UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

The effects of fish waste and oxytetracycline on the microbenthos Wu, Henry C. 1992

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_1992_spring_wu_henry.pdf [ 7.1MB ]
Metadata
JSON: 831-1.0086466.json
JSON-LD: 831-1.0086466-ld.json
RDF/XML (Pretty): 831-1.0086466-rdf.xml
RDF/JSON: 831-1.0086466-rdf.json
Turtle: 831-1.0086466-turtle.txt
N-Triples: 831-1.0086466-rdf-ntriples.txt
Original Record: 831-1.0086466-source.json
Full Text
831-1.0086466-fulltext.txt
Citation
831-1.0086466.ris

Full Text

THE EFFECTS OF FISH WASTE AND OXYTETRACYCLINEON THE MICROBENTHOSByHENRY C. WUB.Sc.(H), The University of British Columbia, 1988A THESIS SUBMITTED IN PARTIAL FULFILLMENT OFTHE REQUIREMENTS FOR THE DEGREE OFMASTERS OF SCIENCEinTHE FACULTY OF GRADUATE STUDIES(Department of Zoology)We accept this thesis as conformingto the required standardTHE UNIVERSITY OF BRITISH COLUMBIAApril 1992© Henry C. Wu, 1992In presenting this thesis in partial fulfilment of the requirements for an advanceddegree at the University of British Columbia, I agree that the Library shall make itfreely available for reference and study. I further agree that permission for extensivecopying of this thesis for scholarly purposes may be granted by the head of mydepartment or by his or her representatives. It is understood that copying orpublication of this thesis for financial gain shall not be allowed without my writtenpermission. (Signature)Department ofThe University of British ColumbiaVancouver, CanadaDate  ftr-^3o /97-DE-6 (2/88)11ABSTRACTFish feed, feces and oxytetracycline (OTC) were added to sediment microcosm tanks in1990, and the physical and biological changes in the benthic environment were followed over aperiod of 28 days. The main objective was to study the response of the microfauna (benthic microbialand protozoan communities) to fish waste. Results indicated that anoxic and highly reducingconditions were quickly reached in the sediments within a day, with a shift away from microbiallymediated sulphate reduction to methanogenesis in the degradation of fish waste. Most of the OTCwas quickly washed out of the fish waste when added to the sediments, with small quantities of theantibiotic persisting throughout the duration of the study. Bacterial abundance rapidly increased inthe sediments when fish waste and oxytetracycline were added. While there was a potential for OTCto act on the microfaunal communities, this effect could not be conclusively demonstrated.A second study was conducted in which fish feed, feces and oxytetracycline were addedgradually to sediment microcosm tanks in 1991 for 20 days. The main objective of this study wasto examine the changes in the physical-chemical regime and the benthic microfaunal communityover time. The gradual addition of fish waste to the sediments resulted in microbially mediatedsulphate reduction being the main route for the degradation of fish waste. As most of theoxytetracycline was quickly washed out of the fish waste, the amount of the antibiotic wasinsufficient to affect the bacterial and protozoan communities. There were two rapid increases inciliate abundance, with the initial bloom occurring before the increase in bacterial abundance, andthe decrease in the ciliate population was correlated to increased reducing conditions in thesediments. A large number of heterotrophic microflagellates developed after the ciliate populationdeclined, and these flagellates were observed to be actively grazing on the bacterial population. Theshort time interval of the study was considered to be insufficient for noticeable changes to haveoccurred in the meio- and macrofauna. The rapid response of the microfauna to fish waste suggeststhat changes in the protozoan community could be used as a biological tool in monitoring the impactof fish waste on the environment. More efficient management at aquaculture facilities couldminimize environmental problems.TABLE OF CONTENTSABSTRACT^TABLE OF CONTENTS^LIST OF FIGURES viLIST OF TABLES^ viiiACKNOWLEDGEMENTS^ ixGENERAL INTRODUCTION 1CHAPTER 1. THE EFFECTS OF FISH WASTE AND OXYTETRACYCLINE ON THEMICROBENTHOS. A PRELIMINARY INVESTIGATION.1.1 INTRODUCTION^ 51.2 OBJECTIVES 91.3 MATERIALS AND METHODS1.3.1 EXPERIMENTAL DESIGN^ 91.3.2 APPLICATION OF FISH WASTE AND OXYTETRACYCLINE^ 101.3.3 SAMPLING DESIGN^ 121.3.4 DATA ANALYSIS 171.4 RESULTS1.4.1 SEDIMENT COMPOSITION^ 171.4.2 VISUAL OBSERVATIONS 171.4.3 REDOX PROFILES (Eh)^ 201.4.4 DISSOLVED OXYGEN 231.4.5 CARBON: NITROGEN RATIOS^ 231.4.6 OXYTETRACYCLINE^ 271.4.7 MICROBIAL ABUNDANCE 271.4.8 PROTOZOAN ABUNDANCE^ 30iii1.5 DISCUSSION1.5.1 REDOX MEASUREMENTS^ 301.5.2 MICROBIAL MAT 321.5.3 DISSOLVED OXYGEN^ 331.5.4 CARBON:NITROGEN RATIOS 331.5.5 OXYTETRACYCLINE^ 351.5.6 MICROBIAL ABUNDANCE 371.5.7 POTENTIAL EFFECTS OF OTC ON MICROFAUNA^ 381.5.8 ENVIRONMENTAL CONCERNS^ 44CHAPTER 2. THE EFFECTS OF FISH WASTE AND OXYTETRACYCLINE ON THEMICRO- AND MEIOFAUNA.2.1 INTRODUCTION^ 492.2 OBJECTIVES 512.3 MATERIALS AND METHODS2.3.1 EXPERIMENTAL DESIGN^ 512.3.2 THEORMCAL CALCULATION OF ORGANIC LOADING AND OTC^ 542.3.3 APPLICATION OF FISH WASTE AND OTC^ 562.3.4 SAMPLING DESIGN^ 562.4 RESULTS2.4.1 VISUAL OBSERVATIONS^ 582.4.2 REDOX PROFILES (Eh) 592.4.3 DISSOLVED OXYGEN^ 612.4.4 OXYTETRACYCLINE 652.4.5 BACTERIAL ABUNDANCE^ 652.4.6 MICROBIAL MAT 672.4.7 PROTOZOAN ABUNDANCE^ 682.4.8 DIATOM, NEMATODE AND COPEPOD ABUNDANCE^ 70ivV2.5 DISCUSSION2.5.1 REDOX MEASUREMENTS ^ 722.5.2 MICROBIAL MAT^ 762.5.3 DISSOLVED OXYGEN 762.5.4 OXYTETRACYCLINE^ 772.5.5 MICROFAUNAL ABUNDANCE^ 782.5.6 MICROFAUNAL INTERACTIONS 812.5.7 DIATOMS^ 842.5.8 MEIOFAUNA 852.5.9 INTERACTIONS BETWEEN MICRO- AND MEIOFAUNA^ 872.5.1 0 MACROFAUNA^ 882.5.11 PROTOZOANS IN BIOLOGICAL MONITORING^ 892.6 SUMMARY^ 912.7 RECOMMENDATIONS FOR FUTURE STUDIES^ 92APPENDIX A. EXTRACTION OF ATTACHED BACTERIA IN SEDIMENTSBY ULTRASONICATION^ 94APPENDIX B. EXTRACTION OF OXYTETRACYCLINE FROM SEDIMENTS BYHIGH-PERFORMANCE LIQUID CHROMATOGRAPHY.A CALIBRATION AND RECOVERY STUDY^  103REFERENCES^  111LIST OF FIGURESFigure 1.1. Generalized SAB diagram of changes along a gradient of organic enrichment.Where S=Species Diversity, A=Total Abundance, B=Total Biomass, -PO=Peak of Opportunists, E=Ecotone Point, TR=Transition Zone.(From Pearson & Rosenberg 1978)^ 6Figure 1.2. The arrangement of the ten, 20 litre cylindrical benthic microcosm tanks withblack plastic covers, side profile. A header tank supplies a common source ofseawater to each microcosm^ 11Figure 1.3. A schematic of the 7 transect lines used for randomly sampling the benthos,with the central drainage pipe. The sediment cores would be taken along thetransect line^ 13Figure 1.4. A 3 cc syringe-corer, modified to obtain Eh depth profiles at every 0.5 cm^ 14Figure 1.5. Grain size composition by relative percentages of sediment obtained from 30to 40 ft depth near Spanish Banks/U.B.C. The major constituents were the fineto medium sand at 31% (from 180-355 gm), followed by the very fine sand at29% (from 75-180 gm), and the coarse silt/mud at 26% (< 53 gm)^ 18Figure 1.6. Reduction-oxidation time-series depth profiles of the replicate tanks for theControl Treatment in 1990. Values < +100 mV (dashed line) indicates reducingconditions. Measurements began at 0.5 cm above the sediment surface 21Figure 1.7. Reduction-oxidation time-series depth profiles of the replicate tanks forTreatment A in 1990, a low dosage of oxytetracycline. Values < +100 mV(dashed line) indicates reducing conditions. Measurements began at 0.5 cmabove the sediment^ 22Figure 1.8. Reduction-oxidation time-series depth profiles of the replicate tanks forTreatment B in 1990, a high dosage of oxytetracycline. Values < +100 mV(dashed line) indicates reducing conditions. Measurements began at 0.5 cmabove the sediment 24Figure 1.9. Comparison of the dissolved oxygen levels in seawater between the Control(fish waste added), Treatment A (fish waste and 2.53 g of OTC added), andTreatment B (fish waste and 10.13 g of OTC added). Plotted values are theMean ± 1 Standard Deviation, n=3^ 25Figure 1.10. Carbon:nitrogen ratios of the Control (fish waste added), Treatment A(fish waste and 2.53 g of OTC added), and Treatment B (fish waste and 10.13 gof OTC added). Plotted values are the Mean ± 1 Standard Deviation, n=3^26Figure 1.11. The concentrations of oxytetracycline (ppm) for Treatment A (2.53 g of OTC)and Treatment B (10.13 g of OTC). Plotted values are the Mean ± 1 StandardDeviation, n=3^ 28Figure 1.12. Bacterial abundance (cells x 10 9 • 1 sediment dry weight) between the Control(fish waste added), Treatment A (fish waste and 2.53 g of OTC added), andTreatment B (fish waste and 10.13 g of OTC added). Plotted values are theMean ± 1 Standard Deviation, n=3^ 29viFigure 1.13. Electron free-energy diagram for the biologically mediated redox sequencesof the degradation of organic carbon, CH2O (acting as the electron donor). Note therelationship between the change in energy (G°), redox processes and the Eh (mV)(from Zehnder & Stumm 1988) 31Figure 2.1. The arrangement of the 200 L tanks in which fish waste and oxytetracyclinewere applied. The order of the tanks for the treatments was randomly chosen'^ 52Figure 2.2. Grain size composition by relative percentages, of the sediment obtained fromthe intertidal zone during low tide at Spanish Banks/U.B.C. The majority of thesediment is comprised of fine to medium sand (65%), followed by a medium tocoarse sand component (26%)^ 53Figure 2.3. A schematic of the division of grids used to randomly sample the sediments ofthe 200 L tanks. The sediment is divided into a system of 3 columns (A-C) and5 rows (1-5) with a central drainage pipe and 1.5" border which was not sampledto avoid edge effects^ 55Figure 2.4. Reduction-oxidation time-series depth profiles of the replicate tanks for theBlank Treatment, followed over 20 days in which no fish waste or oxytetracyclinewas added. Values < +100 mV (dashed line) indicates reducing conditions.Measurements began 0.5 cm above the sediment surface^ 60Figure 2.5. Reduction-oxidation time-series depth profiles of the replicate tanks for theControl Treatment, followed over 20 days in which fish waste but nooxytetracycline was added. Values < +100 mV (dashed line) indicates reducingconditions. Measurements began 0.5 cm above the sediment surface^ 62Figure 2.6. Reduction-oxidation time-series depth profiles of the replicate tanks for theOTC Treatment, followed over 20 days in which fish waste and oxytetracycline wasadded. Values < +100 mV (dashed line) indicates reducing conditions.Measurements began 0.5 cm above the sediment surface^ 63Figure 2.7. Comparison of the dissolved oxygen levels in seawater between the BlankTreatment (fish waste and OTC not added), Control Treatment (fish waste addedbut not OTC), and OTC Treatment (fish waste and OTC added). Plotted valuesare the Mean ± 1 Standard Deviation, n=3^Figure 2.8. Comparison of bacterial abundance (cell x 10 9 -g- ' sediment dry weight) over20 days, between the Blank Treatment (fish waste and OTC not added), ControlTreatment (fish waste added but not OTC), and OTC Treatment (fish waste andOTC added). Plotted values are the Mean ± 1 Standard Deviation 66Figure 2.9. Comparison of protozoan abundance over 20 days, between the BlankTreatment (fish waste and OTC not added), Control Treatment (fish waste addedbut not OTC), and OTC Treatment (fish waste and OTC added); where A=ciliateabundance (numbers x 10 3 -cm-2); B=heterotrophic microflagellate abundance(numbers x 106-cm-2). Plotted values are the Mean ± 1 Standard Deviation^69Figure 2.10. Comparison of diatom abundance (numbers x 10 2-cm-2) over 20 days, betweenthe Blank Treatment (fish waste and OTC not added), Control Treatment (fish wasteadded but not OTC), and OTC Treatment (fish waste and OTC added). Plottedvalues are the Mean ± 1 Standard Deviation^ 71vii64Figure 2.11. Comparison of nematode abundance (numbers•cm -2) over 20 days, between theBlank Treatment (fish waste and OTC not added), Control Treatment (fish wasteadded but not OTC), and OTC Treatment (fish waste and OTC added). Plottedvalues are the Mean ± 1 Standard Deviation^ 73Figure 2.12. Comparison of harpacticoid copepod abundance (numbers•cm -2) over 20 days,between the Blank Treatment (fish waste and OTC not added), Control Treatment(fish waste added but not OTC), and OTC Treatment (fish waste and OTC added).Plotted values are the Mean ± 1 Standard Deviation 74APPENDIX AFigure 1. Accumulation of bacteria attached to sediments following ultrasonication timesof 30, 60, 90, 120, 180, 240 and 300 seconds, of a formalin preserved sand sample.Values presented are the Mean ± 1 Standard Deviation, n=5^ 97Figure 2. The destructive effects on a bacterial suspension extracted from the sediment,following ultrasonication times of 30, 60, 90, 120, 180, 240 and 300 seconds.Values presented are the Mean ± 1 Standard Deviation, n=4^ 98Figure 3. The percentage of bacteria remaining on sediments, following ultrasonicationperiods of of 90, 150, 210 and 270 seconds. The relationship is linearized by logtransformation and fitted with a first order regression and 95% confidence limits,=16 100LIST OF TABLESTABLE 1. Total Carbon content (percent) of the sediments in the Control, Treatment A,and Treatment B. Values presented are the Mean % ± 1 Standard Deviation (SD)....34viiiixACKNOWLEDGEMENTSI would like to thank my supervisor Dr. T.R. Parsons, and the other members of my ResearchCommittee (Dr. C.D. Levings and Dr. R. Petrell) for their advice and continuing support throughoutthis study. My thanks to the following people for their invaluable assistance; S. Mattice in the set-up and maintenance of the microcosm tanks, Dr. H. Rogers who kindly provided access to his laband HPLC (DFO, West Vancouver) and M. Sadar in the analysis of oxytetracycline in 1990, Dr. K.McErlane who provided an HPLC in 1991 and to R. Aoyama for his suggestions during the analysisof oxytetracycline.I am especially grateful for the useful criticisms of my Research Committee, and from Dr.M. St. John, Dr. C. Lalli and J. Berges on various drafts of this thesis. The suggestions provided bySCARL (Statistical Consulting and Research Laboratory, Dept. Statistics), J. Berges, D. Montagnes,T. Sutherland and M. Adl during the course of this research were particularly helpful.Funding for this project was made available from NSERC operating grants to Dr. T.R.Parsons with additional support by the Department of Fisheries and Oceans through Dr. C.D.Levings, and from operating grants provided by the Donner Foundation.My deepest gratitude to my parents and Jo for their encouragment, patience and understand-ing, and to whom I would like to dedicate this thesis.GENERAL INTRODUCTIONWorld aquaculture (the production of finfish, crustaceans, molluscs and seaweeds)has increased greatly in the last three decades (Iwama 1990), from 5.6 million t in 1975 to aprojected value of over 22.2 million t in the year 2000. The socioeconomic aspects ofaquaculture can be considerable, being an important industry in some countries (e.g.Norway) and employing many people (Ridler 1984, Ford 1984, Folke & Kautsky 1989).In British Columbia, the aquaculture industry dates back to 1912 with Pacific oystercultivation, while trout (freshwater) culture began in the 1950s and salmon (marine) culturein the 1970s. The farming of other aquatic species in B.C. such as clams, scallops, mussels,Arctic char, black cod (sablefish), sturgeon and marine plants (e.g. kelp) is just beginning.By far the largest and most successful activity (total value) has been in marine fish culture.While salmon farming is a relatively young industry, there have been large increases in thenumber of farms and production, from eight operating farms producing 107 t of chinookand coho salmon in 1985 (Korman 1989), to 125 farms in 1991 producing 16,500 t ofchinook, coho, and Atlantic salmon worth $105 million (B.C. Salmon Farmers Association1992). An additional 68 farm sites could be in operation within the next few years oncelicensing and development is completed.Salmon farming along the coast of British Columbia is intensive in local areas. Therearing of fish occurs in floating net pens and the diet is composed almost exclusively ofartificial feed. Unfortunately, the output from this intensive culturing practice isunavoidably released locally into the marine environment, affecting water and sedimentquality. This output is composed of the uneaten feed and metabolic by-products of the fish,of which the major constituents are organic carbon and nitrogen compounds (carbohydrate,lipid and protein), followed by phosphate, ammonium, urea, bicarbonate, vitamins,therapeutants and pigments (e.g. Gowen & Bradbury 1987, Enell 1987, AquametrixResearch Ltd. 1988). The particulate excess feed and fecal material sink to the sediments in1the immediate area around the fish farms, while the soluble components (e.g. ammonium,urea) are diluted and dispersed to the water column (e.g. Hall & Holby 1986, Enell 1987).The magnitude and nature of this fish waste is affected by the biomass of the farm andfeeding methods used, which determine the amount of food fed and wastage levels fromuneaten feed and fecal material (Iwama 1990). The size of the fish determines the size ofthe feed and types of feed used (e.g. starter feed vs production feed, dry vs moist feeds) allof which are involved in the rate of breakdown in the excess feed and the type of fecalmatter produced. These are important characteristics to consider as the size of the particlesare inversely related to the areal dispersal pattern (smaller, lighter particles sink slower andare dispersed farther) which further depends on the physical oceanography of the region(currents, tides, and bottom topography) and on the distance of the net pen above the oceanfloor (Weston 1986).Salmon farming will result in changes to the nearby marine environment, alteringthe physical, chemical and biological regimes. There are several potential environmentaleffects (as reviewed by Weston 1986, with some modifications as increased research hasbroadened the scope of problems):(1) Changes in water circulation.(2) Sedimentation and accumulation of the fish waste beneath the culture operation.(3) Alterations in water chemistry.(4) Changes in plankton and benthic algal biomass and productivity.(5) Changes in structure of the benthic faunal communities.(6) Introduction of exotic species and subsequent changes in the genetic fitness of wildstocks.(7) The effects of antibiotics on the surrounding biota and development of bacteriapathogenic to humans.(8) Disease transmission from cultured to wild stocks.2Since the review of the environmental impacts of mariculture by Weston (1986),there has been a great deal more interest worldwide of the potential environmental problemsassociated with this rapidly expanding industry (Gowen & Bradbury 1987, Eng et al. 1989,Munro 1990, Iwama 1990). Further research to quantify the nutrient flux from maricultureactivities has resulted in the development of models (Hall et al. 1990, Holby & Hall 1991,Ackefors & Enell 1990). The possibilities for disease and parasite transmission, and forgene transfer between wild and escaped farmed fish resulting in reduced genetic variabilityhave also been examined (Windsor & Hutchinson 1990, Carss 1990, Egidius et al. 1991).There have been recent investigations on the effects of antibiotics (used to treat diseaseoutbreaks at fish farms) on the microbial community in marine sediments in Norway(oxytetracycline, oxolinic acid, furazolidone; e.g. Samuelsen et al. 1991 & 1992, Nygaardet al. 1992). Additional work continues on the behaviour of these antibiotics in the marineenvironment and fauna (Samuelsen et al. 1988, Lunestad & Goksoyr 1990, Bjorklund et al.1990 & 1991) and on the potential for development of antibiotic resistance (Torsvik et al.1988, Husevag et al. 1991). The physical and chemical changes of the sediments (due tothe high sedimentation rates of fish waste from net and cage pens) have been studied (Hall& Holby 1986, Samuelsen et al. 1988, Kaspar et al. 1988). The possibility that nutrient-enriched environments (a condition known as hypernutrification or eutrophication) aroundaquaculture facilities may support increases in the planktonic and benthic algal communitieshas also been considered (Weglenska et al. 1987, Parsons et al. 1990, Korman 1989,Stirling & Dey 1990, Carr & Goulder 1990). The response of the macrofauna toenrichment of the benthic environment by excess fish feed and fecal matter has been wellestablished (Brown et al. 1987, O'Connor et al. 1989, Lumb 1989, Lumb & Fowler 1989,Frid & Mercer 1989, Ye et al. 1991).3However, the effects of fish waste should not be restricted solely to the macrofauna,as an integral benthic community consists of micro- and meiofaunal components as well(Fenchel 1978). But changes in the micro- and meiofauna due to organic enrichmentresulting from aquaculture activities have never been considered before. Thus, the overallobjective of this study was to examine the responses of the microfaunal community (and themeiofaunal community to a limited extent) to fish waste and the antibiotic oxytetracycline.Oxytetracycline was chosen as it has been the most commonly used drug in the treatment ofdiseases at fish farms for approximately the last two decades (Grave et al. 1990). For acomprehensive review of the environmental impacts associated with aquaculture, the readershould refer to Weston (1986), Gowen & Bradbury (1987), Iwama (1991) and Levings(1992).41. THE EI4ECTS OF FISH WASTE ANDOXYTETRACYCLINE ON THE MICROBENTHOS.A PRELIMINARY INVESTIGATION.1.1 INTRODUCTIONThere is a lack of comprehensive research on changes in the microbenthiccommunities (bacterial and protozoan populations) beneath fish farms caused by fish wasteand oxytetracycline, a commonly used antibiotic. Previous studies have concentratedmainly on changes in meio- and macrofaunal communities in the sediments due to fishwastes (e.g. polychaetes, molluscs, amphipods, copepods, snails; Weston 1986 & 1990,Moriarty 1986, Brown et al. 1987, Gowen & Bradbury 1987, Lumb 1989, O'Connor et al.1989, Ritz et al. 1989). The terms micro-, meio- and macrofauna refer to the faunalcommunities of the benthic environment, and are used interchangably with micro-, meio-and macrobenthos. Only a few recent reports have examined the benthic diatomcommunities around aquaculture sites (Stirling & Dey 1990, Carr & Goulder 1990), and theremaining studies have focused on bacterial abundance and resistance to antibiotics (e.g.Torsvik et al. 1988, Husevag et al. 1991, Samuelsen et al. 1988, 1991 & 1992).Because benthic biota are tolerant of a certain amount of environmental variation,the responses of each benthic species will be determined by the magnitude of organicenrichment (Blackstock 1984). Characteristic changes in the species, abundance andbiomass of the meio- and macrobenthic communities in response to pollution are bothtemporal and spatial (e.g. Pearson 1975, Pearson & Rosenberg 1978, Bagheri & McLusky1982, Oviatt et al. 1987, Warwick et al. 1987, Whitehurst & Lindsey 1990, Moore &Rodger 1991). As organic enrichment decreases along a gradient from a central source(Fig. 1.1), there is an initial small increase in biomass corresponding to large abundances ofsmall opportunistic species, followed by a decline to the ecotone point. This ecotone pointis a transition zone in which a community is poor in species, abundance and biomass5E• 4r....■■■■•• TR elr.■.............►PO•INCREASING ORGANIC INPUT6Figure 1.1. Generalized SAB diagram of changes along a gradient of organic enrichment.Where S =Species Diversity, A =Total Abundance, B=Total Biomass, PO=Peak ofOpportunists, E=Ecotone Point, TR=Transition Zone. (From Pearson & Rosenberg1978).