of pollutants on aquatic organisms, A.P.M. Lockwood (ed.) CambridgeUniv . Press,pp . 145-162.

44. Schelske,C.L. , Stoermer,E.F . (1972). Phosphorus, silica and eutro- phication of lake Michigan. In:Nutrient s and eutrophication, G.E. Likens (ed.) Spec. Symp.Amer . Soc.Limnol .Oceanogr. , pp. 157-170. 3 45. Sonntag,N.C . and Parsons,T.R . (1979). Mixing an enclosed, 1300m water column: effects on the planktonic foodweb.J . Plankton Res.J_ :85-102 .

46. Steele,J.H . and Henderson,E.W . (1976). Simulation of vertical structure in aplanktoni c ecosystem, Scottish Fisheries Research Rept. 5_:1-27 .

47. Steele,J.H. , Farmer,D.M . and Henderson, E.W. (1977). Circulation and temperature structure in large marine enclosures. J. Fish. Res.Bd . Can. 34:1095-1104 .

48. Strickland,J.D.H . and Terhune,L.D.B . (1961). The study of in situ marine photosynthesis using a large plastic bag.Limnol . Oceanogr.6^ : 93-96.

49. Takahashi,M. , Thomas,W.H. , Seibert,D.L.R. ,Beers ,J. , Koeller, P. and Parsons,T.R . (1975). The replication of biological events in enclosed water columns.Arch . Hydrobiol. 76:5-23 .

50. Takahashi,M . and Whitney,F.A . (1977). Temperature, salinity and light penetration structures:Controlle d Ecosystem Pollution Experiment. Bull.Mar . Sei. 27:8-16 .

51. Takahashi,M. ,Wallace ,G.T. ,Whitney ,F.A . and Menzel,D.W . (1977). Controlled Ecosystem Pollution Experiment:Effec t of mercury on enclosed water columns. IManupulatio n of experimental enclosures. Mar. Sei. Coram. 3: 291-312.

52. Topping,G . and Windom,H.L.W . (1977). Biological transport of copper at Loch Ewe and Saanich Inlet: controlled ecosystem pollution experiment. Bull.Mar . Sei. 27:135-141 .

53. Verduin,J . (1969). Critique of methods involving plastic bags in aquatic environments.Trans .Amer . Fish. Soc. 98.: 335-336.

54. Vervelde,G.J . and Ringelberg,J . (1977). Experimentation with eco­ systems. Agro-ecosystems 3:261-267 .

55. Wade,T.L . and Quinn,J.G . (1980). Incorporation, distribution and fate of saturated petroleum hydrocarbons in sediments from a controlled marine ecosystem. Mar.Environm . Res.3 : 15-33.

A9-24 Curriculum vitae

Jan Kuiper werd op 24 mei 1950 inDe n Helder geboren.N a het behalenva n het diploma gymnasium beta aan het Gemeentelijk Lyceum te Den Helder, werd in 1968begonne nme t de studie in de richting milieuhygiëne (die toen nogwater ­ zuivering heette)aa n de landbouwhogeschool teWageningen . In de doctoraal faseva n de studie werden alsvakke nwaterzuivering , kolloid- chemie, natuurbehoud en natuurbeheer en pedagogiek en algemene didactiekge ­ kozen. Een deel van het werkwer d buiten de landbouwhogeschool gedaan op het Nederlands Instituut voor Onderzoek der Zee, de afdeling Hydrobiologie van hetRijk s Instituut voor Natuurbeheer en deVisserijdiens t van het Ministerie van Landbouw en Visserij te Paramaribo. In 1974 werd het ingenieursdiploma behaald.

Vanaf dat jaar is hij werkzaam alsmarie n ecotoxicoloog bij het Laboratorium voor Toegepast Marien Onderzoek van de Hoofdgroep Maatschappelijke Techno­ logie TNO te Den Helder. Daar is ook het in ditproefschrif t beschrevenon ­ derzoek verricht. Het onderzoek is deels gefinancieerd in het kader van het Ie en 2e Environmental Research Programme van de EEG onder contractnr. 110-75-1 ENV N en 227-77-1 ENV N (Al,A2 ,A3 ,A5 ,A7 ,A8) ,deel s doorRijks - waterstaat-direktie Noordzee (A8)e nverde r doorTNO . Publications

J. Kuiper (1973)D e rolva n protozoën in dewaterzuiverin g H„0 6: 491-496 J. Kuiper (1977)Developmen t ofNort h Sea coastal plankton communities in separate plastic bags under identical conditions. Mar. Biol. 44: 97-107. J. Kuiper (1977)A n experimental approach in studying the influence of mercury on aNort h Sea coastal plankton community. Helgolander. wiss. Meeresunters. 30:652-665 . J. Kuiper (1978)Experimente d milieuonderzoek metplanktonisch e ecosystemen. Innovatie 8: 11-12. H.J. Hueck,D.M.M .Adema ,W.Chr . deKoe k and J. Kuiper (1978)Experience s with the validation of ecotoxicological tests. UBA-Berichte 10: 159-167. T. Grolle and J. Kuiper (1980)Developmen t ofmarin e periphyton undermer ­ cury stress in a controlled ecosystem experiment. Bull. Environm. Contam. Toxicol. 24:858-865 . J. Kuiper (1981)Fat e and effects ofmercur y inmarin e plankton communities inexperimenta l enclosures.Ecotox .Environm . Safety 5: 106-134. J. Kuiper (1981)Fat e and effects of cadmium inmarin e plankton communities inexperimenta l enclosures.Mar .Ecol . Prog. Ser. 6: 161-174. W.Chr. deKoe k and J. Kuiper (1981)Possibilitie s formarin e pollution research at the ecosystem level.Chemospher e 10:575-603 . U.H. Brockmann,K.P .Koltermann ,E .Dahl ,A . Dahle,K . Eberlein,A . Gaertner, G. Gassmann, K.D. Hammer, J. Jahnke, G. Kattner, M. Krause, J. Kuiper, M. Laake and K. Nagel (1981) Water exchange in Ros- fjorden during spring '79,a detailed account ofphysical , chemical and biological variations. In: R. Saetre and M. Mork (eds.) The Norwegian Coastal Current.Universit y of Bergen,pp . 93-130. J. Kuiper (1982)Ecotoxicologica l experiments withmarin e plankton communi­ ties in plastic bags. In: G.D. Grice and M.R. Reeve (eds.)Marin e Mesocosms,Springe rVerlag ,Ne w York,pp .181-193 . D.M.M.Adema ,J . Kuiper and A.O.Hanstvei t (1982)Consecutiv e systems of tests for assessment of the effects of chemical agents in the aquatic environment. Proc. 5th International Congress of Pesticide Chemistry (IUPAC),Kyoto , inpres s J. Kuiper (1982)Fat e and effects of 5-nitrofuroicacid-2 (NFA)o n amarin e plankton community in experimental enclosures. Proc. 6th Congress on aquatic toxicology of the ASTM, October 1981, St. Louis, inpres s J. Kuiper,H . van het Groenewoud en G. Hoornsman (1982)Diurna l variation of someplankto n parameters in an enclosed marine community. Neth.J.Sea Res. Inpress .