(Pearson & Rosenberg 1978). After the ecotone point and continuing along the gradient ofdecreasing organic enrichment, biomass increases to a second maximal level with a greatervariety of species and lowered abundance. Species diversity, abundance and biomass levelseventually stabilize to resemble non-polluted communities at the outer boundaries of thezone of pollution. Changes in part, or all of the microbenthic community due toanthropogenic pollution (domestic, agricultural and industrial) are also similar to changes inthe meio- and macrobenthic communities (e.g. Small 1973, Curds 1973, Pearson 1975,Madoni & Ghetti 1981, Wyatt & Pearson 1982, Heip et al. 1985, Hul 1987 & 1988,Stoessel 1989). Those benthic species that are highly sensitive or are opportunistic inorganically enriched sediments have been used as indicator organisms to identify the degreeof pollution (e.g. mussels, nematode/copepod ratio, polychaetes, protozoans; Raffaelli &Mason 1981, Raffaelli 1987, Grabacka 1985, Dale 1987, Warwick et al. 1988, Manru etal. 1988, Gray et al. 1988, Warwick et al. 1990).There has not been a comprehensive study of the effects of fish waste and OTC onthe microbenthic community for a number of reasons. Firstly, aquaculture has not been alarge industry worldwide until the last two decades (see reviews by Phillips et al. 1985,Weston 1986, Windsor & Hutchinson 1990, Munro 1990, Iwama 1990, Svealv 1991).Only recently has more attention been directed toward calculating the different componentsand quantities of the pollutants from aquaculture facilities (i.e. fluxes and mass balances;Enell 1987, Korman 1989, Aure & Stigebrandt 1990, Hall et al. 1990, Holby & Hall1991). Secondly, macrofauna have been traditionally used in pollution monitoring duemainly to the practical advantages of handling larger organisms in the field and laboratory,and because the taxonomy and general biology is more widely known (Gray et al. 1988).Thirdly, the main drawback involved in using microfauna (and meiofauna) in pollutionstudies are the taxonomic difficulties of the many different species. Finally, the logistics ofquantitative biological surveys require that soft-bodied organisms (e.g. protozoans) beexamined live, or fixed and preserved either in the sediment or immediately after extraction7(Heip et al. 1988). While the sediment microbial population can be easily preserved andcounted (Montagna 1982), quantitative extraction, identification, fixation and preservationof benthic protozoans (also referred to as 'ciliates'), is an ongoing problem (Schwinghamer1981, Alongi 1986). However, the advantages in using the microfauna are that smallersediment samples are needed (i.e. field sampling is less labour intensive), and shortergeneration times can result in potentially faster responses to pollution. Thus, the benthicmicrofauna can serve as a more sensitive indicator of pollution from mariculture sites (e.g.Finaly et al. 1979).Benthic microcosm tanks were used to assess the changes in the microfaunalcommunity that may result from pollution of the sediments with fish waste andoxytetracycline. Although microcosm tanks may not provide a totally realistic benthiccommunity structure because they may uncouple the physical and biological processes(Pritchard & Bourquin 1984, Federle et al. 1986, Findlay et al. 1990a), they have still beenused extensively to study the benthos (e.g. Hargrave 1972, Kelly & Nixon 1984, Alongi &Tenore 1985, Nedwell & Lawson 1990, Hansen & Blackburn 1991, Sundback et al. 1991).Controlled conditions are always required to determine a definitive cause and effectrelationship between the organisms and pollutants involved, whereas this relationship canonly be inferred from field studies (Underwood & Peterson 1988, Lasserre 1990, Pilson1990). Directions for possible future studies and interpretations of general interest can thusbe obtained from the use of benthic microcosms in this study. Any experimental resultsobtained from a laboratory setting must eventually be applied towards understandinginteractions in the field.81.2 OBJECTIVESFollowing enrichment of the sediments with fish waste and oxytetracycline, theobjectives of this initial study were:(1) To evaluate the physical-chemical changes to the benthos such as (i) oxygen levels ofseawater due to the biological oxygen demand of the sediments, (ii) sediment depthprofiles of the reduction-oxidation potentials, and (iii) carbon: nitrogen ratios of thesediments.(2) To follow the decrease of oxytetracycline in the sediments over time.(3) To determine possible changes in the abundance of microfauna in the sediments(bacterial and protozoan populations).1.3 MATERIALS AND METHODS1.3.1 EXPERIMENTAL DESIGNA controlled experimental ecosystem approach was developed in which a series ofmicrocosms (< 1 m 3 volume, Lasserre 1990) were used to create a benthic environmentenriched in fish waste. Seawater and tank facilities were located at West VancouverLaboratories, Department of Fisheries and Oceans, West Vancouver, British Columbia,Canada. The duration of this experiment was 28 days, from July to August 1990.Sediment was obtained with a Shipek grab sampler (3.3 L), near Spanish Banks/U.B.C.,from 30 to 40 ft depth during low tide. This ensured that the sediment and benthiccommunity was from a depth similar to that expected from net pens (e.g. Weston 1986),and any changes to the benthos could be understood better in relation to natural fieldconditions. Three to four grabs were combined to yield enough sediment for themicrocosm tanks (4 to 6 cm depth). A period of two weeks was allowed for the biologicaland chemical properties of the benthic system to be re-established (Federle et al. 1986).9Ten, 20 L cylindrical tanks were used (Fig. 1.2), with a header tank supplying acontinuous flow of seawater to each tank at 0.5 L.min -1 . This was the maximum rate offlow that would not disturb the sediments in the tanks. Drainage was by a central pipe inthe centre of each tank, with the outflow near the water surface. Water temperature andsalinity ranged from 13 to 16°C and 27 to 29 0/. throughout the study period, respectively.Based on 9 tanks, 3 treatments were designed with 3 replicates per treatment, all with thesame levels of fish waste (feces and feed) but with differing levels of oxytetracycline(OTC). The treatments consisted of a relatively low and high concentration of OTC, with acontrol series (no OTC added) for comparison.1.3.2 APPLICATION OF FISH WASTE AND OXYTETRACYCLINEFish feces and excess feed were collected from 10,000 L outdoor tanks at WestVancouver Laboratories, West Vancouver, B.C. with the assistance of Mr. S. Mattice. Theantibiotic oxytetracycline was provided by Argent Laboratories, Richmond, B.C., in theform of Oxysol 440, which contains 100 g of OTC per 227 g Oxysol 440. A treatment levelof 11.5 g Oxysol 440 per kg of fish feed was prescribed (Dr. E.L. Dahl, D.V.M., pers.comm.).The fish waste was applied in all tanks to a depth of 2 cm, to resemble mild toheavy loading beneath fish farms (Jacobsen & Berglind 1988). The fish waste wasmeasured using a 10 L container and poured through a funnel and 1.5" diameter plastichose, with the end of the hose held just above the sediment surface. To avoid disruption ofthe established reduction-oxidation characteristics, care was taken to avoid disturbing thesediment as the fish waste was evenly and carefully layered on the sediment surface. TheControl treatments consisted of only fish waste, while the low concentration treatment ofOTC (Treatment A) contained 5.75 g of Oxysol 440 (or 2.53 g OTC) in the fish waste.The high concentration treatment of OTC (Treatment B) contained 23 g Oxysol 440 (or10.13 g OTC). The Oxysol 440 was added directly to the fish waste and stirred well to1 011Figure 1.2. The arrangement of the ten, 20 litre cylindrical benthic microcosm tanks withblack plastic covers, side profile. A header tank supplies a common source of seawaterto each microcosm.ensure complete mixing before addition to the sediments. The replicate tanks for eachtreatment were randomly chosen and dark covers placed over all of the tanks to minimize(i) autotrophic activity due to poor and variable indoor lighting conditions, and (ii)epiphytic growth along the sides of the tanks.The design of this experiment thus monitored the recovery of the benthicenvironment following organic enrichment and chemotherapeutic treatment.1.3.3 SAMPLING DESIGNInitial samples were taken before the fish waste and oxytetracycline were added (onday 0). Subsequent samples were taken on the 2nd, 4th, 7th, 14th, 21st and 28th dayfollowing organic enrichment of the benthos. To avoid biased samples due to edge effects,sampling was completed along a series of 7 transect lines, radiating outwards from thecentre of the tank and arranged equi-distance from each other (Fig. 1.3). Four cores weretaken per tank on each sampling day for (i) redox measurements, (ii) bacterial analysis, (iii)total carbon and nitrogen and OTC, and (iv) microfauna (ciliates). These four cores weretaken from one transect line, with the transect line and the order of the cores along thetransect randomly determined for that day and maintained consistently for all the tanks.Reduction-Oxidation Profiles (Eh)A reduction-oxidation (redox) depth profile was taken by inserting a 3 ml syringe(i.d. 0.85 cm, area sampled 0.567 cm 2) into the sediment to obtain Eh readings from anundisturbed core sample (Fig. 1.4). The rubber plunger ring was removed and fitted with athin glass tube, to which was attached a short piece of rubber tubing with an autoclaveclamp. Silicone sealant applied around the glass and rubber ensured an airtight seal and therubber ring was refitted onto the syringe. A series of 1 mm diameter holes were drilledthrough one side of the syringe, at 0.5 cm intervals. Before the syringe corer was insertedinto the sediment, parafilmTM was tightly wrapped around the corer to cover the holes, andthe clamp was opened. After the corer was pushed into the sediment, the clamp was closed1238 cmTransect Linesfor Sampling13Figure 1.3. A schematic of the 7 transect lines used for randomly sampling the benthos, withthe central drainage pipe. The sediment cores would be taken along the transect line.3 mm i.d.Surgical Tubing14 Autoclave ClampGlass Tube17-71-^  Rubber Plunger RingiE37•-^ • lmm Hole•••••••• +0.5 cm Spacing•Luer End of SyringeRemoved^—Figure 1.4. A 3 cc syringe-corer, modified to obtain Eh depth profiles at every 0.5 cm.and the resulting vacuum within the syringe acted to maintain the sediment in positionwithin the corer as it was withdrawn from the sediment. To prevent the sediment fromslipping out of the corer during Eh measurements, another piece of wax paper was wrappedaround the bottom end of the corer. The tip of the electrode was then inserted through thewax paper and holes of the syringe to obtain the Eh reading. Readings were taken at 0.5cm above the sediment, then at the sediment-water interface, and at 0.5 cm depth intervalsthereafter. Although the Eh profile has been shown to change with 5. 1 mm intervals(Reimers et al. 1984, Revsbech et al. 1989), the size of the core and electrode tip-width didnot allow measurements at smaller intervals without excessive disruption of the core. Theredox potentials were measured with an Accumet Model 320 pH meter fitted with aPlatinum Combination Electrode (Fisher #13-620-82, electrode tip approximately 1 cmlong, 0.8 mm in width), and calibrated with ZoBell's solution prior to, and during,measurement of the redox potentials (ZoBell 1946). This solution was used primarily toprevent electrode poisoning and secondarily to clean the tip (Whitfield 1969), and it wascomposed of 0.0033 M potassium ferrocyanide to 0.0033 M ferricyanide in 0.10 Mpotassium chloride (e.g. ZoBell 1946, Pearson & Stanley 1979, Brown et al. 1987).Dissolved OxygenWater overlying the sediments was measured for dissolved oxygen by colourimetricanalysis (Parsons et al. 1984a). This measurement of oxygen can be taken as a roughindication of the changes occurring in the sediment chemistry due to the biological oxygendemand (BOD) of the sediments, as the exchange of gases (oxygen and carbon dioxide)occurs between the sediment surface and water column (Pearson & Rosenberg 1978; Brownet al. 1987). This measurement was expressed as the percent saturation of dissolvedoxygen in seawater.ProtozoansAnother 3 ml corer was used to collect the sediment, using a PercollTm-sorbitol gel(Sigma Chemical Co., St. Louis, MO) and centrifugation technique (Schwinghamer 1981,15Alongi 1986). This relatively fast technique has been shown to have better than a 90%extraction efficiency from muddy and sandy sediments for most microfauna. Due to thetendency of Percollim-sorbitol to gel in the presence of formalin or Lugol's solution(personal observations, Alongi 1990), the Percollml-sorbitol was drained by filtrationthrough 8 gm filters at low pressures (1/2 atm), and the organic matter remaining on thefilter was then washed into 20 ml scintillation vials with a 2% formalin-seawater solution(Bullough 1962).BacteriaThe core for the bacterial sample was taken with a 1 ml syringe (i.d. 0.4 cm, areasampled 0.126 cm2 , volume 0.628 cm3), with the rubber stopper ring modified as for theEh corers. The sediment was preserved with 3.6 ml of 0.22 Am filtered artificial seawaterand 0.25 ml of 37% formaldehyde (final concentration of 2.1% formaldehyde), and it wasanalyzed for bacteria after two weeks, as outlined in Appendix A.Carbon:Nitrogen RatiosA separate core from a 3 ml syringe (with the rubber stopper ring modifiedaccordingly) was taken and immediately frozen at -12°C. The core was then thawed andmixed well, and a small subsample was taken (<20 mg) for analysis of total organic carbonand nitrogen. The remainder of the sample was used for analysis of oxytetracycline. Thesubsample of sediment was dried at 60°C for 3 days, ground to a fine powder, and thenVanadium Pentoxide V205 (BDH Ltd., Poole England) was added to the sediment samplesbefore being subject to an autoanalyzer (Carlo Erba, Model NA 1500) for determination oftotal carbon and nitrogen.Oxytetracycline (OTC)The remaining sediment from the previous core was used for analysis of OTC,following a highly sensitive method (Jacobsen & Berglind 1988) outlined in Appendix 2.An external standard was calculated by dissolving 100 mg of OTC in 10 ml of methanol(2x10-4 M). Analysis of samples was on a Hewlett Packard system using a 1084A liquid16chromatograph, 1030B variable-wavlength UV detector operated at a wavelength of 350 nmand connected to a 79850A LC terminal. The volumes of the samples that were analyzedhad to be increased to 200 pi to obtain quantifiable readings.1.3.4 DATA ANALYSISA repeated measures ANOVA was used in the analysis of data followed by a TukeyMultiple Range Test when appropriate, using the statistical package SYSTAT (Ver 5.0,Evanston, IL). When the assumptions of equality of variance and normality were violatedand transformations were unable to correct the problems, the non-parametric Kolmogorov-Smirnov Two-Sample Test was used.1.4 RESULTS1.4.1 SEDIMENT COMPOSITIONGrain sizes of the sediment were determined by sieve analysis (as in Harrison 1981).A sample of sediment was dried at 80°C for 5 days before being sieved through a nestedseries of Canada Standard sieves (The W.S. Tyler Co. Ltd., St. Catharines, Ont.) of 595,355, 180, 75 and 53 Am mesh diameter. The sediment retained on each screen wasweighed and the weight of the subsieve material was estimated by the difference (Fig. 1.5).Based on the Udden-Wentworth scale, the sediment composition may be described as beingof fine to medium sand (75 - 355 Am) with a large silt (or mud) component (<54m)(Buchanan 1984).1.4.2 VISUAL OBSERVATIONSThe sediment surface was initially a grey colour and devoid of benthic epifauna. Byday 2, a smooth white microbial film had begun to develop on the sediment surfaces of theControl tanks, whereas the sediment surfaces of the tanks in Treatments A and B were1718403530og)o 25----LS 20--4a)c)(1) 15(i)1050<53^53-75^75-180^180-355^355-595^>595Sediment Size Range (urn)Figure 1.5. Grain size composition by relative percentages of sediment obtained from 30 to 40ft depth near Spanish Banks/U.B.C. The major constituents were the fine to mediumsand at 31% (from 180-355 Am), followed by the very fine sand at 29% (from 75-180Am), and the coarse silt/mud at 26% (< 53 Am).beginning to turn a dark brown colour. Numerous macrofauna, not quantified but mainlycomprised of large polychaetes and a few brittlestars, appeared on the sediment surfaces ofthe tanks in all treatments.By day 4, the white microbial mat covered approximately three-fourths of thesediment surface of the tanks in the Control treatment. Only small patches of this microbialfilm had begun to develop in the tanks of Treatment A, and the bacterial layer was even lessdeveloped in the tanks of Treatment B. All of the macrofauna had disappeared from thesediment surfaces, and were assumed to have re-entered the sediment. An attempt wasmade to clean the tank walls of epiphytic growth that had appeared at this time, but due tothe small volume of the tank and the physical agitation of the water column, the whitemicrobial film developing on the sediment surfaces was easily disrupted, and further effortsto clean the tank walls were not made.By day 7, the microbial film had declined in all tanks, only partially covering thesediment surfaces of tanks 2 and 3 in the Control treatment and with very little appearing intank 1. This patchy distribution of the white microbial layer was also prevalent in the tanksof Treatments A and B. The sediment surfaces in tank 1 of the Control remained a darkbrown by day 14, while tanks 2 and 3 were again covered by a white film. This microbialmat was also well developed in all tanks of Treatment A, but was only partially present intank 2 of Treatment B. The sediment surfaces of tanks 1 and 3 of Treatment B were grey-brown in colour, but the white mat began to be re-established on the surfaces of these tanksby day 21. The colour of the sediment to a depth of 4 cm in Treatments A and B was avery dark brown to black, with a black layer in the top 2 cm. The colour of the sediment inthe Control was a light brown at depth, but also with a 2 cm thick black surface layer.After day 21, the microbial layer began to disappear from the sediment surfaces ofall tanks, but remained for a longer period of time and disappeared slower in Treatments Aand B. By the final day of sampling, day 28, the sediment surfaces of all tanks were a lightto dark brown colouration. Patches of the white microbial mat remained in some of the19tanks of Treatments A and B, but only in areas of the sediment disturbed by previoussampling.1.4.3 REDOX PROFILES (Eh)The reduction-oxidation (redox) potential is a quantitative measure of the energy ofthe electron escaping tendency of a reversible system. This can be measured with anelectrode system as Eh in millivolts (mV, hydrogen scale), E being the potential differencebetween the standard hydrogen electrode and the system in which the redox potential isbeing measured (Zobell 1946, Whitfield 1969). The redox profiles exhibited by the tanksof the Control treatment are all similar in pattern (Fig. 1.6). Organic enrichment of thesediment caused strongly reducing conditions by day 2 at the sediment surfaces and atdepth, as indicated by Eh values < +100 mV. This point is the lower boundary of freeoxygen in the oxidized zone (Jorgensen & Fenchel 1974) below which the sediment isreducing (Fenchel 1969, Jone 1979, 1981, Harrison 1981). Redox conditions were mostreducing on day 4 at -160 mV and -318 mV for replicate tanks 1 and 2 respectively, and onday 21 at -316 mV for tank 3; thereafter the sediment became less reducing overall.Reducing conditions initially increased with sediment depth before stabilizing, but after day14 in tanks 2 and 3 the redox profile began to become less reducing at depth - this conditionnever occurred in tank 1. A redox potential discontinuity (RPD) layer was always presentbetween 0 to 1 cm beneath the sediment surface, a zone where an abrupt change in Ehvalues occurs with depth.Redox values at the sediment surfaces at day 0 in tanks 2 and 3 of Treatment Aindicated that oxidizing conditions were present, with the RPD layer extending down to 2cm and becoming very reducing at this depth (Fig. 1.7). By day 2 after the addition of fishwaste and oxytetracycline, the sediment surfaces had become reducing. Redox profileswere similar to those in the Control treatment, with a low Eh of -285 on day 4 measured intank 1, -235 on day 28 in tank 2, and -200 on day 21 in tank 3. The sediment became less201.00.50.00.51.01.52.02.53.03.54.00 Replicate 2a v 4 ., Day 0not available0 • Day 20•4 .7 Day 4a', Day 70 Day 14• ,nia^• • Day 210• Day 28---------0 --7• - - ----------- ------•--.2,• -44 't 0• 41 0v 04• is •- --------^0- ----Replicate 3Day 0not available• Day 2^ Day 4• Day 7O Day 14• Day 21• Day 28Eh (mV)210CNI1.00.50.00.51.01.52.02.53.03.5 -4.0NN NI^I0^00 01.0^NI^0^0^ 0^0^N7^7^OD V. 0^et COI^II^I^I^I^I^I^t^I.^I^I^1^I^I^I^5^I^J^I:CI.. • - 11Replicate 1Day 0not available• Day 2^ Day 4^ Day 7O Day 14• Day 21• Day 288 0N N N-6;t? 0 V70 • • 44Elr' ‘ ,4,•■k7^vsib4.N N■^I°O 8^cv^o^0^c.,^2^•0 0 0.^0^2N ^• 0^• 0^(NN^N 7^7^I I^.II.I ^I^1.1,1■I■1■1N01.00.5 -0.0 -0.5 -1.0 -1.5 -2.0 -2.5 -.3.0 -3.5 -4.0  C^o^o^0^0^ 0^0^0^0^0^0CO et 0 CD N 0^0^ 0^0^N LO 0 et CO NN^N^N O ea. 0^et- CO N^N^NI I I^I^7^I^I1^i^,^I^1^I^I i^,^I^,^i^,^I^.1 Figure 1.6. Reduction-oxidation time-series depth profiles of the replicate tanks for theControl Treatment in 1990. Values < +100 mV (dashed line) indicates reducingconditions. Measurements began at 0.5 cm above the sediment surface.1^1^1^1^1^.1^•^1^1-a_..p• C7 -v,.......-.^•^o rriV\ off,•pDay 21Day 28Replicate 1Day 0not available• Day 2^ Day 4^ Day 7^ Day 148NNNONN0NN mO1N^010o o 8N NO^o^N^ON .- mI^I^I^I0N0N0 0O0N0n 0m m0^0a 0N^NI It0^N^00^0▪^O^oI IDay 0Day 2Day 4Day 7Day 14Day 21Day 281.00.50.00.51.01.52.02.53.03.54.0Replicate 3o Day 0• Day 2V Day 4V Day 7^ Day 14• Day 21L Day 281.00.5 -0.0 -0.5 -1.0 -1.5 -2.0 -2.5 -3.0 -3.5 -4.0  I ^i^l^l^I^1^i•^i^i^i^•l^11.00.50.00.51.01.52.02.53.03.54.0Eh (mV)N^N^N^.^N^ V^OI I I ^I IFigure 1.7. Reduction-oxidation time-series depth profiles of the replicate tanks for TreatmentA in 1990, a low dosage of oxytetracycline. Values < +100 mV (dashed line)indicates reducing conditions. Measurements began at 0.5 cm above the sediment.22On 0 0 Ob ON ON ON OM OMreducing after day 21 in tanks 1 and 3, but remained highly reducing on day 28 in tank 2.In comparison to the Control treatment, the sediment surfaces of Treatment A did notbecome reducing as quickly and recovered earlier; the sediment remained reducing at depthfor a longer period of time and was also less reducing overall.Redox potentials on day 0 were^+200 mV (Fig. 1.8), much greater than in thetanks of the Control and Treatment A, with reducing conditions reached by 1 cm depth.After the addition of fish waste and a higher dosage of oxytetracycline, the redox profile ofthe sediments closely resembled the profiles in Treatment A. Overall, the sedimentsprogressed to less reducing conditions, with the lowest surface Eh of -120 mV on day 7 intank 1, and -120 and -170 mV on day 4 in tanks 2 and 3 respectively. This was also whenthe overall redox of the sediments was the lowest, at -260 mV in tank 3. Although thesediments in this treatment remained reducing for a longer period, and were less reducingoverall than in the Control treatment (similar to Treatment A), the sediment surfacerecovered sufficiently to become oxidized, which never occurred in either the Control orTreatment A.1.4.4 DISSOLVED OXYGENChanges in the levels of dissolved oxygen of seawater were not significantlydifferent among the three treatments (ANOVA and Tukey test, P > 0.05, Fig. 1.9).Dissolved oxygen levels were initially 97% saturation of seawater, and declined rapidly byday 7, to fluctuate between 88% to 78% saturation of seawater for the remainder of thestudy. Analyzed on a daily basis, only the decrease in dissolved oxygen on day 2 for theControl was significantly different from Treatments A and B (Tukey test, P <0.05).1.4.5 CARBON:NITROGEN RATIOSThe carbon:nitrogen (C:N) ratios of all treatments varied in range from 13.5 to 19(Fig. 1.10). The C:N ratios decreased in value from 17-24 to 13.5-15 by day 2 after2300 00^0OO00 0CON 8I^I^IIII^I^I^II^IFt'NI^I/^I01.00.5 -0.0 -U0.5 -a. 1 0 -Ncz^1.5 -TT:^2.0 -g 2.5 -.brill'^3.0 -3.5 -4.0 ^0^0 0 0000*^0^l0^N^0^0^0^0N^N N CO * 0^* 03I ^7^7^1^1I^I.^I^I^1^I^,^I^,^I,^1^■^I^i^,i^,^i^,^I0^0^0^0^0 0CON^N^N^■,)Eh (mV)0N1.00.5 -0.0 -•-?- 0.5 --0-1 1.0'0^1.5 -4i:^2.0 -2.5 -'Vkl)^3.0 -v)3.5 -4.0 ^0 0 0 0*^0^N 0.-^CO^g^0^CO0I^I^I^I^I^I^I^1^I^I^I^I^I 003I0 0 0 0 00^CON0tet1.00.50.00.5 -Q. 1' 0 -A^1.5 -2.0 -2.50)^3.0 -3.5 -4.0 ^Replicate 20 Day 0• Day 2Day 4Day 7^ Day 14■ Day 21o Day 28N24Figure 1.8. Reduction-oxidation time-series depth profiles of the replicate tanks for TreatmentB in 1990, a high dosage of oxytetracycline. Values < +100 mV (dashed line)indicates reducing conditions. Measurements began at 0.5 cm above the sediment.95 -90--85 :80 :75 --70 :65 --:60 ---55 -50  1 , 11 00o Control• Treatment Av Treatment B1^I^I^I^I^1^1'1'1^1'1'1'10^2^4 6 8 10 12 14 16 18 20 22 24 26 28 30Time (Days)25Figure 1.9. Comparison of the dissolved oxygen levels in seawater between the Control (fishwaste added), Treatment A (fish waste and 2.53 g of OTC added), and Treatment B(fish waste and 10.13 g of OTC added). Plotted values are the Mean ± 1 StandardDeviation, n=3.38 —36 —34c 32o"-«-1^30czX 28co 26 —tali0s. 24 —...)..-‘^22 —o 20,c)s.^18 —ccst.) 16 —14 —12 —10 —8 I^1^1^1^1'^1'^1'^1'^1^1^1^1^1^f^1^10^2^4^6^8 10 12 14 16 18 20 22 24 26 28 30Time (Days)26o Control• Treatment Av Treatment BFigure 1.10. Carbon:nitrogen ratios of the Control (fish waste added), Treatment A (fish wasteand 2.53 g of OTC added), and Treatment B (fish waste and 10.13 g of OTC added).Plotted values are the Mean ± 1 Standard Deviation, n=3.addition of fish waste, before increasing slightly by day 28. None of the treatments weresignificantly different from each other (Kolmogorov-Smirnov test, P < 0.05). Forcomparison to previous studies, the total carbon content of the sediments is presented aswell (Table 1), and shows that the carbon content of the sediments increased after theaddition of the fish waste.1.4.6 OXYTETRACYCLINELevels of OTC declined quickly after addition to the sediments, from 260 ppm and1080 ppm to < 1 ppm and 120 ppm by day 2, in Treatments A and B respectively (Fig.1.11). OTC levels continued to decline slowly in both treatments for the remainder of thestudy period. After day 14, trace levels of OTC remained in the sediments of Treatment A,but could no longer be detected. OTC concentrations were significantly different betweenthe two treatments (P >0.05, Kolmogorov-Smirnov test).1.4.7 MICROBIAL ABUNDANCEFollowing addition of fish waste to the Control tanks, bacterial abundance decreasedfrom 9x109 cells. g-1 to 5.5x109 cells. g-1 sediment on day 4, before increasing to 22x10 9cells•g-1 on day 21 and declining again on the final day of the experiment (Fig. 1.12). Arelatively low dosage of oxytetracycline with the fish waste to Treatment A caused an initialrise in bacterial numbers to 12x10 9 •g-1 sediment by day 2, followed by an immediatedecrease in abundance on day 4 as in the Control tanks. Changes in bacterial abundancewere similar to the Control treatment, increasing to the highest level on day 21 (to20x109 .g-1 ) and declining again. The change in bacterial abundance between Treatment Awas not significantly different from the Control (P >0.05, Tukey test). After the highdosage of OTC in Treatment B, bacterial cells increased rapidly to peak at 26x10 9 •g-1sediment on day 4, and dropping to about 14x10 9 • -1 on day 7, remaining near this level2712001100281000a)900300c.)250200a., 1501 0050 ---0 -o Treatment A• Treatment B1!Ir^ I^'^I^'^I^1^'^I0^2^4^6 8 10 12 14 16 18 20 22 24 26 28 30Time (Days)Figure 1.11. The concentrations of oxytetracycline (ppm) for Treatment A (2.53 g of OTC)and Treatment B (10.13 g of OTC). Plotted values are the Mean ± 1 StandardDeviation, n=3.-o Control• Treatment Av Treatment B--40°-)sa)cno3020U°' 100 I^I^I^I^I^I^1^1^I^I^I^I^1,10^2^4^6^8 10 12 14 16 18 20 22 24 26 28 30Time (Days)29Figure 1.12. Bacterial abundance (cells x 109 .g-1 sediment dry weight) between the Control(fish waste added), Treatment A (fish waste and 2.53 g of OTC added), and TreatmentB (fish waste and 10.13 g of OTC added). Plotted values are the Mean ± 1 StandardDeviation, n=3.for the remainder of the study. The high dosage of OTC resulted in significant changes inbacterial abundance between Treatment A and the Control treatment (P <0.05, Tukey test).Bacterial were initially small, coccoid and short rod shaped cells. Chains composedof larger rods occurred during the first several days of the experiment, but were no longernoticeable by the end of the experimental period.1.4.8 PROTOZOAN ABUNDANCEAlthough the PercollTM-sorbital gel was vacuumed off, the remaining gel on thefilter still coagulated with the 2% formalin-seawater preservative solution, rendering allsamples unidentifiable.1.5 DISCUSSION1.5.1 REDOX MEASUREMENTSThe development of reducing conditions in the sediments by the addition of fishwaste was clearly demonstrated by the changes in the reduction-oxidation profiles (Eh) inthe tanks of all 3 treatments. The redox potentials < +100 mV at the surface and at depth,concurrent with the shift in the RPD layer to the surface, were indicative of reducingconditions being rapidly reached (Fig. 1.6 - 1.8). As oxygen is depleted within a short timein the surface zones of the sediments, a series of anaerobic oxidation processes occur, in thegeneral sequence of nitrate reduction, sulphate reduction and methane formation (Zehnder& Stumm 1988, Legal & Fauque 1988). The energy released during the redox reactionscan be measured (Eh), and the intensity of the reducing conditions can thus be related to theredox potential (Fig. 1.13). Microbially mediated reduction reactions thus occurred belowredox potentials of -100 mV to -200 mV (e.g. Battersby & Brown 1982, Stanley et al.1981), in several genera of sulphate reducing bacteria (e.g. Desulfovibrio,Desulfotomaculum, Desulfobacter, Desulfonema; Postgate 1979, Nedwell 1982, Parsons et30[CH20]C O2[H]-420 H2H*CH4CO HSFe"Fe3*s of -ENH4NO;( Mn2.Mn`'7 0 ( N2810^NO (- H2°L 0231Redox Potential^-AG°(pH 7)[mV^ [kJ/mot el- 600-20------:'` NADH -400\ D -320 FADHred.^+200Cyt ox.-1 +230'red.Cyt +600+ 800NA D-200FAD-'red.Cyt b )+30 ox .0+20+40+100+80+60+120+1000Figure 1.13. Electron free energy diagram for the biologically mediated redox sequences ofthe degradation of organic carbon, CH2O (acting as the electron donor). Note therelationship between the change in energy (G°), redox processes and the Eh (mV)(from Zehnder & Stumm 1988).al. 1984b, Widdell 1988, Fauque et al. 1991, Bak & Pfennig 1991). This was reinforcedby visual inspection of the sediment cores, which revealed that the upper 1 to 2 cm of thesediment surfaces were black, becoming grey-brown at depth. The black colouration waslikely due to the presence of ferrous sulphide (FeS), and the FeOOH-FeS zone is associatedwith the brown and black colour change in the sediments (Jorgensen 1989, Jorgensen &Revsbech 1989). The values of the redox potentials measured in this study were within theranges found in sediments beneath fish farms in a Scottish sea loch by Brown et al. (1987)(as low as -185 mV at 4 cm depth), and down to -200 mV within 2 cm of the sedimentsurface by Lumb & Fowler (1989). The sediment was less reducing in tank 1 of theControl treatment, compared to tanks 2 and 3, due to the inadvertent mixing of thesediments and fish waste at the time of addition (day 0). This mixing aerated the sedimentand created a less reducing environment.1.5.2 MICROBIAL MATThe white microbial mat that developed on the sediment surfaces was composed ofthe non-phototrophic sulphide oxidizing filamentous bacteria, primarily Beggiatoa spp.(e.g. Jorgensen & Fenchel 1974, Jorgensen 1977a, Nelson & Castenholz 1981, Jorgensen& Revsbech 1983 & 1989, Nelson et al. 1986, Kuenen 1989, Lumb & Fowler 1989)oxidizing the large amounts of H2S produced in the sediments during sulphate reduction.The presence of anoxygenic phototrophic bacteria is doubtful, as light was excluded fromthe tanks by dark covers except when sampling (Caumette 1989). The rapid developmentof the microbial mat was another indication of how quickly the sediments had becomeanoxic and reducing. The decline of the white microbial mat on day 7 was caused bywiping down the tank walls on day 4, which resulted in mild aeration of the sedimentsurfaces. But anoxic and reducing conditions redeveloped quickly as the microbial layerrapidly reformed.32Based solely on morphology, the small cocci and short rod shaped bacterial cells thatwere prevalent throughout the study are characteristic of sulphate reducing bacteria (Widdel1988, Fauque et al. 1991), while the larger rods that comprised chain and colonies arecharacteristic of the chemoautotrophic white sulphur bacteria, Beggiatoa spp. (Nelson &Castenholz 1981, Jorgensen & Revsbech 1983).Benthic infauna surfaced from the sediments after day 2 to escape sulphide poisoningin the sediments (Jorgensen 1977b & 1980), and had presumably returned to the benthos byday 4, although burrows were not observed. Previous observations from field studies havenoted that the presence of the white microbial mat did not exclude all epibenthic meio- andmacrofauna, with sea anemones and geoducks present beneath fish pens (Aquametrix 1988).Under extremely reducing conditions, even these organisms will eventually die (Jorgensen1980).1.5.3 DISSOLVED OXYGENAll three treatments exhibited the same pattern of changes in the dissolved oxygen ofseawater due to the BOD (Fig. 1.9), which has been consistently found for bottom waterfrom sediments loaded with fish waste (Hall & Holby 1986, Brown et al. 1987, Gowen &Bradbury 1987, Parsons et al. 1990). There was a significant decrease in the dissolvedoxygen levels of the Control treatment on day 2 compared to Treatments A and B. Thismay be attributed to the greater reducing conditions that were reached more rapidly by thesediments of the Control treatment, perhaps as the microbial population was not inhibited insome way by the OTC as in Treatments A and B.1.5.4 CARBON:NITROGEN RATIOSThe carbon content of the sediments prior to addition of fish waste varied between 1to 2% (Table 1). This was within the range expected for estuarine sediments alongVancouver Harbour (Johnson 1990). Enrichment with fish waste increased carbon content,33TABLE 1. Total Carbon content (percent) of the sediments in the Control, Treatment Aand Treatment B. Values presented are the Mean % ± 1 Standard Deviation (SD).Day Control Treatment A Treatment BMean % SD Mean % SD Mean % SD0 1.529 0.255 1.517 0.672 1.849 0.9092 1.522 0.380 1.749 0.801 1.956 0.3824 1.808 0.372 2.689 1.218 2.112 0.3307 2.075 0.272 1.543 0.449 1.839 0.29514 1.523 0.239 2.118 0.619 1.213 0.43721 1.596 0.276 1.806 0.704 1.526 0.29128 1.555 0.126 1.646 0.252 2.282 1.11334with values in the range of 1.5 to 2.7%. In comparison, the carbon composition ofsediments found by Brown et al. (1987) directly beneath fish farms was as high as 9.35%,but averaged 4% composition 15-1400 m away from the farm sites. However, thesediments of this study were characteristically low in organic matter due to the quartz originof the sediments and tidal action (Carolyn Jones, pers. comm.). Comparisons of the totalcarbon content of the sediments between this study and others should be made cautiously(e.g. Brown et al. 1987). In field studies, phytodetritus would have accumulated in thesediments throughout the year, in addition to the fish waste released from the cultureoperations (e.g. Aquametrix 1988). Only fish waste was added to the sediments in thisstudy, and insufficient phytodetritus would have accumulated in the microcosm tanks afteronly a few weeks for the carbon content of these sediments to resemble conditions in thefield.The decrease in the C:N ratios of all treatments reflected the lower total carbon andnitrogen of the fish waste (Fig. 1.10). OTC did not affect the degradation of fish waste, asthe C:N ratios did not differ between treatments. The C:N ratios were similar to thosefound by Samuelsen et al. (1988), of 10 and 14 at fish farm sites in Norway. The C:Nratios at the end of the study remained approximately the same as when the fish waste wasinitially added, suggesting that any changes in C:N ratios would require a longer durationthan 4 weeks to become noticeable due to biological degradation processes by benthicfauna. Fish wastes are readily degradable by meio- and macrofauna (Weston 1986 & 1990,Tsutsumi 1987), and are thus assumed not to be refractory.1.5.5 OXYTETRACYCLINEThe sediments with oxytetracycline added to the fish waste appeared to experienceless reducing conditions (Fig. 1.6 - 1.8), although the sediments remained reducing overallfor a greater period of time at depth, with oxidizing conditions reached at the sedimentsurfaces by the end of this study. Sediments with fish waste and OTC have been known to35remain reducing for a longer period of time (Samuelsen et al. 1988, Jacobsen & Berglind1988, Bjorklund et al. 1990), but this differential effect at the sediment surface and at depthhas not been previously recorded. This increase in the Eh of the Control treatment to a lessreducing environment suggests that there was a shift in the microbially mediateddegradation processes, away from sulphate reduction.The large decreases of OTC in Treatments A and B (Fig. 1.11), were due partly tothe OTC being washed out of the fish waste during addition to the sediments, rather thanbeing degraded by bacterial action (Samuelsen et al. 1988, Samuelsen 1989, Jacobsen &Berglind 1988, Husevag et al. 1991). But as discussed in Appendix 2, the oxytetracyclinewas also underestimated in the sediments because of problems experienced with themethodology. Even though analysis of the OTC samples was incomplete, the smallconcentrations of OTC would likely have persisted for a long period of time,^419 daysunder conditions of sediment anoxia and low turbulence in the water column (Bjorklund etal. 1990). Samuelsen et al. (1988) theorized that the development of the white sulphuroxidizing bacterial plaques on the sediment surfaces may have decreased diffusion of theoxytetracycline into the water column.As oxytetracycline is a hydrophilic compound with an octanol/water partitioncoefficient of 0.025 (Chopra 1985), it should tend not be incorporated into fish waste, andaccumulate in the sediments (unlike the behaviour of hydrophobic organic pollutants,review by Jaffe 1991). This is offset by the preferred accumulation of organic pollutants inmuddy and fine sediments that are high in organic carbon (Hiraizumi et al. 1979), whilehigh amounts of dissolved organic matter (DOM) enhances the solubility of pollutants in theaquatic environment (Jaffe 1991). This suggests that the accumulation and persistance ofOTC in the sediments beneath a fish farm would occur under high rates of sedimentation.Chemical degradation of OTC by photochemical decomposition can occur (Oka et al.1989), but would be unlikely at the water depths of the sediments beneath fish farms wherelight is greatly reduced. Redox reactions and biodegradation by microorganisms of OTC36under highly reducing conditions are not known (Lunestad & Goksoyr 1990), and thedisappearance of OTC over time due to diffusion and washing out from the sediments iscurrently the accepted explanation.1.5.6 MICROBIAL ABUNDANCETotal bacterial counts were within the range found for sediments beneath fish farms,0.3 - 1.4x109 •g-1 (Enger et al. 1989), 1.4 - 6.3x109 •g-1 (Samuelsen et al. 1988), 1.4 -2.2x109 .g-1 (Torsvik et al. 1988), and 3.7 - 10.1x109 .g-1 (Carr & Goulder 1990). Totalcell counts in this study were to 2 to 10 times higher (e.g. 26x109 .g-1 in Treatment B) thanin the previous studies which had not utilized ultrasonication to release all the bacteriaattached to the sediment. The large standard deviations present (Fig. 1.12) were due to acombination of patchiness, sampling and subsampling variability that often occurs withbenthic analysis (Venrick 1971, Montagna 1982).From the strongly reducing conditions present in the sediments (Fig. 1.6 - 1.8), thelarge increases in bacterial abundance on Days 2 and 4 of Treatment B (Fig. 1.12) werelikely due to the sulphate (and nitrate) reducing fractions of the microbial population(Battersby & Brown 1982). Increases in the abundance and activity of sulphate reducingbacteria have been associated with increases in redox potentials (Jones 1979, Aller & Yingst1980, Jorgensen 1989). Sulphate reduction in marine benthos can be the principal pathwaycontrolling the degradation of organic matter (e.g. Jorgensen & Fenchel 1974, Pearson1982, Battersby & Brown 1982, Thode-Anderson & Jorgensen 1989, Sampou & Oviatt1991). Stanley et al. (1981) examined sediments loaded with organic waste from a pulpmill in Scotland, and concluded that while total aerobic/anaerobic heterotrophic bacteria andnitrate reducing bacterial cells did not increase in abundance, sulphate reducing bacteriawere 2 to 3 times higher than in unpolluted regions. Samuelsen et al. (1988) also foundsulphate reducing bacteria to be 2 to 4 orders of magnitude more at fish farm sites (104 and106), compared to unpolluted sites (10 2).37The cell counts in all of the treatments (Fig. 1.12) did not begin to converge untilthe sediments began to become more oxidized by day 28. The overall increase in bacterialabundance on day 21 in the Control and Treatment A was due mainly to Beggiatoa spp., asday 21 corresponded to the maximal development of the white microbial mat.1.5.7 POTENTIAL EFFECTS OF OTC ON MICROFAUNAThe smaller dose of OTC (1.27 •' 1 fish waste) resulted in a small increase inbacterial cell counts on day 2 in Treatment A (12x109 bacteria•g-1 sediment, Fig. 1.11),while the larger dose of OTC (5.07 g•1 -1 fish waste) resulted in a large increase in cellcounts on day 2 and 4 in Treatment B (to 26x109 bacteria•g4 sediment). Samuelsen et al.(1988) also found total bacterial counts to be 1.4x10 7 • -1 at a fish farm site abandoned fora year and without chemotherapy for 18 months, and 1.9x107•g-1 at another fish farm siteduring a disease outbreak treated with antibiotics. These cell counts were higher incomparison to a control site where the bacterial counts were only 4.3x10 5 .g-1 . The largeincreases in microbial abundance during the first few days of the experiment in TreatmentB, the slower development of reducing conditions within the sediments and of the whitemicrobial mat (Section 1.4.2), and the persistence of the white microbial mat at the end ofthe experiment can be explained by the inhibition of benthic bacterivorous protozoans andthe development of OTC resistance. Although only bacterial abundance and the redox stateof the sediments was conclusively established in this preliminary investigation, thepossibility of OTC resistance developing and the inhibition of the microbial and protozoancommunities (and interactions between these microfauna) should also be considered. Thecell counts were higher in Treatment B with the higher dosage of OTC, than in TreatmentA and the lower dosage of OTC, despite the presence of the white microbial mat in theControl treatment and Treatment A. This would suggest that the increases in cell numberswere in another fraction of the anaerobic bacterial community, perhaps due to a38combination of OTC resistance developing and the potential inhibition of protozoangrazing.Development of OTC ResistanceThe large increases in bacterial abundance in Treatment B, the overall increase inbacterial abundance despite the presence of OTC in Treatment A, and the stable bacterialpopulation from day 6 to 28 in Treatment B suggested that there was a potential for OTCresistant strains to have developed among the microbial population. The frequent and longterm use of OTC, has resulted in a majority of bacteria being resistant to this antibiotic(Levy 1984 & 1988), either from the sediments or in animals near fish farms (Aoki et al.1974 & 1981, Austin 1985, Bjorklund et al. 1990 & 1991, Grave et al. 1990, Husevag etal. 1991, Nygaard et al. 1992). The bacteria carry and transfer this resistance throughplasmids (e.g. Chopra 1985, Terzaghi & O'Hara 1990, Al-Masaudi & Russell 1991),assumed to be mainly through conjugation and possibly transduction in aquaticenvironments (e.g. Trevors et al. 1987, Fernandez-Astorga et al. 1992). Torsvik et al.(1988) isolated 2 distinct strains of antibiotic resistant bacteria from sediments beneath afish farm, with one strain susceptible to low concentrations of OTC (2.4 tig•m1-1 ), and theother resistant to high concentrations of OTC (300 Ag•m1-1). This double peak ofresistance to OTC has also been reported in Aeromonas salmonicida (causing furunculosis),at 1.26 and 80 Ag•m1 -1 OTC (Inglis & Richards 1991). There is thus a possibility for themicrobial community to have developed resistance in Treatments A and B.The minimum inhibitory concentration of OTC required to elicit an antibioticresistant response in the microbial community was not measured. After the majority of theadded OTC in the fish waste was washed out after addition to the sediments (Appendix B),the presence and activity of the OTC was probably diminished even further by complexformation with magnesium and calcium ions in seawater (Lunestad & Goksoyr 1990). Theactive form of OTC in the sediments and seawater can be as little as 5 %, with the OTC39bound mostly in the form of an OTC-magnesium complex, and some as an OTC-calciumcomplex. The microbial population could have been inhibited by the OTC or by thegrazing pressures exerted by the bacterivorous protozoans.Inhibition of Protozoan PredationThe loss of all benthic ciliate samples precluded examination of protozoanabundance. However, the oscillations in bacterial abundance (Fig. 1.12) are characteristicof grazing effects by bactivorous protozoa (e.g. Tietjen 1980, Berninger et al. 1991, Weisse& Scheffel-Moser 1991). Protozoa are abundant in sewage (e.g. Pike & Curds 1971,Kinner & Curds 1987, Manru et al. 1988, Kosciuszko & Prajer 1990, Esteban et al. 1991),and inhibition of protozoa by antibiotics has been shown to increase total bacterialabundance (Mallory et al. 1983, Wiggins et al. 1987, Wiggins & Alexander 1988). Thepotential effects of OTC on protozoans and their grazing pressure should also beconsidered.The effects of OTC on benthic ciliates (Class Kaiyorelictea) indicated a chemotacticresponse (personal observations), with the ciliates attempting to move away from theantibiotic. At high dosages of OTC (e.g. 4 drops of 0.25 M OTC, or 1.24x104 mg OTC.1 -1 seawater) these highly contractile ciliates exhibited a toxic reaction, swelling and bursting.In Treatment B, a sufficiently high and active amount of OTC could have been initiallypresent to have inhibited the bactivorous ciliate population in the sediments. This couldhave been either by an immediate toxic effect, or a temporary cessation of proliferationfollowed by resumed growth at a decreased rate (Nilsson 1989). Tremaine & Mills (1987)determined that protozoans were inhibited by as little as 25 mg•1 -1 cycloheximide , withlarge protozoans inhibited within 24 hrs, and smaller protozoans active until 48 hrs. Thedosages of OTC used in the Treatment A and B were initially 1.27 g•1-1 and 5.07 g•1-1 fishwaste, respectively, perhaps sufficient to inhibit the protozoan populations. The toxic effect40of OTC on protozoans resulting in the organism bursting is not known, and requires furtherresearch.OTC is able to both enter eukaryotic cells and act on the mitochondria, and byimplication OTC is thus able to enter protozoans (eukaryotes) as well. Oxytetracyclineinhibits protein synthesis at the 70S ribosome (Levy 1984, Chopra 1985), and the mode ofaction in the mitochondria of ciliates may be analagous to the action of chloramphenicol(CAP, Chopra 1985) which also inhibits protein synthesis at the 70S ribosome. Studies onthe ciliate Tetrahymena have shown that the action of CAP is confined to the mitochondria(Nilsson 1989), resulting in a decrease of mitochondrial DNA and volume, inhibition inprotein synthesis and the synthesis of mitochondrial ribosomes. Recovery can be rapid, 6hrs after CAP is removed, and different strains of Tetrahymena can tolerate higher levels ofCAP and achieve cell doublings (Nilsson 1989). The mitochondrion in protozoans use 02as a terminal electron acceptor to oxidize reducing compounds passing down the electrontransport chain to produce energy (Finlay 1990). However, protozoans in reducingsediments that rely on anaerobic metabolic pathways for the release of energy do not havemitochondria, but contain endo- and ectosymbiotic bacteria and hydrogenosomes (Fenchelet al. 1977). The symbiotic bacteria are methanogens, utilizing the H2 gas produced fromthe catabolism of pyruvate in the hydrogenosomes to produce methane that is expelled fromthe ciliate. The hydrogenosomes function like mitochondria, and yield 02, CO2, acetateand energy used by the protozoan (Finlay 1990). The action of OTC on the symbioticbacteria and hydrogenosomes of anaerobic ciliates is not known, and additional research isrequired in this area.Although most of the OTC was rapidly washed out of the sediments (with lowamounts of OTC persisting in the sediments, Fig. 1.11), the effects of the higherconcentrations of OTC on the protozoan community could have extended over the first fewdays (day 2 and 4). During this period of ciliate inhibition, the microbial community wouldhave responded rapidly by increasing in abundance. After day 4 the lower amounts of OTC41present (50 ppm) may not have been enough to inhibit the bactivorous ciliate population anyfurther, resulting in the bacterial community being grazed down. The lower dosage of OTCin Treatment A may have only been enough to have had a small inhibitory effect on anyciliates present, which would account for the small rise in bacterial abundance on day 2(and with the bacteria displaying OTC resistance). Similarly, Pratt & Cairns (1985) alsofound that a toxicant (cadmium) reduced numbers of photosynthetic and bactivorous-detritivores ciliates at low concentrations, and high toxicant levels eliminated photosyntheticspecies and reduced bactivorous-detritivore protozoan populations by one-half.The potential effects of OTC within and between the microfaunal communities couldhave occurred as follows. Protozoan grazing on the sulphate reducing bacterial communitycould have been inhibited by a high dose of OTC. Concurrently, the greater increase inbacterial abundance indicated that growth of the anaerobic bacterial population was notinhibited by the OTC, perhaps due to the development of OTC resistance. Althoughbacterial abundance did increase, the slower increase in the reducing conditions of thesediments treated with OTC (Treatment B), along with the slower development of the whitemicrobial mat at the end of the study showed that there was initially a lack of H2S availablefrom the anaerobic bacterial community (i.e. sulphate reduction was not the dominantprocess in the degradation of fish waste). If some sulphate reduction had occurred, the H2Scould have been bound in the sediments as FeS and FeS 2 (Jorgensen et al. 1990, Bak &Pfennig 1991, Kristensen et al. 1991) making it unavailable to the sulphur oxidizingbacteria, or the sulphur oxidizing bacteria may not have been able to use the H2S due topossible inhibition by OTC. The increase in bacterial abundance in Treatment B, but lackof available H2S for the sulphur oxidizing bacteria, suggests that while the sulphatereducing bacteria increased in abundance the metabolic rates of the bacterial communitywere inhibited, again through the lack of grazing pressure. Grazing by protozoans on themicrobial community can stimulate bacterial production (Kemp 1990), which could have42occurred in the Control treatment, as indicated by the highly reducing conditions that werequickly reached and the rapid appearance of the white microbial mat. As well, anyincreases in the bacterial population could have been grazed down and never observed.Bacterial metabolic rates may not always be coupled to growth processes, and the increasein cell numbers on day 2 and 4 of Treatment B may not necessarily mean an increase inrates of bacterial activity (Hanson 1980, Pearson 1982). When the OTC washed out of thefish waste and sediments after the first day (and with the recovery of the protozoancommunity due to the potential inhibitory effects of OTC), along with the possibility ofreduced grazing pressure on the sulphate reducing population but stimulated metabolicrates, sufficient H2S could have been produced for the sulphur oxidizing bacterialcommunity to have become established, and with increased redox conditions in thesediments. However, these theories are only speculative, and cannot be conclusivelydemonstrated due to the lack of data on ciliate abundance and the inadequacy of studies onthe effects of OTC on protozoan physiology and OTC resistance among the microbialcommunity in this study.The greater reducing conditions at the end of the study, with the continued presenceof the white microbial mat implied that H2S was still available from sulphate reduction(sulphate reducing bacteria remained active). In the Control treatment, the disappearance ofthe white microbial mat indicates that sulphate reduction was no longer the primary route inthe degradation of fish waste (H2S was no longer available), and perhaps there was asubsequent switch to methanogenesis. In marine systems with high rates of organic matterdeposition, sulphate can be depleted to the extent that methanogenesis becomes a significantpathway in anaerobic carbon catabolism (Capone & Kiene 1988). The composition ofsedimentary gas at fish farms has been found to be composed of mainly methane (64%) andH2S (22%) (Kaspar et al. 1988). This is confirmed by Aure & Stigebrandt (1990) whoestimated that in sediments strongly impacted by fish waste and experiencing reducingconditions from -200 to -320 mV, up to 70% of the biochemical decomposition should be43by methanogenesis, and the remaining 30% through sulphate reduction. Methaneproduction would require a sedimentation rate greater than 8 g C•m 2 •d-1 beneath net pens(Iwama 1990). Nitrate reduction in the sediment is negligible under these conditions(Kaspar et al. 1990). Heterotrophic growth and activity would need to be measured inorder to confirm these conjectures.1.5.8 ENVIRONMENTAL CONCERNSThe impact of oxytetracycline on the microbenthos may result in two potentialproblems, the enhancement of disease transmission in the microbial community, anddecrease in the mineralization rate of detrital carbon.Disease TransmissionThe persistence of low amounts of OTC in sediments ( 419 days), may cause along-term antibiotic resistance response (Chopra 1985, Austin 1985), making it impossibleto control bacterial diseases of cultured fish (Toranzo et al. 1984). Antibiotic resistance bybacteria has often been correlated to fish farms undergoing antimicrobial therapy, andresistance to the tetracycline antibiotics reached 69.1% among bacterial strains isolated fromcatfish farms with cross-resistance to the tetracyclines, chlorapmphenicol, kanamycin,ampicillin and nitrofurantoin commonly occurring (McPhearson et al. 1991). Nygaard etal. (1992) found that among bacteria isolated from sediments beneath fish farms, OTCresistance increased from 5% to 16% after 12 months. The presence of other fishpathogenic bacteria (see Sindermann 1984, Egidius 1984, Austin & Allen-Austin 1985,Cahill 1990) in the sediments also raises the possibility of horizontal gene transfer ofantibiotic resistance between different pathogenic bacteria (Husevag 1991). With fishpathogenic bacteria remaining viable in fish farm sediments for long periods of time (Vibriosalmonicida causing vibriosis, Enger et al. 1989) and developing resistance to antibiotics(Husevag et al. 1991), there is a possibility for the reinfection of fish through degassing ofthe sediments returning the bacterium into the water column (Levings 1992). There is also44a potential for the development of cross resistance between oxytetracycline, oxalinic acidand furazolidone in bacteria, creating a greater risk for passing this resistance on betweendifferent bacterium (Trevors et al. 1986, Lunestad & Goksoyr 1990, Inglis & Richards1991, Nygaard et al. 1992). Coliform bacteria such as Escherichia coli and Klebsiella spp.are present at fish culture sites and sewage contaminated waters (McPhearson et al. 1991,Qureshi & Qureshi 1992), with 10% of the coliform strains resistant to 8 types ofantibiotics. Increased cross resistance in a greater portion of human pathogenic bacteriacould result with increased usage of antibiotics, increasing a risk for human infections thatcannot be treated by the usual antibiotics. The inability to treat bacterial diseases wouldalso greatly curb the potential for cross culturing species (e.g. salmon and shellfish culture),particularly in the presence of human pathogenic bacteria. Preliminary results indicate thatthere is an overall increase in antibiotic resistance of bacterial isolates (Family Vibrionaceathat cause gastroenteritis, septicemia, meningitis and skin/eye/ear infections in humans),from cultured fish and surrounding waters of B.C. coastal fish farms (Dr. M.T. Kelly,unpublished results). These bacterial isolates exhibited cross-resistance to the drugsampicillin, tetracycline and cotrimoxazole. There was also increased resistance by V.anguillarum to ampicillin, the causative agent of vibriosis in farmed salmon. Harvesting ofwild fauna (fish and shellfish) around fish farms is another possible route for humaninfections. The consumption of raw cultured and wild shellfish which concentrate V.vulnificus by filtration (e.g. Kelly & Dinuzzo 1985, Kelly & Stroh 1988), is a commoncause of gastroenteritis (DuPont 1986). A large outbreak of gastroenteritis in Singapore in1979 was traced to shellfish imported from the Phillipines, where Manila Bay is one of themajor aquaculture sites for shellfish but is also one of the most polluted water bodies inSoutheast Asia (Eng et al. 1989). The presence of antibiotics and resistant bacteria havebeen isolated from wild fish and shellfish which were feeding near the culture sites duringmedicated feed treatments (Bjorklund et al. 1990, Samuelsen et al. 1992).45The experimental evidence of Toranzo et al. (1984), suggests that plasmid genetransfer between fish and the aquatic environment is possible. They determined antibioticresistance patterns and resistant plasmids of bacteria isolated from the water in hatcherytanks and the skin of rainbow trout. Of 170 bacterial isolates (belonging to 8 bacterialgroups) from the cultured fish, 87.6% of the isolates were resistant to at least 1 drug, andthe majority of these strains (85.2%) were multiresistant. The plasmid coding for resistanceto chloramphenicol was found in Vibrio spp., Citrobacter spp., and Enterobacter spp., andresistance to sulfadiazine, tetracycline, nitrofurantoin and ampicillin were also cotransferredsome of the time. If the contaminated hatcheries were not adequately cleaned prior totransferring in uninfected rainbow trout, there would be a great potential for water-bornecontamination of the new fish. While there is concern about the potential inability to treathuman pathogenic diseases contracted from cultured and wild fish and shellfish due toantibiotic resistance through the increased use of drugs in aquaculture, this has never beensubstantiated in the natural environment.Mineralization of Detrital CarbonThe addition of antibiotics to the microbenthos could affect the microbial andbactivorous protozoan populations, which in turn would affect the mineralization of detritalcarbon. Providing that the microbial population displays resistance to the antibiotics used,elimination of the protozoan population could allow the bacteria to rapidly multiply andincrease the degradation of organic compounds (Wiggins et al. 1987). However, the resultsfrom this study appeared to indicate that the presence of protozoans may have actuallystimulated bacterial growth and activity, while bacteria increased in the absence ofprotozoans but probably not in activity (Section 1.5.7). From feeding experiments withisopods, Smock & Harlowe (1983) showed that the nutritional quality of detritus wasdecreased when microbial growth was inhibited by antibiotics. Inhibition of bacterivorousprotozoans may be undesirable because of their role in nutrient cycling (Lighthart 1969,46Barsdate et al. 1974, Stout 1981, Kemp 1990). Benthic protozoan grazing can contributeindirectly to the mineralization of detritus, possibly by facilitating nutrient availability andthe rate of nutrient turnover by lowering bacterial biomass (Pratt & Cairns 1985, Wiggins& Alexander 1988). This could in turn stimulate the growth rate of bacteria and increasethe uptake of nonlimiting nutrients (Stout 1980). The half-life of E. coli in sewage wastetreatment systems have been shown to decrease from 16 hrs to 1.8 hrs, while the uptake ofdissolved phosphorous was also found to be 4 times greater by bacteria being grazed on byprotozoans (Stout 1980). This was attributed to a higher proportion of younger, morerapidly dividing bacterial cells, compared to a more static population. However, protozoanscan serve directly as a source of carbon in the flow of energy through the trophic web, byconcentrating nutrients within themselves and making themselves available to otherheterotrophs (e.g. Fenchel & Jorgensen 1977, Tenore & Coull 1980, Meyer-Reil & Faubel1980, Stout 1980, Lopez & Levinton 1987, Sanders 1987, Stoecker & Capuzzo 1990).The importance of the microfauna (including the meio- and macrofauna) in thedegradation of organic matter has been demonstrated (Small 1973, Curds 1973, Tietjen1980, Mann 1988, Plante et al. 1989, Toerien et al. 1990). The input of the microfauna tothe detrital food web may depend on the types of sediment, organic matter and trophic webspresent (e.g. Tenore et al. 1982, Rieper 1985, Hansen et al. 1987, Alongi 1988, Kemp1987 & 1988, Jumars et al. 1989, Kuuppo-Leinikki 1990, Riddle et al. 1990, Nilsson et al.1991). While the extent of the contribution of the microfauna in the degradation of organiccarbon and nutrient cycling in this study cannot be determined, the importance of the fishwaste in enhancing growth of the microbial community, and of the bactivorous protozoansas a potential grazing control is apparent.The addition of fish waste to the sediments created a highly anoxic and reducingenvironment within a day (this study). This was an indication of the consumption of freeoxygen from the sediments in the degradation of the fish waste, and a subsequent shift in47anaerobic microbial metabolism from manganese, nitrate and iron reduction to sulphatereduction. The OTC was rapidly washed out of the fish waste when added to thesediments, with small quantities of the antibiotic persisting until the end of the study.Although the dissolved oxygen content of seawater decreased due to BOD from the fishwaste, the fluctuations in the dissolved oxygen levels between treatments were notsignificantly different. Carbon:nitrogen ratios decreased with the addition of fish waste,reflecting the greater nitrogen to carbon content of the fish waste.Bacterial abundance peaked at 26x109 •g-1 sediment by day 4 in Treatment B, andwas significantly different from the Control and Treatment A. The rapid rise in cell countsduring the initial few days may have been due to the potential inhibition of the bactivorousprotozoan population and development of OTC resistance in the anaerobic bacterialcommunity. The amount of OTC in Treatment A was significantly less than in TreatmentB, thus affecting any grazing on the microbial community by ciliates to a comparativelylesser degree. While bacteria increased in the absence of protozoans due to the OTC,metabolic rates may not have been affected, potentially slowing mineralization of the fishwaste. While the presence of protozoans may have stimulated growth and activity in thesulphate reducing bacterial community, only the decrease of OTC and redox state of thesediments, and changes in bacterial abundance can be conclusively demonstrated. Thepossible effects of OTC on the microbial and protozoan communities must remainspeculative until further research has been completed.To this point, this thesis has dealt with only a portion of the microfaunalcommunity. The next chapter will present the results of a pollution experiment in which anattempt was made to simulate the sedimentation rates of fish waste and OTC beneath a fishfarm. Oscillations in the abundance of microfauna (ciliates, bacteria), diatoms, andmeiofauna (nematodes, copepods) are presented.482. THE Eli ECTS OF FISH WASTE ANDOXYTETRACYCLINE ON THE MICRO- AND MEIOFAUNA.2.1 INTRODUCTIONIntensive cultivation of fish, crustaceans and molluscs generates large amounts ofsolid wastes (uneaten feed and feces) that can accumulate on the sediment bottom (e.g.Weston 1986, Gowen & Bradbury 1987, Iwama 1991). Ecological assessment of theimpact of aquacultural activities on the benthos has been restricted in the past to themacrofauna or macrobenthic invertebrates (Weston 1986). (Macro referring to thoseorganisms which are retained on a 0.5 mm sieve mesh; Eleftheriou & Holme 1984). Thelarger size of the invertebrates allows them to be more easily observed and sampled.In zones of heavy organic enrichment beneath and immediately surrounding fishfarms, species diversity generally decreases, with an increase in biomass and abundancefrom opportunistic species (Weston 1986). The dominant macrofauna (or infauna, as theyburrow within the sediments) is commonly the small polychaete, Capitella capitata, from8x103 to over 25x103 individuals•m -2 (Mattson & Linden 1983, Tsutsumi 1987, Brown etal. 1987, Weston 1990, Aquametrix 1990 & 1992). However, one of the disadvantages tothe use of macrofauna (especially polychaetes) as a biological indicator of organic pollution,is the relatively long interval required to observe large changes in abundance. This is dueto the longer reproductive cycles of the polychaetes (from 4 to 6 weeks) and environmentalmonitoring can take several months to a year (Pearson et al. 1982, Aquametrix 1990 &1992, Tsutsumi 1990a & 1990b). Another drawback is that the sieve mesh sizes used inisolating macrofauna from sediments are often too large, and do not sample the smallerfraction of the macrofaunal community, missing the micro- and meiofauna (Bachelete1990). This leads to potentially inaccurate estimates of abundance, and erroneous reportsthat polychaetes are the only infauna present (Kaspar et al. 1985; isolating macrofauna from49sediments beneath a mussel farm using a 1.68 mm mesh sieve). Using a 0.5 mm meshsieve, Weston (1990) found that nematodes, and C. capitata comprised over 99% of thetotal number of individuals beneath a salmon farm in Puget Sound (Washington, U.S.A.).Biological monitoring with macrofauna thus requires long time intervals, which canhamper the identification of polluted sediments, and the rapidity of preventative orrestorative management measures (Weston 1986). These actions may involve periodicallyrotating culture sites, submersible mixers to disperse the fish waste, and the collection ofdead fish and fish waste by the use of funnels beneath net pens (Gowen & Bradbury 1987).Other methods include the use of extruded rather than pressed feeds, which remain morestable in water and sink slower allowing the fish a greater opportunity to ingest the feed,and more efficient feeding techniques such as hand feeding to minimize wastage (Seymour& Bergheim 1991). In addition, the inherent weakness in sampling the macrobenthos (i.e.sieving techniques that miss smaller organisms) assures that comprehensive and accuratestudies of the benthic community cannot be obtained. Sampling of the microfaunalcommunity (and perhaps meiofauna as well) can provide a solution to these limitations.While identifying the microfauna requires a greater amount of time and expertise (asdiscussed in Chapter 1), the rapid response of the microfauna to a community stress (withina few days) makes them a potential biological monitoring tool in outlining managementobjectives in minimizing organic enrichment of sediments at fish farms. The microfaunawill include the microbial and protozoan communities, while the meiofauna communityincludes nematodes and harpacticoid copepods. In addition benthic diatoms will also beenumerated. While some of the meiofauna surrounding aquaculture sites have beensampled briefly before (Weston 1990), a comprehensive microfaunal survey should includethe meiofauna due to possible micro- or meiofaunal interactions (Fenchel 1978, Tietjen1980, Alongi 1988). In this experiment, fish waste and oxytetracycline was applied over 20days in 200 L microcosm tanks, to simulate organic enrichment of the benthos at fish50farms. The rapid changes in the microfaunal community (and the meio- and macrofauna toa limited extent) were examined.2.2 OBJECTIVESThe objectives of this study were:(1) To establish the responses of the microfauna (bacteria and protozoan communities)to fish waste and oxytetracycline.(2) To evaluate the physical-chemical changes to the benthos, including (i) oxygenlevels in seawater due to the biological oxygen demand of the sediments, and (ii)sediment depth profiles of the reduction-oxidation potentials.(3)^To follow the decrease of oxytetracycline over time.2.3 MATERIALS AND METHODS2.3.1 EXPERIMENTAL DESIGNThe experimental period was from July to August in 1991. There were threetreatments, a Blank treatment without any fish waste or OTC added, a Control treatmentwith fish waste but no OTC added, and an OTC treatment with fish waste and OTC added.The Blank treatment consisted of three, 20 L sediment tanks, while three, 200 L sedimenttanks were used for each of the Control and Oxytetracycline (OTC) treatments (for a totalof six, 200 L tanks) (Fig. 2.1). The sediment was collected from Spanish Banks at low tide(Vancouver Harbour) and added to the tanks to an average depth of 5 cm. Sedimentcomposition ranged from fine to coarse sand (Fig. 2.2). The sediment was acclimatized inthe tanks for one month, to allow the biological, physical and chemical factors toreestablish. Water temperature and salinity during the study period varied from 11 to51Figure 2.1. The arrangement of the 200 L tanks in which fish waste and oxytetracycline wereapplied. The order of the tanks for the treatments was randomly chosen.52070 —65 =60 =5550045=0 40 =0 3530 25 =U20 7(1115 =10 -5 --0 =53<53^53-75^75-180 180-355 355-595^>595Sediment Size Range (urn)Figure 2.2. Grain size composition by relative percentages, of the sediment obtained from theintertidal zone during low tide at Spanish Banks/U.B.C. The majority of the sedimentis comprised of fine to medium sand (65%), followed by a medium to coarse sandcomponent (26%).13.5°C and 26.5 to 29.5 04,0 , respectively. As the tanks were indoors with poor andvariable fluorescent lighting, covers were provided for all tanks to eliminate any differencesdue to autotrophic activity and epiphytic growth on the walls of the tanks.Due to the rapid increases in the bacterial population in the first phase of this studyin July and August of 1990 (Chapter I) and of smaller ciliates (Fenchel 1982), the tankswere sampled every second day to ensure that a peak in microfaunal abundance would notbe missed. The duration of the experiment was thus limited to 20 days (July 13 - August 2)due to the practical restraints in time required for sampling and analysis. Fish waste wasapplied to the Control and OTC treatment tanks every day, while only oxytetracycline wasadded to the OTC treatment tanks for 4 days (from Day 9 to Day 12 of the experiment).2.3.2 THEORITICAL CALCULATION OF ORGANIC LOADING AND OTCThe discharge of excess fish feed and feces from an aquaculture facility can becalculated, based on the size of the cage or net pen and stocking densities, husbandrypractices and the feed ingested and excreted (Iwama 1991, Hajen 1990, Henderson andBromage 1988, Weston 1986). However, the accuracy of these estimates will depend onwater quality conditions around the fish farm. A mean sedimentation rate was estimated tobe 62 g dry wt•m-2 •d-1 , by suspending sediment traps directly beneath 7 salmon farms and1 smolt farm in B.C. (Aquametrix 1988). Based on this figure and the sediment surfacearea of 0.385 m2 for the 200 L tanks (Fig. 2.3), 23.9 g dry wt of fish waste was requiredto be applied to each 200 L tank per day.Oxytetracycline was contained in a coarse mixture by the trade name of Oxysol 440(supplied by Argent Laboratories, Richmond, B.C.). Treatment levels of fish with Oxysol440 is prescribed to be at 100 mg•kg-1 of biomass, or 11.5 g Oxysol 440-kg-1 fish feed(Dr. Dahl, D.V.M., pers. comm.). One gram of Oxysol 440 contains 0.44 g OTC.Although fish are usually treated for a period of 10 days with OTC (e.g. Bjorklund et al.1990), the logistics of this study limited the application of OTC to only 4 days. From 11.554Figure 2.3. A schematic of the division of grids used to randomly sample the sediments of the200 L tanks. The sediment is divided into a system of 3 columns (A-C) and 5 rows(1-5), with a central drainage pipe and a 1.5" border which was not sampled to avoidedge effects.g Oxysol 440•kg-1 fish feed and 0.0239 kg dry wt of fish waste•m -2 •d-1 , approximately0.275 g Oxysol 440.m-2 .d-1 was required to be applied to the 200 L OTC treatment tanks(or 0.005 mg OTC•mg -1 dry wt of fish waste).2.3.3 APPLICATION OF FISH WASTE AND OTCFish waste was collected from 10,000 L outdoor circular tanks, concentrated in 500ml plastic bottles and frozen at -12°C until just before use. The dryweight to volume ratioof the fish waste was predetermined. After the fish waste was thawed, the appropriateamount was measured out into smaller containers, the OTC added if necessary, and thecontainers capped. The fish waste was only added to the 200 L tanks after sampling wascompleted each day. The containers of fish waste were lowered into the water above thesediment, and the cap opened. The fish waste was slowly and evenly distributed over thesediment surface, and care was taken to avoid disturbing the sediment. The flow of waterto each tank was turned off to allow the fish waste time to settle to the sediment, and wasturned on again after half an hour.2.3.4 SAMPLING DESIGNThe 20 L sediment tanks were sampled following the grid system previouslyestablished (Chapter 1). The 200 L tanks were also randomly sampled based on a system ofsquares (Fig. 2.3) with a 1.5" border to avoid edge effects. The sides of the tanks were notwiped down to avoid disruption of the benthos. A single large sediment core was taken(i.d. 5 cm) from which smaller subcores could be drawn with the 1 ml and 3 ml syringecorers. (The cores for Eh measurements were taken directly from the 200 L tanks to ensurethat the depth profile of the sediments remained undisturbed). Bacterial, dissolved oxygen,OTC and Eh samples were all collected and measured as previously discussed (Chapter 1,Appendix 1 and 2) with bacterial samples enumerated the same day. Dissolved oxygen andEh were measured every fourth day.56Ciliate, nematode and copepod samples were taken from the same 3 ml core, and thetotal abundance recorded. Ciliates were extracted by the seawater-ice technique (Uhlig etal. 1973) using a 180gm sized mesh and collecting the ciliates in a petri dish. Thistechnique was modified by adding 8% MgC12 to the crushed ice to enhance extraction byrelaxing the ciliates (Dr. J. Berger and M. Adl, pers. comm., Kirby 1950). The ciliateswere immediately preserved in Bouin's solution (Lee et al. 1985, Montagnes & Lynn1987). The ciliate samples were later placed in 25 ml settling chambers, and enumeratedafter 24 hrs using an inverted microscope. The entire microscope field was scanned andcounted, and transects were only necessary when the smaller ciliates (<50 Am) were veryabundant.The sediment was then washed into a petri dish and 10% ethanol was added toanesthetize the nematodes and copepods. After 10 minutes, nematodes and copepods werecollected by the washing and decantation technique into a 250 m mesh sieve (McIntyre &Warwick 1984). Three washes ensured that all organisms had been extracted from the sand(personal observations). The samples were preserved in 5% formalin buffered with boraxand stained with Rose Bengal. Each sample was enumerated in a Bogorov tray, and theentire sample counted under a dissecting microscope.Diatoms were collected with a separate 3 ml corer and the sediment wasimmediately placed in a 20 ml scintillation vial, filled with filtered artificial seawater andLugol's solution was added. Each sediment sample was diluted 10 times, and three, 3 mlsubsamples were taken using a 5 ml automatic pipette as the sediment was mixed with astirring bar. Each subsample was further diluted 15 times, and a 3 ml volume from eachsubsample was placed in a 5 ml settling chamber for enumeration beneath an invertedmicroscope. Data analysis was as previously described (Chapter 1).572.4 RESULTS2.4.1 VISUAL OBSERVATIONSThe surface sediment was light brown in all tanks, and sufficient fish waste hadaccumulated by day 4 in the 200 L tanks to become visible as a fine, dark layer of organicmatter. Nudibranchs were present on the sediment surfaces of all tanks, a common speciesfound along the coast of the Pacific Northwest (Thick Horned Aeolid, Hermissendacrassicornis) that feed on coelenterates (Kozloff 1987). The nudibranchs were probablyintroduced from the seawater pipe intake system. A burrow from a marine invertebrateworm (unidentified) was present in tank 1 of the Control treatment which was avoidedduring sampling. The physical-chemical and biological characteristics of the sedimentaround the burrow will be different due to the introduction of oxygen into the sediment andfeeding activities (e.g. Findlay et al. 1990b). A small fish, the Pacific sandlance,Ammodytes hexapterus Pallas (approximately 10 cm long) was in tank 3 of the Controltreatment. The sandlance was observed to spend the majority of its time in the sediment.A. hexapterus has been documented to feed on plankton, especially small crustaceans(Clemens & Wilby 1961).It was not until day 8 (beginning of OTC addition) and day 9 that sporadic patchesof the white microbial layer began to develop on the sediment surfaces. This white layerwas smooth in appearance, while the accumulated fish waste was flocculent in nature (darkgreen to black in colour). By day 12 at the end of the oxytetracycline treatment, all of thetanks in the OTC treatment had a well developed white microbial layer on the sedimentsurface, underlain with dark brown sediment. In comparison, the tanks for the Controltreatment had a less well developed, patchier distribution of the white microbial layer. Atthis time small benthic polychaetes (Phylum Annelida) had appeared on the sedimentsurfaces of the Control and OTC treatment tanks. The polychaetes ranged in size from 0.4to 1.5 cm in length, with an average density of 1,079 individuals•m 2 (standard deviation of58377 individuals). A benthic worm (0.6 cm long) was isolated and tentatively identified as asipunculid (Dr. C. Lalli) a detritus feeder. The nudibranchs had begun to avoid thesediment surfaces and had moved to the walls of the tanks.On day 16 the Control and OTC treatment tanks were characterized by the absenceof all nudibranchs, and the white microbial layer in the OTC treatment tanks had begun tochange from being smooth to spotted in areas (resembling small tufts). The smooth whitemicrobial layer in the Control tanks was now well developed. The sediment surface of theBlank treatment tanks remained brown with nudibranchs present throughout the studyperiod. By day 18, the white microbial layer of the Control tanks had also begun to appearspotted, and the nudibranchs had reappeared on the walls of the Control and OTC tanks. Anudibranch in tank 3 of the OTC treatment was observed to be laying eggs, and egg patcheswere present in all of the 200 L tanks. A small, younger nudibranch was observed in tank3 of the OTC treatment.Sedimentary conditions had not changed for any of the tanks by the last day of theexperiment, day 20. The burrow in tank 1 and the sandlance in tank 3 of the Controltreatment remained active throughout the study. When the water supply was shut off formore than half an hour, the sandlance would emerge from the sediment until the water wasturned on again.2.4.2 REDOX PROFILES (Eh)As previously discussed (Chapter 1) the Eh measurements are related to themicrobially mediated reducing or oxidizing conditions occurring in the sediments. Theredox profiles for all of the tanks in the Blank treatment indicated that the sediment surfaces(and 0.5 cm above the surface) remained positive and highly oxidizing throughout the entirestudy period at approximately +250 mV (Fig. 2.4). The redox potential discontinuity(RPD) layer remained broad and deep, down to 2.5 cm depth throughout the study. Theredox potentials were positive at most depths in all tanks, and was oxidizing ( > +100 my)59[ Replicate 3Day 0not available• Day 477 Day 8^ Day 12^ Day 16• Day 200^N_^8^07^7I^IEh (mV)00^:11I^I^I 0N OaaNI0^0^ON c0r")^,0")I^.^I^■Replicate 10 Day 0• Day 4^ Day 8^ Day 12^ Day 16■ Day 208NONEa)a)0)1 .00.50.00.51 .01.52.02.53.03.54.00^0^00 te, IC^0^0I^I0atCO0 8♦co 8 X0♦70 0^0 01 a)a)0)1.00.5 -0.0 -0.5 -1.0  -1.5 -2.0  -2.5 -3.0 -3.5 -4.0  E"atCZIa)Ea)0)0^0N8 r.., ,,,,,,-, PnI ■ I IReplicate 2Day 0not available• Day 47 Day 8^ Day 12^ Day 16• Day 2081.00.5 -0.0 -0.5 -1.0  -1.5 -2.0 -2.5 -3.0 -3.5 -4.0Figure 2.4. Reduction-oxidation time-series depth profiles of the replicate tanks for the BlankTreatment, followed over 20 days in which no fish waste or oxytetracycline was added.Values < +100 mV (dashed line) indicates reducing conditions. Measurements began0.5 cm above the sediment surface.60in tank 2 at all depths for the duration of the study. Although the Eh below 2 cm depthbecame reducing in tanks 1 and 3, a pattern to the changes in the redox profiles was notapparent.The sediments in the replicate tanks of the Control treatment were initially similar tothose of the Blank treatment, with oxidizing conditions on day 4 and 8 (Fig. 2.5). But thesediments gradually became more reducing as fish waste was continually added to thesediments. This was strongly apparent by day 12 and 16, where the Eh readings at thesurface and 0.5 cm above the surface were strongly reducing and negative. Reducingconditions of -200 mV in tanks 1 and 3, and -40 mV in tank 2 were reached at the end ofthe study by day 20. The sediments in all of the tanks were more reducing near the surfacethan at depth at this time (as in the prior study), although this reverse RPD layer remaineddeep (down to 3 cm).The changes in the Eh of the sediments for the OTC treatment (Fig. 2.6) followedthe same pattern in the Control treatment. The surface sediments were initially welloxygenated, from the highly positive Eh values ( > +300 mV, Fig. ) and were -160 mVin tanks 1 and 3, and -120 mV at the sediment surface of tank 2 by day 20. The reverseRPD layer was also present, as the Eh near the sediment surface was more reducing than atdepth by day 16 for all tanks. The addition of OTC did not make any difference in thebetween this treatment and the Control.2.4.3 DISSOLVED OXYGENThe oscillation in dissolved oxygen levels over the experimental period followed asimilar pattern for all three treatments (Fig. 2.7), decreasing immediately and peaking againon day 12 before dropping again. The percent saturation of oxygen in seawater variedbetween 78% and 92% for all 3 treatments. Dissolved oxygen levels in the Blank treatmentwere significantly different from the Control treatment (P=0.009, Tukey Test) but not with61Cesc00C8 00 08Eh (mV)Co4^27^7I^I^I▪ 0I^I 0so. roC0^0CO0"to 0^0• toFe,0^0 0n.CEV)1.00.50.00.51.01.5 -2.0 -2.5 -3.0 -3.5 -4.0  I,^I Replicate 2Day 0not available• Day 4Day 8Day 12^ Day 16■ Day 200^0^0^ 0^ 0^0^0^0^00 47 CV 0^0^ 0^0^eV 00^0 1. 0 IN tO•-•^CO 1. 0^V' CO CV^CC^N^h^•")t^I,^I^ItI^I^II^I^I^I^I,^I^I IReplicate 3Day 0not available• Day 47 Day 8^ Day 12^ Day 16■ Day 20-a1 .00.5 -0.0 -0.5 -1.0  -1.5 -2.0  -E 2.5 -a.> 3.0 -3.5 -4.0 ^621.00.5 -0.0 -• 0.5 -.4 1.0a 1.5 -2.0-2.5-E:45• 3.0 -• 3.5 -4.0 ^Figure 2.5. Reduction-oxidation time-series depth profiles of the replicate tanks for theControl Treatment, followed over 20 days in which fish waste but no oxytetracyclinewas added. Values < +100 mV (dashed line) indicates reducing conditions.Measurements began 0.5 cm above the sediment surface.0^0Fr^!= 2 00 0ro 0CNI 0 0OI^I^I^I^I^I^I^I^I1.00.50.00.51.01.52.02.53.03.54.0Eh (my)630^07,^I001.00.50.0c.)^0.51.0• 1.5 -2.02.5• 3.0 -3.5 -^4.0 ^Day 0not available• Day 4^•^ Day 8^ Day 12^ Day 16■ Day 20 Replicate 181.00.5 -E 0.0 -c.) 0.5 -1.0 -cal▪ 1.5 -• 2.02.5 -v 3.0 -up3.5 -4.0 ^t0^N 0 F, NN^c2^03 ^1. o^CCI^I II 1I^I^I^1^I^I^I^It^l^l^l^I^I^I 0^..... -• --- -- --- --- - -------- -^_10'■^.^Day 0/ ^not available7 0/• Day 4I •v Day 8\,v^ • Day 120 Day 16Replicate 2 ■ Day 201o 0^0N ot c,^0cv^- c2)^a^oI^1 ;I,^I^I^I^I0^0^0C^0^0 mr; cci'D pr;I ,^1^I^IDay 0not available-• • Day 4^ Day 8♦ Day 12Da 16,Replicate 3 0 Day  20•Figure 2.6. Reduction-oxidation time-series depth profiles of the replicate tanks for the OTCTreatment, followed over 20 days in which fish waste and oxytetracycline was added.Values < +100 mV (dashed line) indicates reducing conditions. Measurements began0.5 cm above the sediment surface.o Blank• Controlv OTC Treated10095--Vcutx›.,k900-cia)85 -7.0w0.. 80 -75 -:70 1^1^1^i^I^[^I^I^1^I^10^2^4^6^8^10^12^14^16^18^20Time (Days)64Figure 2.7. Comparison of the dissolved oxygen levels in seawater between the BlankTreatment (fish waste and OTC not added), Control Treatment (fish waste added butnot OTC), and OTC Treatment (fish waste and OTC added). Plotted values are theMean ± 1 Standard Deviation, n=3.the OTC treatment. Dissolved oxygen levels did not differ between the Control and OTCtreatments (P > 0.05).2.4.4 OXYTETRACYCLINEOTC was only detectable on day 10, at 13.82 ± 22.87 ppm. The initialconcentration of OTC found in fish waste was 939 ± 68.7 ppm for a total of 0.12 g OTC(mean and standard deviation, n=3), but was 1100 ppm for 10.13 g OTC in Treatment B inthe previous study (Chapter 1). The initial amount of OTC should have been much lowerin this study or much higher in the previous study. This difference is due to the use of anexternal standard (OTC was injected directly on the HPLC column) in the previous study,compared to an internal standard (TC extracted from the sediments) used in this study forthe calculation of the concentration of OTC. Consequently, the concentration of OTC inthe previous study would be underestimated due to the use of the external standard.Residual amounts of OTC were detected in the sediments of this study for the remainder ofthe experimental period.2.4.5 BACTERIAL ABUNDANCETotal bacterial abundance of the Blank, Control and OTC treatments on day 0 wasvery similar (Fig. 2.8), between 6 - 7x109 • -1 sediment (corrected cell counts, AppendixA). Bacterial abundance for the Blank treatment remained approximately constantthroughout the treatment period, while both the Control and OTC treatments exhibited apeak in abundance by day 8, 1.44x10 10 bacteria•g-1 and 1.42x10 10 bacteria•g-1 sedimentrespectively. Microbial abundance in the Control treatment appeared to peak 4 days earlierthan in the OTC treatment, day 8 and day 12 respectively, but this trend was not readilyapparent due to the large variances present. After peaking, cell counts decreased in theControl and OTC treatments, remaining constant in the Control treatment but continued todrop in the OTC treatment. Overall, bacterial abundance was significantly enhanced by the65o Blank• Control^ OTC20191817161514• 13127^11rn o0 9• 8• 7-6 6543210660^2^4^6^8^10^12^14^16^18^20Time (Days)Figure 2.8. Comparison of bacterial abundance (cell x 109 • -1 sediment dry weight) over 20days, between the Blank Treatment (fish waste and OTC not added), Control Treatment(fish waste added but not OTC), and OTC Treatment (fish waste and OTC added).Plotted values are the Mean ± 1 Standard Deviation.addition of fish waste in both the Control and OTC treatments when compared to the Blanktreatment (log transformed data, Tukey Test, p< 0.05). Addition of oxytetracycline to thefish waste did not significantly affect microbial abundance between the Control and OTCtreatments (P> 0.05).2.4.6 MICROBIAL MATThe development of the microbial mat occurred simultaneously in both the Controland OTC treatments. The bacterial community were initially small, individual rod andcocci shaped bacteria 1.5 Am in length. On day 6 in both Control and OTC tanks, largerrods began to appear in the sediment and were approximately 2.5Am in length. By day 8when the smooth, white microbial layer began to be established, the microbial communitywas composed of both small cocci and large rods, with the rods forming many dense,clustered chains and colonies. After oxytetracycline was added to the OTC tanks for 4days, there were less bacteria present overall after day 12, and fewer large rods and chains.The addition of oxytetracycline appeared to be correlated to the slower development of thelarge rod shaped bacterial chains in the microbial community. Visual observations ofbacterial cells also seemed to indicate that the larger rods were associated with the whitemicrobial film, as the smaller rods and cocci attached directly to sand grains while thelarger chain forming rods were not associated directly with the sediment. It was alsoobserved that fine layers of the bacterial film formed in the stagnant water column abovethe sediment surfaces of the cores after several hours. Examination of this microbial filmindicated that it was indeed composed of only the large, chain forming rods.By day 12, the large rods had begun to decline, and the smaller cocci and rodsbegan to predominate again. The morphological characteristics of the microbial communityagain changed on day 16 in the OTC treatment, when the smooth microbial film began toappear spotted (small tufts of growth). This spotted microbial film was not as wellestablished in the Control treatment. Under the epifluorescent microscope, these tufts of67bacterial growth consisted of large, thick walled units that formed strands (approximately10 gm wide and up to 2 mm long), which were filled with small rod shaped granules (3 to8µm in length) that stained orange. Examination of the bacterial tufts with an invertedmicroscope showed that the tufts were capable of limited motion. By day 20 the whitebacterial tufts were well established in both the Control and OTC tanks, but with an unevendistribution on the sediment surface.2.4.7 PROTOZOAN ABUNDANCEThe use of Bouin's Solution as a fixative caused extreme contraction in the benthicciliates, which made accurate identification impossible. Only total ciliate abundance wasestimated (Fig. 2.9a) and identification was based on the limited morphologicalcharacteristics visible through the inverted microscope. Overall, ciliate abundance in theControl and OTC treatments was significantly different from the Blank treatment (logtransformed data, Tukey Test, P <0.05). While the addition of fish waste greatly increasedciliate abundance after 4 days in both the Control and OTC treatments, oxytetracycline didnot cause any significant changes in ciliate abundance between these two treatments(P > 0.05), although the initial peak in ciliate abundance in the OTC treatment laggedbehind the Control treatment by 2 days. A bimodal peak in abundance occurred duringboth of these treatments, with a greater initial peak in abundance of 35x103 individua1s•cm-2 on day 6 and 26x103 individuals•cm -2 day 8 for the Control and OTC tanks, respectively.The second peak in abundance was lower, with 14x10 3 and 22x103 individuals•cm -2 onday 12 in the Control and OTC treatments, respectively.A bloom of heterotrophic microflagellates appeared on day 16, after the ciliatepopulation declined on day 14 (Fig. 2.9b). The microflagellates consisted of small, circularcells (3-6 gm in diameter) that were observed to be feeding on the bacterial communitywithin the fish waste. The microflagellates reached densities of 13x106 individuals•cm-2 in6820460-40-o^-30-Ao Blank• Controlv OTCB• Controlv OTC69a)z.020-10 -,..^0 •---•---10---0 ♦^0 • • • •I^.^I^,^I^,^1^I^,^I^i^,^I^,^1^,^I^10^2^4^6^8^10^12^14^16^18^200^2^4^6^8^10^12^14^16^18^20Time (Days)Figure 2.9. Comparison of protozoan abundance over 20 days, between the Blank Treatment(fish waste and OTC not added), Control Treatment (fish waste added but not OTC),and OTC Treatment (fish waste and OTC added); where A =ciliate abundance(numbers x 10 .cm-2); B=heterotrophic microflagellate abundance (numbers x106•cm-2). Plotted values are the Mean ± 1 Standard Deviation.the Control treatment, and 7.3x10 6 individuals•cm-2 in the OTC treatment, but the 2treatments were not significantly different (T-Test, P >0.05). The microflagellates did notbecome noticeable until abundance increased significantly, and were thus not enumeratedprior to day 14.In this organically poor sediment, there were initially few ciliates in the sedimenttanks from day 0 to 4. The ciliates present ranged in length from 50 to > 300 Am, andincreased numbers in the 50-150 Am size range appeared with the continued addition of fishwaste. Some of the larger ciliates that appeared on day 6 belonged to class Karyorelictea(Order Protostomatida N. Ord. and Order Loxodida Jankowski) and subclass Haptoria (Leeet al. 1985, Corliss 1979). Small ciliates (approximately 20-40 Am in length) appeared onday 4, and were the major contributing factor for the large increases in ciliate abundance.Rotifers appeared by day 10, but were not enumerated.Examination of live samples was accomplished by placing a drop of sand (obtainedwith an eyedropper) in 5 ml settling chambers and observed through the invertedmicroscope. Seven morphologically dissimilar benthic ciliates were readily distinguishablein the Control and OTC treatments up to day 14 and 16, but had declined to only 3 or 4morphologically different ciliates at the end of the study period. There did not appear to beany differences in ciliate species diversity between the Control and OTC treatments.Benthic ciliate diversity in the Blank treatment always remained low, with never more than3 or 4 distinct ciliates present.2.4.8 DIATOM, NEMATODE AND COPEPOD ABUNDANCEWhile diatom abundance increased slightly at the end of the study period from8x102 to 26x102 cells•cm-2 (Fig. 2.10), no significant difference was found between the 3treatments (P >0.05, Tukey Test). All of the diatoms were of the pennate variety, andappeared colourless or with the cytoplasm greatly reduced. Only those cells with a majorityof cytoplasm were counted.7036 —32 —28 —c,21E0 24coo--, 20v)s..^16a>,..012Z0 Blank• ControlV OTC80^r^ 1 10^2^4^6^8^10^12^14^16^18^20Time (Days)71Figure 2.10. Comparison of diatom abundance (numbers x 10 2 •cm-2) over 20 days, betweenthe Blank Treatment (fish waste and OTC not added), Control Treatment (fish wasteadded but not OTC), and OTC Treatment (fish waste and OTC added). Plotted valuesare the Mean ± 1 Standard Deviation.Nematode densities were highly variable for all 3 treatments, with the highestabundance in the Blank treatment at 53 individuals•cm -2 on day 16 (Fig. 2.11). By the endof the study, nematode abundance declined in the Blank treatment and rose in the Controland OTC treatments, but was still not significantly different between treatments (P >0.05,Kolmogorov-Smirnov Test). Nematode sizes varied greatly from several hundred m to 1cm in length.Values for harpacticoid copepod abundance were significantly different between all3 treatments (P <0.05, Kolmogorov-Smirnov Test, Fig. 2.12). Densities initially decreasedfor all treatments, but remained highest in the Blank treatment at 23 individuals•cm -2 onday 16, while copepod abundance in the Control and OTC treatments remained at depressedlevels until near the end of the treatment period, when abundance in all 3 treatmentsconverged. Copepods with egg cases and nauplii were present throughout the study period.2.5 DISCUSSION2.5.1 REDOX MEASUREMENTSThe sediment in the Blank treatment remained oxidized with very little change inredox conditions during the experimental period (Fig. 2.4), as expected when the fish wastewas not added. In comparison, the sediments of the Control and OTC treatments steadilybecame more reducing with time at the surface and at depth by day 12 and 16 (Fig. 2.5 and2.6). The highly reducing conditions, visual observations of the changes in colour of thesediment and the development of the white microbial mat confirmed that microbiallymediated anaerobic redox reactions (i.e. sulphate reduction) were occurring (as in Chapter1; e.g. Jones 1980, Nedwell 1984, Revsbech & Jorgensen 1986, Bak & Pfennig 1991). Upto 86% of the sulphate reduction has been shown to occur in the upper 1 cm of the sediment(Skyring & Bauld 1990). The redox conditions of the sediments in the OTC treatment didnot appear to be different from that in the Control treatment, as the Eh of both treatments7270 -656055 -E50 =C■245 -240 =cl) 35E 3025 -20 -1510 =5o Blank• Control7 OTCI^'^I^I^I^ I^'^1^'^1' 1^I0^2^4^6^8^10^12^14^16 18^20Time (Days)73Figure 2.11. Comparison of nematode abundance (numbers•cm-2) over 20 days, between theBlank Treatment (fish waste and OTC not added), Control Treatment (fish waste addedbut not OTC), and OTC Treatment (fish waste and OTC added). Plotted values are theMean ± 1 Standard Deviation.o Blank• Controlv OTC24E• 20a• 16.)• 12zI^'0^2^4^6^8^10^12^14^16^18^20Time (Days)74Figure 2.12. Comparison of harpacticoid copepod abundance (numbers•cm -2) over 20 days,between the Blank Treatment (fish waste and OTC not added), Control Treatment (fishwaste added but not OTC), and OTC Treatment (fish waste and OTC added). Plottedvalues are the Mean + 1 Standard Deviation.approached -200 mV in value by the end of the experiment. These values wereconsiderably less reducing than found in the Control and OTC treatments (Fig. 1.6, Chapter1), approaching -320 mV. However, the pattern of changes to the redox profile betweenthe 2 studies were the same, with the sediment becoming more reducing at the surface thanat depth, and the RPD layer shifting closer towards the surface. The redox potentials foundin this study were very similar, to the Eh of sediments beneath fish farms in Scottish sealochs at -185 and -200 mV (Brown et al. 1987 and Lumb & Fowler 1989, respectively).This would suggest that with regard to redox potentials, the addition of fish wastes to thesediments in the microcosms closely matched the input of fish wastes from aquaculture sitesto the benthic environment, and sulphate reduction remained the primary route in thebiodegradation of fish waste. The fish waste that was cumulatively added in this study wasapproximately 1.24 kg•m-2 at the end of the experiment, with a total of 1.06 kg•m-2 in thefirst study. As the sediments were less reducing than previously found (the Eh at -200 mVinstead of -320 mV in Chapter 1) and the white microbial mat was present throughout theexperimental period, this would suggest that the sulphate in the fish waste (and thus theH2S produced by sulphate reduction) remained available throughout the experiment. Thesedimentation rate of the fish waste was insufficient for methanogenesis to have occurred.With continued accumulation of the fish waste in the sediments, greater reducing conditionscould eventually have been reached. Perhaps the sulphate was not exhausted in the fishwaste due to a diminished metabolic rate of the anaerobic microbial community in responseto the gradual addition of the fish waste, or the stimulatory effect of the grazing pressureexerted by the bacterivorous protozoan population was not as great. That the OTC did notappear to make any difference in the redox of the sediments between the Control and OTCtreatments could have been due to an insufficient quantity of OTC to affect the microfauna.752.5.2 MICROBIAL MATThe morphological changes of the microbial population in the fish waste areremarkably similar to those described by Gonzalez & Biddanda (1990), during microbialtransformation of isopod (Idotea granulosa) feces. The large colonial cells in the whitemicrobial mat and filaments that formed by day 16 could have been due to the sulphuroxidizing bacteria Beggiatoa spp. For instance, the large cells were observed to be from 3-8 Am in length, while the tufted filaments were approximately 10 Am wide and up to 2 mmlong with limited motion. These cellular dimensions are within the range described forBeggiatoa spp. (Jorgensen 1977a) in which cell sizes range from 4 to 26 Am long, withfilaments composed of hundreds of cells up to 1 cm long and 50 Am wide. As well, thefilamentous strands were similar in appearance to the Beggiatoa cultures presented by Strohl& Larkin (1978) although the "gliding" motion was not observed. Other colourless sulphuroxidizing bacteria that form white films or veils are the Thiovulum spp., but these are large,spherical cells (8-16 Am wide; Wirsen & Jannasch 1978, Jorgensen & Revsbech 1983).While filamentous cyanobacteria have also been described from benthic environments(Kuenen et al. 1985, Caumette 1989), the limited amount of light in the tanks wouldpreclude this possibility.2.5.3 DISSOLVED OXYGENThe biological oxygen demand (BOD) of the sediments should have resulted in thedissolved oxygen (DO) levels in the Blank treatment being higher than in the Control andOTC treatment tanks in which fish waste was placed. Yet the DO level in the Blanktreatment was significantly lower than in the Control treatment (Tukey Test, P > 0.05, Fig.2.7). This reversal in the expected change of DO was due to the different design of themicrocosm tanks used in the Blank treatment and the Control and OTC treatments. Forinstance, the seawater supply for the 200 L Control and OTC treatment tanks was provideddirectly from pipes (Fig. 2.1), while the seawater supply for the Blank treatment tanks had76to be piped from a further distance and settled in a header tank before emptying into the 20L tanks. Measurements of DO levels in seawater from the Blank treatment (66.46+1.87%)were significantly different from the Control and OTC treatments (78.72+0.56%; P <0.05,T-Test). Perhaps some BOD may have occurred in the header tank as a great deal ofphytodetritus was observed to have accumulated in the header tank. If this 12% differencein the DO of seawater was accounted for in the Blank treatment tanks, then the dissolvedoxygen levels would be higher overall in the Blank treatment as expected, although similarfluctuations in the DO levels of seawater for all treatments had occurred. The sharpdecrease in the DO levels of seawater on day 4 following the addition of fish waste was notdue to the BOD from the sediments, as this drop in DO also occurred for the Blanktreatment which did not have any fish waste added. The DO levels between the Controland OTC treatments were not significantly different from each other. The oxygen demandfrom sediments impacted by fish waste can be much greater than from nonpolluted sites(0.74 g.d - 1 .m-2 compared to 0.10 g.d - l.m-2 , Aquametrix 1990), which should result innoticeable differences in the dissolved oxygen content of bottom seawater. As thedifferences in DO levels were minimal between the treatments in this study, it can beconcluded that the microcosm systems were flushed with well oxygenated seawater at alltimes.2.5.4 OXYTETRACYCLINEThe total amount of OTC that was added to the fish waste in this study was 0.48 g(spread over 4 days) or 0.73 g•1 -1 total, compared to the 2.53 g and 10.13 g that wereadded to Treatment A and B of the first study all at once, respectively. A yield of 13.82 ±22.17 ppm OTC (mean ± 1 standard deviation, n=3) was obtained on day 10 from theOTC treatment, and the residual amounts of OTC detected throughout the rest of the studycould have remained within the sediments for several months (up to 419 days, Bjorklund etal. 1990 & 1991). The majority of the OTC likely washed out of the fish waste during77application, with continued diffusion and leaching from the sediment (Samuelsen et al.1988). OTC was not detected in water samples taken from near the sediment surface. Asdiscussed in Appendix B, OTC levels were underestimated due to difficulties with theextraction procedure. The redox profile of the OTC treatment appeared to be the same asin the Control treatment, again suggesting that the quantity of OTC applied to the sedimentswas insufficient to inhibit the bacterial or protozoan community.2.5.5 MICROFAUNAL ABUNDANCEBacteriaTotal cell counts in sediments not impacted by fish waste remained constant at 5 -7x1091-1 sediment (Blank treatment, Fig. 2.8), with elevated bacterial numbers up to1.4x10 10 .g-1 in the Control and OTC treatments with fish waste. Microbial abundancefrom sediments polluted by fish waste in this study (1.4 - 2.6x10 101-1 ), were several timeshigher than in previous studies (0.03 - 1.01x10 10 .g-1 ; Enger et al. 1989, Samuelsen et al.1988, Torsvik et al. 1988, Carr & Goulder 1990), which had not used ultrasonication toloosen attached sedimentary bacteria.CiliatesInitial ciliate counts were fairly low, < 50 individuals•cm-2 , composed of the larger50-300 gm ciliates. It is not unusual to find low numbers of ciliates in organically poorsediments (Parker 1981) which could have been the situation in this study, where the sandwas from the intertidal zone was strongly impacted by wave and tidal action. Theapplication of oxytetracycline did not affect ciliate abundance between the Control and OTCtreatments, as both treatments exhibited bimodal peaks in abundance from a small ciliate,20-40 gm in length (Fig. 2.9a). In both treatments, the first peak in abundance was thehighest but 2 days apart, at 35x103 individuals•cm -2 on day 6 in the Control treatment, and7826x103 individuals•cm -2 on day 8 in the OTC treatment. The second peak in ciliateabundance occurred on day 12 in both treatments, 14x103 and 22x103 individuals•cm-2 inthe Control and OTC treatments, respectively. As data on protozoan abundance insediments beneath fish farms have not been previously collected, it is difficult to assesswhether the ciliate numbers observed in this study were within the range for sedimentspolluted by fish waste. In comparison, ciliate numbers in different benthic habitats canrange from a few hundred to several thousands individuals•cm -2 (e.g. Kemp 1990), whichwould suggest that the values for ciliate abundance observed were much greater than usuallyfound. In exceptional cases, ciliate numbers can reach as high as 75x103 -cm-2 inorganically enriched sediments (Finlay 1978). This is within the range of values found inthis study.As both treatments exhibited the same oscillations in ciliate numbers (rapid increasesand decreases in abundance) these changes could not be attributed to the OTC. The 2 daylag by the ciliates in the OTC treatment to reach the first maxima on day 8 (compared today 6 in the Control treatment), was not due to the OTC either as this lag period occurredbefore the addition of OTC. The lag period was thus due to natural variations within thebenthic ecosystem. As the concentration of OTC in this study was over 5 times less than inTreatment A of the previous study (in which the protozoan community was not consideredto have been inhibited, Section 1.5.7), the amount of OTC in this study is thus consideredto have been insufficient to have inhibited the ciliate community. The OTC also could havebeen diluted to a greater extent during the gradual addition of the OTC over 4 days,compared to the addition of OTC at one time. The low concentrations of OTC found in thesediments beneath fish farms after 12 months (up to 5 ppm; Bjorklund et al. 1990, Nygaardet al. 1992), thus may not be sufficient to inhibit bacterial or protozoan activity.79Ciliate DiversityIn general, macro-, meio- and microfaunal populations exhibit similar changes inspecies diversity, numbers and biomass to different sources of organic pollution (e.g.Pearson & Rosenberg 1978, Weston 1986, Gowen & Bradbury 1987, Ritz et al. 1989).When protozoan communities are environmentally stressed (i.e. by a pollutant) there is anelimination in sensitive species and an increase in the numbers of tolerant species, leadingto higher numbers and lower species diversity (Atlas 1984). Qualitative changes in thisstudy showed that the larger ciliates ( > 250 Am) disappeared quickly with organicenrichment of the benthos, and the protozoan community soon became dominated by thesmaller, much more numerous ciliates. While the diversity of the protozoans increasedfrom a few morphologically distinct ciliates to several distinct ciliates during the first 2weeks of sediment enrichment, this declined to only a few ciliate types again (althoughthese ciliates were different than those initially present at the start of the experiment).Species numbers are higher in transition zones where deposition of nutrients are occuring(Pratt & Cairns 1985). Studies of protozoan colonization in nutrient enriched systems (Prattet al. 1987) have shown that colonizing protozoans often "overshoot" eventual speciesequilibrium, where a species maximum can be reached within a few days followed by adecrease in species numbers after longer intervals. The seawater-ice technique forextracting benthic ciliates was not considered quantitative, as observations of drops ofunpreserved sediment would often show some types of ciliates that were not present in thepreserved samples. While nearly all benthic ciliates can be extracted with this technique(Fenchel 1969), extraction efficiences have been found to be low with underestimations ofciliate abundance by over 90% (Alongi 1986). It is possible that the number of ciliatespecies was underestimated, as it would have been easy to miss the rare species.While the diversity of ciliate species has been shown to decrease in response toagricultural and domestic pollution with an associated increase in numbers (Hul 1987 &1988, Grabacka 1988, Kosciuszko & Prajer 1990), a higher ciliate diversity, abundance and80biomass was found in the sediments of a Scottish Loch polluted by effluent discharge froma pulp and paper mill (Wyatt & Pearson 1982). This greater ciliate diversity was attributedto not only the increase in available prey (with the sulphur bactera serving as the major foodsource), but also to the structural changes in the sediment surface layers from the cellulosefibre, that provided a greater number of potential niches available for colonization.Heterotrophic MicroflagellatesMicroflagellate abundance remained relatively low in the sediments throughout thestudy period (Fig. 2.9b, < 1x105 individuals-cm-2), until after the ciliate populationdeclined on day 14. A bloom of the heterotrophic microflagellates occurred in the Controland OTC treatments, to a maximum of 12.5x10 6 individuals-cm-2 on day 18, and wereabsent in the Blank treatment. The addition of oxytetracycline did not affect microflagellateabundance between the Control and OTC treatments. Comparison with flagellateabundance in other types of sediment indicates that the values from this study were high(Kemp 1990) but not exceptionally so (Fenchel 1975, Kemp 1988).2.5.6 MICROFAUNAL INTERACTIONSThe dynamics of protozoa associated with the degradation of fish waste indicated apredator-prey relationship with bacteria (Gonzalez & Biddanda 1990). There was a positivecorrelation between the initial peaks in ciliate and bacterial numbers (Fig. 2.8 and 2.9a)with ciliate abundance in the Control and OTC treatments occurring 2 and 4 days,respectively, before the peak in bacterial abundance. This suggested that bacterial numbersdid not peak as quickly in the previous study (Chapter 1) partly due to this grazing pressure.Other causes may have been the differences in the application of fish waste and OTC,which was added gradually in this study instead of all at once; or that the fish waste wasstored frozen prior to use in this study, but was added to the sediments without freezing inthe previous experiment. The greater amount, and the unfrozen state of the fish waste81would have allowed a greater proportion of the bacterial and protozoan community to havebeen initially present in the first study (Chapter 1), which could have resulted in a differentmicrofaunal community developing with different responses to the benthic ecosystem andOTC than in this study. While ciliate numbers always decreased after reaching a maxima,bacterial abundance never peaked again and was likely continually grazed back down(Fenchel 1982). The second, smaller peak in ciliate numbers may reflect a decrease inbacterial prey availability. Although bacterial abundance remained depressed, bacterialactivity could have been high, which has been shown to be stimulated by macro-, meio- andmicrofaunal grazing (e.g. Barsdate et al. 1974, Fenchel & Jorgensen 1977, Gerlach 1978,Finlay 1978, Tietjen 1980, Montagna 1984). A combination of the fish waste (as a supplyof organic matter) and the grazing effects exerted by the bacterivorous ciliates appear to bethe controlling factors among the microfauna (i.e. a substrate, substrate and grazing, andgrazing control; Weisse & Scheffel-Moser 1991, Bak & Nieuwland 1989). The rapidresponse of the bacterial and ciliate populations also implies that growth and productivity ofthe microfauna in sediments organically enriched by fish waste can be high.While benthic ciliate bacterivory was potentially important in controlling bacterialproduction in sediments enriched by fish waste, most studies have indicated that of the 5 to50% of benthic bacterial production that could be consumed by protozoans, most grazing isattributed to microflagellates (e.g. Barsdate et al. 1974, Fenchel 1975 & 1986, Kemp 1988& 1990, Patterson & Fenchel 1990). As a microflagellate bloom occurred quickly after theciliate bloom ended on day 14 (Fig. 2.9b), the microflagellates thus became the primarygrazers of bacteria. Kemp (1990) suggests that microflagellates should ingest a largefraction of bacterial production only when the relative abundance of microflagellates tobacteria is on the order of 1:1000. Comparison of bacterial and microflagellate abundancecan be easily accomplished by transforming bacterial abundance (Fig. 2.8) to cells•cm-2 .The mean sediment dry weight for the 1 ml bacterial cores is 0.697 g (0.154 g standard82deviation) and the surface area is 0.126 cm 2 . For instance, bacterial abundance at 6 -14x109 •g-1 sediment can also be presented as 3.32 to 7.74x10 10 •cm-2 . As themicroflagellate:bacteria ratio of 1:1000 is met (8 - 12x10 6 flagellates.cm-2 : 3.32 -7.74x10 10 cells•cm-2), there is a potential capacity for these heterotrophic microflagellatesto be important bacterivores. Existing data suggests that microflagellates are more likely tobe important in organically enriched sediments, while ciliates may be more important in theinterstitial spaces of sandy sediments (Kemp 1990).Interactions within the protozoan community can also occur, as ciliates can feed onflagellates (Fenchel & Finlay 1990), and at least 1 ciliate was observed to contain a fewmicroflagellates (personal observations, this study).Interpretation of how the microfauna interact also depends on sedimentarycomposition and the prevailing chemistry of the sediments (due to the bacterial activityresulting from the influx of organic materials). Oxygen does not penetrate as deeply infiner sediments (or in sandy sediments with a heavy organic load) and organic matteroxidizes slower (Fenchel 1967), with ciliates and protozoans adsorbing to the finerparticles. Microbial activity is greatest near the surfaces, and the strongly reducing natureof the sediments also restricts protozoan populations to the surface layers of the sediment(e.g. Fenchel 1969, Marty 1981, Joint et al. 1982, Novitsky & Karl 1986). Benthicprotozoan abundance has been correlated to organic enrichment (and the resulting bacterialactivity) and other physical factors such as the Eh and temperature of the sediments (Wyatt& Pearson 1982). It is interesting to note that the decline in ciliate abundance (after day14, Fig. 2.9a) occurred when the sediments reached anoxic, reducing conditions by day 16(Fig. 2.6 and 2.6). The small, 20 gm ciliates were thus dependent on an aerobicenvironment, if they were sensitive to reducing conditions. Perhaps ciliate abundance was .correlated to both the increase in bacterial abundance and the development of stronglyreducing conditions. As the characteristics of the sand used in the experiments outlined in83Chapters 1 and 2 were mud/silt and sand, respectively, the initial microfaunal communitymay have been different. However, the heavy organic loading and resulting reducingconditions in the sediments would have restricted the bacterial and protozoan communitiesto the surface of the sediments in both studies.2.5.7 DIATOMSDiatoms (mainly of the pennate variety) increased in abundance from 8x10 3cells•cm-2 to 14 - 24x103 cells•cm-2 (Fig. 2.10), but there was no difference in abundancebetween the three treatments. While nutrients (inorganic phosphorous) released from cagefish farming activites in rivers and lakes have stimulated the growth of benthic diatomcommunities downstream of the fish farms (Carr & Goulder 1990, Stirling & Dey 1990),high concentrations of pollutants will generally reduce diatom species diversity andabundance (Atlas 1984, Kwandrans 1988). Lumb & Fowler (1989) found that diatomswere present in sediments away from fish farms, but declined and disappearedapproximately 15 m away from the aquaculture sites. A gradient from diatoms to a whitemicrobial mat (Beggiatoa spp.) occurred at this point, corresponding to the zone where theredox potential at the sediment surface reached < 0 mV.The appearance of nearly all of the diatoms in this study were colourless or withgreatly reduced cytoplasms. Most of the diatoms could have been dormant or dead, as lightwas excluded from the microcosms. However, Wyatt & Pearson (1982) also reported thepresence of living diatoms in similar conditions from sediments enriched by cellulose fibres,and in the presence of hydrogen sulphide. The viability of diatoms in highly reducingsediments is partly attributed to an ability to tolerate reduced levels of light for longperiods, and perhaps to function as facultative or obligate saprophytes (Lackey 1961), orlive heterotrophically in the dark on dissolved glucose or lactate (Fenchel 1969). Whilediatoms may be able to survive under these conditions, active cells were not observed in livesediment samples, or undergoing division. The diatoms could have been mostly neritic84species, having settled out of the photic zone (Lackey 1961, Wyatt & Pearson 1982). Theabsence or low abundance of diatoms under polluted conditions (e.g. Lumb & Fowler 1989,Kwandrans 1988, Wyatt & Pearson 1982), would tend to confirm that entire algalcommunities can be eliminated if pollution becomes too extreme (Atlas 1984). There isthus a potential for error in this assessment of diatom abundance, as there is a strongpossibility that dead cells were also counted. Recalling that the source of seawater for themicrocosms was bottom water from Vancouver Harbour, and the outlet for the drainagepipe in the microcosm tanks was at the water surface instead of at the bottom, the slightincrease in diatom abundance could have been due to an accumulation of cells rather thanfrom an active and dividing diatom population. Possible trophic interactions could haveoccurred, with some species of the diatoms serving as prey for the ciliate and nematodecommunity (Wyatt & Pearson 1982, Findlay 1982), but the poor condition of the diatomswould suggest that in this study, diatoms were not an important source of carbon.2.5.8 MEIOFAUNANematodes and harpacticoid copepods can respond quickly to pollution, forming thebasis for the nematode/copepod ratio in organic pollution studies (e.g. Raffaelli & Mason1981, Raffaelli 1987), with a subsequent decrease in species diversity (Hockin 1983). Ingeneral, nematodes would increase in abundance along an increasing gradient of organicpollution, but would decrease at the highest levels of enrichment. Mesobenthic (interstitialcopepods) will decrease with increasing pollution due to a combination of the clogging ofinterstitial spaces, low oxygen and increasing sulphide concentrations. However, epi- andendobenthic (surface and shallow burrowing) species are not as affected by pollution, andtend to increase with increasing organic enrichment, being able to swim up to water withhigher dissolved oxygen levels (reviewed by Hicks & Coull 1983, Heip et al. 1985,Raffaelli 1987).85The results from this study would tend to support the expected trends that wouldoccur in the harpacticoid copepod and nematode communities with organic enrichment (Geeet al. 1985, Widbom & Elmgren 1988). Nematode density in the Control and OTCtreatments remained lower than in the Blank treatment (55 individuals•cm -2 on day 16, Fig.2.11), but abundance began to rise at the end of the study and decrease in the Blanktreatment. Copepod abundance in the treatments with fish waste decreased, with the highestnumbers reached in the Blank treatment at 24 individuals•cm -2 on day 16 (Fig. 2.12)although all treatments exhibited similar values by day 20. Weston (1990) found thatnematodes and polychaetes contributed to over 99% of the total number of individuals insediments directly beneath a mariculture site in Puget Sound (Washington, U.S.A.), withsurface and near-surface-dwelling crustacean species (amphipods, cumaceans, isopods andostracods) becoming more abundant along a decreasing gradient in pollution away from thefish farm. The presence of crustacean nauplii (assumed to be harpacticoid copepod larvaeas no other crustacean species were observed) and copepods with egg cases throughout thestudy, were an indication of the ability of the copepod community to survive organicenrichment of the sediments, at the present dosage levels of fish waste and OTC. However,as there were no significant differences found between the different treatments due to thelarge fluctuations in abundance, an absolute pattern cannot be established.The OTC levels in this study did not appear to affect copepod and nematodeabundance, but high concentrations (0.25 M OTC) had an immediate and toxic effect on thenematodes (personal observations). Crustaceans were also affected, but survived for alonger period of time. The long-term effects of low concentrations of OTC on meiofaunalcommunities has never been examined. Studies on the effects of OTC on dauer larva (adormant stage) of the nematode Caenorhabditis elegans, indicates that there is a moderateto strong inhibitory effect on pharyngeal pumping activity at concentrations of 5 mM, withimpaired incorporation of methionine into protein, delaying longitudinal growth (Reape &Burnell 1991). Lower concentrations of OTC slowed growth, but the developmental86pathway of the nematodes eventually became normal once treatment with the antibioticceased. While the OTC concentrations in this study did not appear to reach toxic orstrongly inhibitory concentrations, the concentrations of OTC initially present in the fishwaste during addition to the sediments (and perhaps at aquaculture sites during treatment ofdiseases with medicated feed), could slow growth down for a short period of time. Largepopulations of nematodes can still develop under these conditions though, as Weston (1990)found high densities of nematodes in sediments enriched by fish waste (90,507individuals•m-2).Rotifers were not enumerated, and their contribution to the marine meiofauna is notlarge (Fenchel 1978) and is more important in brackish and freshwater systems (Sanders etal. 1989). Weglenska et al. (1987) established that the rotifer community was important, inthe bottom waters beneath rainbow trout cage cultures of Lake Glebokie (Poland) feedingon bacteria and plankton. Perhaps the impact of rotifer activity in sediments polluted byfish waste should also be assessed for future studies.2.5.9 INTERACTIONS BETWEEN MICRO- AND MEIOFAUNABacteria, ciliates and diatoms have been implicated as food items for nematodes andcopepods, while dead meiofauna can serve as a substrate for bacteria and diatoms (e.g.Fenchel 1978, Alongi & Tietjen 1980, Montagna 1984, Rieper 1985, Eskin & Coull 1987,Stoecker & Capuzzo 1990). However, the importance of the microfauna as a potential foodsource for the meiofauna is debatable, as Montagna et al. (1983) found that the meiofaunadid not respond to changes in potential food abundance. Alongi (1988) also suggested thatprotozoan and meiofaunal populations may not be tightly coupled to the dynamics ofbacterial and microalgal communities in tropical intertidal habitats, and Wyatt & Pearson(1982) never observed the meiofauna to be feeding on the ciliate populations.Physical-chemical factors (temperature and Eh) have also been implicated incontrolling meiofaunal abundance (Montagna et al. 1983), while Finlay (1980) decided that87interactions between the physical-chemical factors (Eh, pH, oxygen availability, daylight)and the biological factors (bacterial, ciliate, benthic chlorophyll-a and nematodepopulations) determined the distribution of the micro- and meiofauna (Heip et al. 1985).However, correlations between changes in the meiofauna, microfauna and Eh of thesediments were not apparent.The addition of fish waste to sediments increased bacterial and protozoan abundance,although the response of the meiofaunal community was less clear. While the meiofaunacould have been feeding on the microfaunal community, potential interactions could not beconcluded with certainty, as correlations between micro- and meiofaunal abundance wereweak or not apparent. Trends in meiofaunal abundance may require a longer period of timeto become apparent, as meiofauna will typically require a relatively greater period of timeto respond to organic enrichment, due to longer life cycles (Hicks & Coull 1983, Heip etal. 1985). For instance, while the bacteria and protozoan populations greatly increasedduring the experimental period of 20 days, the meiofauna may require several weeks to afew months to exhibit such large changes. The lower abundance of the nematode andcopepod populations suggests that their impact on the microfauna will not be important untilhigher population densities are reached.2.5.10 MACROFAUNAThe presence of the nudibranchs and small polychaetes did not appear to disturb thesediment surfaces, as the redox profiles remained relatively consistent for all tanks of eachtreatment. The disappearance of the nudibranchs from the sediment surface may have beenrelated to the development of anoxic, reducing conditions, or perhaps to reproductiveactivity. The presence of nudibranch egg patches would also suggest that prey items(coelenterates) were abundant (but the coelenterates were not collected or enumerated).Polychaete abundance did not approach the densities seen in the field (e.g. 1,079individuals•m-2 compared to over 25,000•m -2 , Tsutsumi 1987) as the duration of this study88was too short for such large changes in polychaete abundance to occur. Perhaps over alonger period of time and with a greater amount of fish waste and potential prey (e.g.bacteria, ciliates, diatoms), the polychaete population could increase greatly. The impact ofOTC on macroinvertebrates has not been considered before, but the high growth andproductivity of macroinvertebrate populations (at least for C. capitella) in sediments aroundfish farms would suggest that toxic or inhibitory concentrations of OTC are not reached. Itis reasonable to expect that OTC is present in the sediments beneath the fish farms duringpart or the entire year when the macrofaunal studies were taken, as the use of medicatedfeed during disease outbreaks at fish farms during the summer can occur more than once(Weston 1986), and low concentrations of OTC can persist for long periods of time in thebenthic environment (e.g. Bjorklund et al. 1991). The activity of the sandlance shouldhave resulted in bioturbation of the sediment, but this was not evident in any changes in theredox profiles. The sandlance was probably collected with the sediments from SpanishBanks, and its ability to survive implies that there was sufficient prey available(crustaceans, probably harpacticoid copepods). However, differences between copepodabundance of the 3 treatments were not significant. In addition, as macrofauna wouldrequire several months for large increases in abundance to become apparent (Mattson &Linden 1983), correlations should not be expected with the micro- and meiofauna, and thephysical-chemical conditions over the experimental period of 20 days in this study.2.5.11 PROTOZOANS IN BIOLOGICAL MONITORINGCiliates have commonly been used in Europe and Asia as a biological tool in theevaluation of the degree of water pollution (saprobiological analysis, Sladecek 1973).Based on the dominating species present, several zones of pollution can be classified (e.g.Grabacka 1985 & 1988), although explaining the temporal variation in the frequency andabundance of individual species can be difficult. Finlay et al. (1981) suggested that insteadof relying on subjective decisions of whether or not large fractions of the species present are89characteristic of the degree of pollution, classification techniques could be used to indicatethe presence or absence of significant associations between known or defined protozoancommunities in polluted areas (association analysis).The short-term response of ciliate populations (a few weeks) presents an idealopportunity to utilize the changes in the ciliate communities as a rapid biological tool inidentifying the organic pollution of sediments around aquaculture sites. Rather than waitingthe required several months to identify severe pollution problems through the majorstructural changes in the meio- or macrofaunal communities, rapid and expedient measurescan be taken to minimize the impact of fish waste on the benthos and environment. Thisstudy has shown that the use of ciliates as a biological monitoring tool, for identifying thepollution of sediments impacted by fish waste can be promising. However, a great deal offurther research is required to identify the major components of the microfaunal communitythat exists beneath fish farms (e.g. bacterial and protozoan species and abundance), andpotential interactions with the meio- and macrofauna.902.6 SUMMARYThe main objectives of these 2 studies were to determine if and how the micro- andmeiofauna would react to organic enrichment of the sediments by fish waste, and anadditional response to the antibiotic oxytetracycline. These investigations have shown that:(1) The benthic bacterial population will greatly increase in magnitude, shortly after theaddition of fish waste to sediments.(2) Protozoan populations will also increase rapidly along with the increase in bacterialabundance, perhaps mediated by the reduction-oxidation conditions in the sediments.(3)^Interactions between the micro-, meio- and macrofauna may have occurred,although there were little or no correlations apparent. The duration of this studywas considered to be insufficient to have allowed substantial changes in meio- andmacrofaunal abundance to occur.(3) Oxytetracycline was found to persist at low concentrations for relatively long periodsof time, with minimal or no inhibitory effect on microbial communities (possiblydue to the expression of oxytetracycline resistance). It is unlikely that highconcentrations of OTC would occur in the sediments beneath fish farms, or wouldbe found for long periods of time in the sediment due to the rapid dispersion of themajority of OTC in the marine environment.(4) The protozoan community did not appear to be affected by OTC levels that would befound at aquaculture sites treating fish with medicated feed. However, higherdosage levels of OTC may have inhibited bacterivorous protozoan grazing, resultingin an increase in bacterial abundance but decrease in metabolic activity, and overallslower mineralization of the fish waste.91( 1 )(5)^The rapid response of the microfauna suggests that protozoan communities would bepromising as a biological tool in pollution assessment studies on the impact of fishwaste on the benthic environment. This in turn could lead to more rapididentification of potential pollution problems and efficient management at fish farmsto minimize these environmental problems.2.7 RECOMMENDATIONS FOR FUTURE STUDIESThe effects of fish waste and oxytetracycline on the microfaunal (and meiofaunal)communities requires further work. Currently, research is strongly underway in studyingthe long-term behaviour and effects of oxytetracycline in the benthos and on the microbialcommunity (e.g. Norway) and on macrofaunal communities. However, there is a lack ofinformation on changes in meiofaunal and protozoan communities.While it was established that fish waste may increase microfaunal abundance, with afurther potential effect by OTC, further research should concentrate on the followingaspects to determine the interactions of the benthic communities in the degradation of fishwaste.If marine benthic protozoan populations are to be used as an indicator of organicenrichment (e.g. by association analysis), precise taxonomy is required, along with atechnique that will provide an accurate assessment of abundance.Determination of short and long-term physiological and developmental responses ofprotozoans to differing concentrations of oxytetracycline is also needed.Measurements of bacterial growth, heterotrophic activity and OTC resistance, alongwith taxonomic studies should be taken to determine the distribution and abundanceof potential prey items for the protozoan population.9293(4) A concurrent study of the meiofaunal community, with a complete assessment ofabundance and species diversity to evaluate potential interactions with themicrofaunal community.(5) The responses of the microfauna to organic enrichment by fish waste and OTC at afish farm should be assessed, when physical and biological factors are not uncoupled(e.g. light, wave/tidal action, increased recruitment, macrofaunal interactions, etc.).The field study should be long-term (over several months or years) to ensure thatseasonal changes in the microfauna will not be misinterpreted.(6)^The physical-chemical and biological factors of the benthic communities should bemonitored after cessation of activities at different culture sites, to determine optimalconditions for the recovery of the sediments.While these recommendations would require a large amount of effort, these studieshave been completed before in different areas and for different sources of organicenrichment. For instance, the ecology of the marine microbenthos has been previouslyestablished (e.g. Fenchel 1967 & 1969) and the ecological, biological and physical-chemicaleffects of cellulose degradation in the benthos have also been well documented (e.g.Pearson 1982 and references therein). Benthic trophic interactions and nutrient studies onmangrove litter in tropical coastal regions have also been completed (e.g. Alongi & Tenore1985, Alongi 1988, Alongi et al. 1989, Alongi 1990). While the required research on theecological, biological and physical-chemical questions associated with fish waste fromaquaculture may require several years and involve many groups and individuals, this wouldnot be an impossible task.APPENDIX AEXTRACTION OF ATTACHED BACTERIAIN SEDIMENTS BY ULTRASONICATIONINTRODUCTIONThe enumeration of bacteria from sediments by ultrasonication or homogenizationhave been shown to be the most effective methods in releasing the attached bacteria (e.g.Montagna 1982, Ellery & Schleyer 1984). The adhesion of bacteria to sediments are acombination of physicochemical processes (Van Loosdrecht et al. 1989, Kemp 1990).These processes involve electrostatic attraction such as Van der Waals forces, coulombicattraction (Krone 1978), hydrophobicity of cells (Fattom & Shilo 1984, Rosenberg &Kjelleberg 1986) as well as special appendages (e.g. pili or fibrils) or polymers produced bybacteria to attach to surfaces (e.g. Marshall et al. 1971, Weise & Rheinheimer 1978, Shilo1989, Mir et al. 1991). Up to 33% of bacteria can be left behind on sand grains afterhomogenization (Newell & Fallon 1982) and Ellery & Schleyer (1984) demonstrated thatultrasonication can be significantly more effective than homogenization in separatingattached sedimentary bacteria (using a Decon FS 100 ultrasonication bath sweeping througha frequency range of 40 to 50 kHz, power output between 100 and 200 W). Dye (1979 &1983) found that in sediments > 200 p,m, sonication was very effective in removingattached bacteria, whereas in silty/clay sediments homogenization proved to be better inremoving attached bacteria. As the major size fraction of the sediments used in this studywere in the 180-355 Am range (Chapters 1 and 2), ultrasonication was the technique used inthis study to release bacteria attached to sediments. The sampling and subsamplingvariability in counting bacteria attached to sediments can be quite large (Montagna 1982).But the effect of strengthening cells with formaldehyde and dispersing them with adeflocculent (e.g. 'Tween 80 or pyrophosphate) and ultrasound can cause bacteria to be94randomly distributed, with subsequently lower variances between subsamples. For thisexperiment a Branson B-220 ultrasonication bath, sweeping through a frequency of 50 to 60kHz and 125 W (at 117 volts) was utilized. The optimum ultrasonication time required forthis ultrasonicator type was assessed prior to the study, and was found to be 90 seconds forthis type of sediment.MATERIALS AND METHODSA sand sample was taken with a 1 ml syringe-corer and placed in a 20 mlscintillation vial with 3.55 ml of filtered seawater (0.22 gm pore size), and 0.25 ml of 37%formaldehyde (2.1 % formaldehyde, final concentration). The preserved sand sample wasplaced in a 125 ml Erlenmeyer flask and shaken vigorously with 80 ml of sterile, filtered(0.22 gm pore size) distilled water with 'Tween 80 added (1x10 -4 %) (supplied by SigmaChemical Co., St. Louis, MO), following the technique of Ellery and Schleyer (1984).Ultrasonication was performed for 30 seconds. The sand sample was shaken again and thesolution was allowed to settle for 30 seconds to allow the heavier sand grains to settle out,which has been determined not to affect abundance estimates of bacteria (Montagna 1982).A 1 ml subsample was then extracted from the solution. Ultrasonication times of 0, 30, 60,90, 120, 180, 240 and 300 seconds were applied in series to the sand sample, with 1 mlsubsamples taken for bacterial enumeration after each sonication period. Acridine orangeepifluorescence microscopy was used to obtain direct cell counts from the diluted samples(Parsons et al. 1984).To ascertain the extent of bacterial destruction by ultrasonication, 80 ml of sterile,filtered distilled water with 'Tween 80 was added to another formalin preserved sandsample. The sample was vigorously shaken and the larger sand particles allowed to settleout after 30 seconds. Eight 2 ml aliquots of the bacterial suspension were removed andplaced in individual test tubes, and ultrasonicated in series for the same time intervals as the95previous experiment. Direct counts were performed following the aforementioned acridineorange epifluorescence microscopy technique.Statistical analysis was performed with SYSTAT (Ver 5.0, Evanston, IL). Squareroot transformations were required to meet the assumptions of normality and homogeneityof variance (Bartlett's test, P 0.05).RESULTS AND DISCUSSIONThe effect of ultrasonication with time on bacterial abundance (numbers•g 4sediment dry weight) was highly significant (ANOVA, p 0.05). An initial abundance of9.8x107 cells•g-1 (Fig. 1) was observed with 4x10 8 cells•g -1 present after 90 seconds ofultrasonication. Ultrasonication for 90 seconds or more yielded consistent results, TukeyTest, p z 0.05, with a plateau in bacterial abundance after 90 seconds. The optimalultrasonication time for extraction of attached sedimentary bacteria in this study wasobserved to be 90 seconds, rather than 2 1/2 minutes in the study by Ellery and Schleyer(1984). This may reflect differences in the ultrasonicator models, sediment sizes and typesused, and bacterial types present. However, in determining their optimal ultrasonicationtime, Ellery and Schleyer (1984) did not appear to rigorously test for the peak in bacterialabundance (as indicated by a lack of error bars). For example, their optimal sonicationtime could have been less, if substantial overlap occurred between the standard deviation ofmeans.The destructive effect of ultrasonication is apparent in Figure 2. There is animmediate, significant decrease in bacteria (p <0.05, ANOVA) with increasedultrasonication time (as in Ellery and Schleyer 1984) to 77.5%. There was no signicantchange in the percent of bacterial abundance for the remaining sonication times. Thegreater variation present at 180 seconds corresponds to the point when large amounts ofbacterial fragments began appearing in the samples. For an ultrasonication time of 90965ao2S.a)Ez19700^30^60^90^120 150 180 210 240 270 300Sonication Time (Seconds)Figure 1. Accumulation of bacteria attached to sediments following ultrasonication times of30, 60, 90, 120, 180, 240 and 300 seconds, of a formalin preserved sand sample.Values presented are the Mean ± 1 Standard Deviation, n=5.100 —---co90 —80 —a)tn 70—060 —c.) 50 —a40 —-30980^30^60^90^120 150 180 210 240 270 300Sonication Time (Seconds)Figure 2. The destructive effects on a bacterial suspension extracted from the sediment,following ultrasonication times of 30, 60, 90, 120, 180, 240 and 300 seconds. Valuespresented are the Mean ± 1 Standard Deviation, n=4.seconds, all bacterial counts should therefore be increased by 143% (standard deviation of26%) to obtain a more accurate estimate of bacterial abundance in the suspension.Two samples had an increase in bacterial abundance after 2 minutes. Directobservation of sand grains at this time still showed attached bacteria. This suggests that alonger ultrasonication time is required to release those bacteria still attached to thesediment. Although 90 seconds was taken to be the optimal ultrasonication time, it isapparent that another source of error may be causing an underestimation of bacterialabundance. To determine the fraction of bacteria (%) still attached to the sediment, sandsamples were sonicated for 90 seconds (in a 1:200 dilution), 1 ml of the suspensionremoved for counting and the rest of the suspension discarded. The sand sample was rinsedtwice and the rinses discarded each time, sonicated again (in a 1:100 dilution) for 60seconds, 1 ml of the suspension removed for counting and the remaining suspensiondiscarded. This procedure was repeated twice more for 60 seconds each time (Fig. 3) for atotal sonication time of 270 seconds.There was an exponential decrease in bacterial abundance with increasingultrasonication time, with approximately 65% of the bacteria present released after the first90 seconds of ultrasonication. An additional 1 minute treatment released only 22% morebacteria, with the remaining 13% extracted during the next 2 minutes of sonication. Thisexponential release of bacteria would seem to indicate that there are portions of themicrobial population that require greater ultrasonication times to detach from the sediments.This may be due to the presence of different types of adhesion processes used (e.g. VanLoosdrecht et al. 1989). For example, as bacteria in marine environments have negativecharges on their surface membranes, in distilled water (low ionic strength) Van der Waal'sattraction between bacteria and particles are weakened considerably (Krone 1978, VanLoosdrecht et al. 1989). This would allow the fraction of the sedimentary microbialpopulation utilizing Van der Waal's forces to be removed quickly, while the bacteria that99Log Y = 2.185 — 0.007Xr2= 0.896 •1000.)7:1rnQ.)a 10 7OCo0E1a)Ua)Ucu▪ 0.1 —4.1001^' ^ I^'^'^I^I^'^I^'^10^30^60^90^120 150 180 210 240 270 300 330 360Sonication Time (Seconds)Figure 3. The percentage of bacteria remaining on sediments, following ultrasonication periodsof 90, 150, 210 and 270 seconds. The relationship is linearized by log transformationand fitted with a first order regression and 95% confidence limits, n=16.are more firmly attached by appendages or polymers may require a longer ultrasonicationtime to be displaced. The initial events in the attachment of non-motile and motile bacteriaprimarily due to physical-chemical forces (Marshall et al. 1971, Scheraga et al. 1979). Torelate this exponential decline in bacterial abundance with increasing ultrasonication time,log transformation of percent bacteria was required to linearize the curve (Fig. 3:r2 =0.896, highly significant, p .0.001):LOG (Y) = 2.185 - 0.007(X)This equation can be used to correct for the additional underestimation of bacteria notreleased during the optimal ultrasonication period of 90 seconds. For instance, with 90seconds of ultrasonication approximately 3.90x108 bacteria•g -1 dry weight of sediment arereleased (Fig. 1). Destruction by ultrasonication is 1.68x10 8 bacterial-1 sediment (Fig.2). Using the above equation (Fig. 3) 35.89% of the bacteria are still attached to thesediment at 90 seconds of ultrasonication. The cells released at 90 seconds (including thosecells destroyed by sonication, for a total of 5.58x10 8 bacterial-1 sediment) only accountfor 64.11 % of the overall bacterial abundance. A further correction is therefore required,resulting in an additional 3.12x10 8 bacteria•g-1 sediment. The total bacterial count shouldtherefore be 8.70x10 8 bacteria• -1 sediment, over twice the abundance of the initialbacterial counts at 90 seconds. It is not certain how prolonged ultrasonication may degradethose bacteria remaining attached to the sediment, and this correction does not account forthat. It may be possible that an underestimation of bacterial abundance still exists. Incomparison to the study by Ellery and Schleyer (1984) the second source for theunderestimation of bacterial abundance was not calculated. This formula is valid onlywithin the range tested for (Ricker 1984, Zar 1984), in this instance the sonication periodsfrom 90 to 330 seconds.101The increase in bacterial abundance after an additional minute of sonication (afterthe initial 90 seconds) is not apparent in Fig. 1. It may be that the rate of destruction ofcells by ultrasonication (from the microbial population released from the sediment in thefirst 90 seconds) is equivalent to the rate of release of bacteria after an additional oneminute of ultrasonication. However, these rates of destruction and release have not beenmeasured.Although Ellery and Schleyer (1984) found a strong trend of decreased bacterialabundance with ultrasonication time, their results do not consider that the bacterialsuspension they used was originally extracted from a homogenized sand sample. Bacterialcells may have been already damaged by 2 minutes of homogenization, thus moresusceptible to destruction by ultrasonication. Losses caused by damage during extractionmay also differ between sediment types as well, and these effects should be evaluated.Ellery and Schleyer (1984) sampled surface sediment from a lagoon, while the sediment inthis study came from a 20 m depth in Vancouver Harbour (ranging from mud to fine sand).Examination of the sediment following 5 minutes of sonication shows negligible amounts ofbacteria attached as in Ellery and Schleyer (1984), suggesting that for the sediment typesexamined for their study and in this one, 5 minutes of ultrasonication is sufficient to loosenthe majority of the sedimentary bacteria.The reasons for differences in optimum ultrasonication times may be due toextraction techniques, differences in sediment types as well as components of the microbialpopulation, and types of ultrasonicators used. The use of ultrasonication in extraction ofattached sedimentary bacteria to find the optimal ultrasonication time (and correction factordue to subsequent destruction) should be determined for each study.102APPENDIX BEXTRACTION OF OXYTETRACYCLINE FROM SEDIMENTSBY HIGH-PERFORMANCE LIQUID CHROMATOGRAPHY.A CALIBRATION AND RECOVERY STUDY.INTRODUCTIONIn aquaculture, oxytetracycline (OTC) is a widely used broad-spectrum antibacterialdrug, active against Gram-positive and Gram-negative microbes of both aerobic andanaerobic species (Neu and Caldwell 1978). The primary effect of tetracyclines is thoughtto be inhibition of protein synthesis, at the 30S ribosomal site (Levy 1984).Oxytetracycline is usually administered orally to fish through medicated feed. However,this method of treatment results in a very low amount of OTC being taken in by the fish(Bjorklund et al. 1991). Reduced feeding by diseased fish due to decreased feed intake isdue in part to the effects of the infection and to the unfavourable taste of the OTCmedicated feed. Feed intake has been shown to be reduced to 61 % of normal (Hustvedt etal. 1991), to complete refusal of medicated feed (Salte 1982) in healthy rainbow trout(Oncorhynchus mykiss) when OTC was present in the feed pellets. When treating farmedfish, the decrease in feed intake is compensated for by increasing the dosage of OTC perkilogram of biomass. Concentrations range from 100 mg OTC/kg fish (about 200 mgOTC/kg feed, Dr. E.L. Dahl, Argent Laboratories Ltd., Richmond, B.C.) to 4000-16000mg OTC/kg feed (Grondel et al. 1987). This obviously results in a greater proportion ofthe uneaten feed pellets and oxytetracycline collecting on the benthos.Of the OTC ingested by the fish, the majority of the OTC remains unabsorbed.Cravedi et al. (1987) estimated a bioavailability of 7-9% for rainbow trout in freshwater,while Rogstad et al. (1991) found a maximum absorption of 2.6%, 72 hours afteradministration of medicated feed to rainbow trout. Grondel et al. (1987) found absorptionlevels to be only 0.6% for carp after 25 days (Cyprinus carpio L). The OTC is excreted to103the environment in the feces (Rogstad et al. 1991) and also accumulates in the benthos.The low bioavailability within the fish may be due to the high affinity of OTC to formcomplexes with di- and trivalent cations (Lunestad and Goksoyr 1990, Grondel et al. 1987),with the OTC binding strongest to the bones and tissues of fish (Rogstad et al. 1991). Thetime required for complete elimination of OTC from fish is temperature dependant(Jacobsen 1989). Times for rainbow trout range from 60 days for temperatures > 10°C to100 days between 7-10°C (Salte 1982, Salte and Liestol 1983). More recently, withdrawaltimes of 27 days at 16°C to 135 days at 5°C have been suggested for rainbow trout(Bjorklund and Bylund 1990). In comparison, Aoyama et al. (1991) could not detect anyOTC in chinook salmon muscle tissue 35 days after treatment of medicated feed (at watertemperatures of 7-10°C). This means that OTC can still be released in fish feces for a longperiod following treatment, and is bound to particles and sediments beneath the fish farm(e.g. Samuelsen 1989, Bjorklund et al. 1990).It has not been until recently that the degradation of oxytetracycline in theenvironment has been examined (Samuelsen et al. 1988, Samuelsen 1989, Jacobsen andBerglind 1988, Bjorklund et al. 1990, Bjorklund and Bylund 1991). The half-life of OTCin sediments ranges from 9 to 419 days depending on anoxic conditions, with a rapid half-life in seawater (5 to 16 days).The objective of this study is to examine the effects of fish waste and oxytetracyclineon the microbenthic community. The methodology for the extraction of OTC fromsediments was adapted from Jacobsen and Berglind (1988). A calibration and recoverystudy of this technique was simultaneously run, to evaluate the sensitivity and replicabilityof the method.104MATERIALS AND METHODSCHEMICALSOxytetracycline hydrochloride (OTC) and tetracycline hydrochloride (TC) wereobtained from Sigma Chemical Co. (St. Louis, MO). Oxalic acid was from Fisher ScienticCo. (Fair Lawn, NJ) while methanol and acetonitrile (Omnisolv, HPLC grade) wasobtained from BDH (Toronto, Ont., Canada). For the Na2-EDTA/Mcllvaine buffer,ethylenediaminetetraacetic acid disodium salt (NA2-EDTA) was from BDH (Toronto,Ont.), citric acid monohydrate from Fisher Scientific Co., and the sodium phosphatedibasic from Matheson Coleman & Bell (Ohio). Pure OTC is not used for treatment of thefood pellets, but is supplied as Oxysol 440 by Argent Chemical Laboratories Inc.(Richmond, B.C., Canada) where 227 g of Oxysol 440 contains 100 g OTC.APPARATUSThe HPLC system that was utilized for the calibration and recovery study, and forthe second year of this study was a Shimadzu SIL-9A autoinjector coupled to a ShimadzuSPD-6A UV Spectrophotometer Detector and a Beckman 110A pump. Integration of theareas under the curves was performed by a Shimadzu C-R3A Chromatopac. The analyticalcolumn was a 254 x 4.6 mm I.D. RP-8 column, packed with Adsorbosphere-5 pm particlesize. The prefiltering element was a direct connect universal guard column cartridge holderfitted with guard column cartridges packed with C8 5/2 Adsorbosphere (supplied by AlltechAssociates Inc., Deerfield, IL).CHROMATOGRAPHYThe mobile phase consisted of a methanol-acetonitrile-0.01 M aqueous oxalic acidsolution (1:1.5:2.5) (v/v/v), pH 2 at a flow rate of 1 ml•min -1 . This solution was degassedunder a vacuum prior to use, to eliminate bubbles from entering the column. Distilled andde-ionized water was used to make all solutions, which were then filtered (0.22 /AM105Millipore filters). HPLC grade methanol and acetonitrile were also filtered before use.The wavelength of the UV detector was set at 365 nm, rather than 350 nm suggested byOka et al. (1985) and Jacobsen and Berglind (1988), as this wavelength has been found tobe more sensitive to OTC (McErlane, pers. comm.).Pure OTC was used to make a stock solution of 500 Ag.m1 -1 OTC, by dissolving 5mg in 10 ml of methanol. Differing volumes of serial dilutions were made with methanol togive concentrations of 0.5, 1.0, 4.0, 10, 20 and 30 Ag•m1 -1 (adapted from Aoyama et al.1991).Pure TC was used as an internal standard, for the calibration curves and in therecovery study. A stock solution of 200 Ag.m1-1 TC was made by dissolving 2.0 mg in 10ml methanol, and a 10 Ag.m1 -1 working solution was obtained by diluting 0.5 ml of the TCstock solution to 10 ml with methanol. Enough of the internal standard TC was added tothe samples for a final concentration of 1.0 Ag•g-1 sediment (Aoyama et al. 1991).The recipe for the Mcllvaine buffer (Oka et al. 1986) required for the extraction ofOTC from sediments was provided by Aoyama (pers. comm.), and was made by dissolving28.41 g dibasic sodium phosphate in distilled water in a 1 L flask and diluting to volume.After dissolving 21.01 g citric acid monohydrate in distilled water and diluting to volume,the 1 L citric acid solution was mixed with 625 ml of the sodium phosphate solution in a 2L flask, pH of 4.0. The 0.1 M Na2-EDTA/Mcllvaine buffer was prepared by adding 60.49g EDTA to the 1.625 mL Mcllvaine mixture.CALIBRATION CURVE AND RECOVERY STUDYA calibration curve was needed to calibrate the UV detector and determine if thetetracycline internal standard and oxytetracycline can be accurately measured over theconcentrations that would be present in the samples. If the OTC and TC are not interactingwith each other or other compounds which may be present, there should be a linearrelationship (Bjorklund 1988, Bjorklund et al. 1990). For an internal standard106concentration of 1.0 gg•g -1 sediment 0.5m1 TC was added to 5 g sediment, and 0.5 ml ofthe following OTC standard solutions was also added to give concentrations of 0.05, 0.1,0.4, 1, 2 and 3 ppm sediment. The calibration curve was made by plotting the ratios of thepeak ares of OTC to the peak areas of the internal standard against the knownconcentrations of OTC.To determine the efficiency of the extraction technique for this type of sediment, arecovery study was made by adding 0.5 ml of 1, 10, and 20 Ag•m1 -1 of the OTC stocksolutions to 5 g of dried sediment, but without an internal standard added. The area of thecurves for the OTC extracted from these samples were plotted against the area of the curvesfor 0.5 ml of the unextracted OTC standard solutions, (adapted from Aoyama et al. 1991).EXTRACTION PROCEDUREThe sediment sample was removed from the scintillation vial, the volume wasmeasured in a 10 ml graduated cylinder and dry weight calculated based on the volume todry weight ratio. Tetracycline was then added as the internal standard, to a concentrationof 1 Ag.g4 sediment (adapted from Aoyama et a/. 1991). The sample was then rinsed intoa 50 ml polypropylene centrifuge tube with 15 ml of the 0.1 M Na2-EDTA/Mcllvainebuffer, and homogenized with a variable speed Tri-R stirrer (Model S63C, 12000 rpm, 115V, Tri-R Instruments Inc., Rockville Centre, N.Y.), at 5000 rpm for 1 minute (adaptedfrom Jacobsen and Berglind 1988). A 5 ml aliquot of Na2-EDTA/Mcllvaine buffer wasadded and the sediment homogenized for 1 minute, after which another 5 ml of the Na2-EDTA/McIlvaine buffer was added and the sediment homogenized for another minute.Two ml of the Na2-EDTA/Mcllvaine buffer was used to rinse the stirring rod, and the 2 mlbuffer was also collected in the 50 ml centrifuge tube.The sample was then centrifuged for 15 minutes at 2500 rpm (RefrigeratedCentrifuge model PRJ, International Equipment Co.). After centrifugation, the supernatantwas decanted into a 15 ml filtration tower, and filtered through a 0.45 Millipore filter. The107filter tower was then rinsed and filtered with 1 1/2 ml Na2-EDTA/Mcllvaine buffer. Aftera 60 ml syringe was used to collect the filtered supernatant, the flask was rinsed with anadditional 1 1/2 ml Na2-EDTA/Mcllvaine buffer, and the buffer rinse was also collected inthe 60 ml syringe.An Alitech C18 900 mg Maxiclean cartridge was activated with 10 ml of methanoland 10 ml of distilled water, and the syringe attached to the female end of the cartridge.The supernatant was drawn through the cartridge by vacuum filtration at the recommendedrate of 1.5 drops per second, maximizing the retention of OTC by the cartridge. Thecartridge was then washed with 20 ml distilled water, and the OTC was eluted by passing10 ml 0.01 M methanolic oxalic acid solution through the cartridge and collecting themethanolic-oxalic acid in a 20 ml scintillation vial. All samples were stored at -12°C forapproximately a week, until ready for analysis.RESULTS AND DISCUSSIONThe peak for OTC eluted from the column at 7.49 minutes (standard deviation =0.05minutes, n=6). The recovery of OTC from the sediments is exceptionally poor followingthis technique. Quantities below 1 ppm did not yield a recognizable peak, while at 1 ppmof OTC the efficiency of recovery was only 3.28%. Generally, there were no quantifiablepeaks, although one OTC sample at 3 ppm did show a 0.68% recovery. For tetracycline,the recovery was even less from the sediment, with no peak recognized even at aconcentration of 4 ppm. The calibration did not yield a linear relationship, but this can beattributable to the poor extraction efficiency. Although small peaks were present up to 1ppm (increasing in peak height), the peak height remained approximately constant to thehighest concentration of 3 ppm.This low extraction efficiency is unusual as Bjorklund et al. (1990) using a similartechnique in extracting OTC from sediments had an 88.6% efficiency, and a lowerextraction efficiency of 72.2% for tetracycline. Loss of OTC in the extraction process108could have occurred due to a number of causes. During filtration of the OTC-Na2-McIlvaine-EDTA supernatant through the C18 cartridges, the flow rate is critical tooptimize extraction of OTC, and the columns should not go dry during the procedure(Aoyama, pers.comm.). White et al. (1974) emphasized the presence of small amount ofEDTA in the mobile phase prevent tetracycline antibiotics from complexing with the metaltubing of the HPLC colunmn (the formation of hydrogen bonds with active sites on thesilica surfaces, The Supelco Reporter 1985). Ciarlone et al. (1990) noted the binding oftetracycline antibiotics to borosilicate glass and polypropylene labware, although for OTCand TC this was not a significant loss. Rapid degradation of OTC occurs when in solution.Cold storage is required, and even at -20°C some degradation still occurs (Aoyama andMcErlane, pers. comm.). Dark storage is also necessary, as tetracyclines photodecomposeeasily (Samuelsen 1989, Oka et al. 1989). All samples were kept in dark, cold storageexcept during the extraction process, and the loss of OTC due to degradation is expected tobe minimal (although this loss was not measured). It is not known if OTC may have beenretained on the 0.45 pm filter, when filtering the supernatant following centrifugation. Thefilter was assumed to collect only particulate matter, allowing the OTC in solution to passthrough. The large loss in extraction efficiency is still not clear, but may have occurredwith the Mcllvaine buffer which is one of the critical steps in extracting OTC (Oka et al.1985). An attempt to raise the pH of the Na2-EDTA/Mcllvaine buffer to 4 with HC1 couldhave destabilized the oxytetracycline (Onji 1984). An excessive amount of HCl may havebeen added, as the acidic buffer (pH 4.0) used to calibrate the pH meter was too acidic.Jacobsen and Berglind (1988) did not show the extraction efficiency of OTC fromsediments, after adapting the technique from Oka et al. (1985) for salmonid tissues. It isnot known if this technique was ever accurate and sensitive for OTC. As this method hadalready been used to extract OTC from the sediment samples, I feel that my results will thusbe inaccurate and will lead to a large underestimation of OTC in the sediments. The results109may be qualitative in indicating that OTC is present, but unfortunately cannot be usedquantitatively.It is suggested that a double extraction step be included (adapted from Aoyama et al.1991) as this may help increase extraction efficiency. After centrifugation and filtration ofthe supernatant, the sediment should be re-homogenized with 15, 5 and 5 ml of Na2-EDTA/McIlvaine solution, centrifuged and filtered. The supernatant can be collected in thesame 125 ml filtration flask used initially, and filtered through the C18 cartridge.110REFERENCESAckefors, H. and M. Enell. 1990. Discharge of nutrients from Swedish fish farming toadjacent sea areas. Ambio. 19(1):28-35.Aller, R.C. and J.Y. Yingst. 1980. Relationships between Microbial Distributions and theAnaerobic Decomposition of Organic Matter in Surface Sediments of Long IslandSound, USA. Mar. Biol. 56:29-42.Al-Masaudi, S.B., M.J. Day and A.D. Russell. 1991. Effect of some antibiotics andbiocides on plasmid transfer in Staphylococcus aureus. J. Appl. Bacter. 71:239-243.Alongi, D.M. 1986. Quantitative estimates of benthic protozoa in tropical marine systemsusing silica gel: a comparison of methods. Est. Coas. Shelf Sci. 23:443-450.Alongi, D.M. 1988. Microbial-meiofaunal interrelationships in some tropical intertidalsediments. J. Mar. Res. 46:349-365.Alongi, D.M. 1990. Abundances of benthic microfauna in relation to outwelling ofmangrove detritus in a tropical coastal region. Mar. Ecol. Prog. Ser. 63:53-63.Alongi, D.M. and K.R. Tenore. 1985. Effect of detritus supply on trophic relationshipswithin experimental benthic food webs. I. Meiofauna-polychaete (Capitella capitata)interactions. J. Exp. Mar. Biol. Ecol. 88:153-166.Alongi, D.M. and J.H. Tietjen. 1980. Population growth and trophic interactions amongfree-living marine nematodes. IN: Marine benthic dynamics. Tenore, K.R. and B.C.Coull. (ed.) University of South Carolina Press, Columbia. p:151-166.Alongi, D.M., K.G. Boto and F. Tirendi. 1989. Effect of exported mangrove litter onbacterial productivity and dissolved organic carbon fluxes in adjacent tropicalnearshore sediments. Mar. Ecol. Prog. Ser. 56:133-144.Aoki, T., S. Egusa and T. Arai. 1974. Detection of R factors in naturally occurring Vibrioanguillarum strains. Antimicrob. Agents Chemother. 6:534-538.Aoki, T., T. Kitao and K. Kawano. 1981. Changes in drug resistance of Vibrio anguillarumin cultured ayu, Plecoglossus altivelis, Temminck and Schlegel, in Japan. J. FishDis. 4:223-230.Aoyama, R.G., K.M. McErlane, H. Erber, D.D. Kitts, and H.M. Burt. 1991. HPLCAnalysis of Oxytetracycline in Chinook Salmon Following Administration ofMedicated Feed. J. Chromatogr. 588:181-186.Aquametrix Research Ltd. 1988. A preliminary study of the deposition and dispersioncharacteristics of waste materials associated with salmon farming operations inBritish Columbia. Prepared for B.C. Ministry of Environment and Parks.Aquametrix Research Ltd., Sidney, B.C. 37pp + app.Aquametrix Research Ltd. 1990. Benthic impacts of salmon farming in British Columbia.Volume 1. Summary Report. Prepared for the British Columbia Ministry ofEnvironment, Water Management Branch. Aquametrix Research Ltd., Sidney, B.C.78pp. + app.111Atlas, R.M. 1984. Diversity of Microbial communities. Marshall, K.C. (ed). Adv. Microb.Ecol. 7:1-40.Aure, J. and A. Stigebrandt. 1990. Quantitative estimates of the eutrophication effects offish farming on fjords. Aquaculture. 90:135-156.Austin, B. 1985. Antibiotic pollution from fish farms: effects on aquatic microflora.Microb. Sci. 2(4): 113-117.Austin, B. and D. Allen-Austin. 1985. A review of bacterial pathogens of fish. J. AppliedBacter. 58:483-506.B.C. Salmon Farmers Association. 1992. Aquaculture: British Columbia's Future.Canadian Cataloguing in Publication Data. 100p.Bachelete, G. 1990. The choice of a sieving mesh size in the quantitative assessment ofmarine macrobenthos: a necessary compromise between aims and constraints. Mar.Environ. Res. 30:21-35.Bagheri, E.A. and D.S. McLusky. 1982. Population dynamics of oligochaetes and smallpolychaetes in the polluted Forth estuary ecosystem. Neth. J. Sea. Res. 16:55-66.Bak, F. and N. Pfennig. 1991. Microbial sulfate reduction in littoral sediment of LakeConstance. FEMS Microbiol. ecol. 85:32-42.Bak, R.P.M. and G. Nieuwland. 1989. Seasonal fluctuations in benthic protozoanpopulations at different depths in marine sediments. Neth. J. Sea Research.24(1):37-44.Barsdate, R.J. R.T. Prentki and T. Fenchel. 1974. Phosphorous cycle of model ecosystems:significance for decomposer food chains and effect of bacterial grazers. Oikos.25:239-251.Battersby, N.S. and C.M. Brown. 1982. Microbial activity in organically enriched marinesediments. IN: Nedwell, D.B. and C.M. Brown (ed.) Sediment microbiology.Academic Press. p:147-170.Berninger, U.G., B.J. Finlay and P. Kuuppo-Leinikki. 1991. Protozoan control of bacterialabundances in freshwater. Limnol. Oceanogr. 36(1):139-147.Bjorklund, H.V. 1988. Determination of Oxytetracycline in Fish by High-PerformanceLiquid Chromatography. J. Chromat. 432:381-387.Bjorklund, H.V. and G. Bylund. 1990. Temperature Related Absorptiona nd Excretion ofOxytetracycline in Rainbow Trout (Salmo gairdneri R.). Aquaculture. 84:363-372.Bjorklund, H., J. Bondestam and G. Bylund. 1990. Residues of oxytetracycline in wild fishand sediments from fish farms. Aquaculture. 86:359-367.Bjorklund, H.V., C.M.I. Rabergh and G. Bylund. 1991. Residues of oxolinic acid andoxytetracycline in fish and sediments from fish farms. Aquaculture. 97:85-96.Blackstock, J. 1984. Biochemical metabolic regulatory responses of marine invertebrates tonatural environmental change and marine pollution. Oceanogr. Mar. Biol. Ann. Rev.22:263-313.112Brown, J.R., R.J. Gowen and D.S. McLusky. 1987. The effect of salmon farming on thebenthos of a Scottish sea loch. J. Exp. Mar. Biol. Ecol. 109:39-51.Buchanan, J.B. 1984. Sediment Analysis. IN: Holme, N.A. and A.D. McIntyre (ed).Methods of study of marine benthos. IBP Handbook 16. Blackwell ScientificPublications, California. p:41-65.Bullough, W.S. 1962. Practical Invertebrate anatomy. 2nd ed. MacMillan and Co., Ltd.483 pp.Cahill, M.M. 1990. Bacterial Flora of fishes: A Review. Microb. Ecol. 19:21-41.Carr, O.J. and R. Goulder. 1990. Fish farm effluents in rivers. II. Effects on inorganicnutrients, algae and the macrophyte Ranunculus penicillatus. Wat. Res. 24(5):639-647.Carss, D.N. 1990. Concentrations of wild and escaped fishes immediately adjacent to fishfarm cages. Aquaculture. 90:29-40.Caumette, P. 1989. Ecology and general physiology of anoxygenic phototrophic bacteria inbenthic environments. IN: Cohen, Y. and E. Rosenberg (ed). Microbial Mats.American Society of Microbiology. p:283-304.Chopra, I. 1985. Mode of action of the tetracyclines and the nature of bacterial resistance tothem. IN: Hlavka, J.J. and J.H. Boothe (ed). The Tetracyclines. Springer-Verlag,Germany. p:317-392.Ciarlone, A.E., B.W. Fry and D.M. Ziemer. 1990. Some Observations on the Adsorptionof Tetracyclines to Glass and Plastic Labware. Microchem. J. 42:250-255.Clemens, W.A. and G.V. Wilby. 1961. Fishes of the Pacific coast of Canada. 2nd Ed.Fisheries Research Board of Canada. Bulletin No. 68. p:394-396.Cravedi, J.P., G. Choubert and G. Delous. 1987. Digestibility of Chloramphenicol,Oxolinic Acid and Oxytetracycline in Rainbow Trout, and the Influence of theseAntibiotics on Lipid Digestibility. Aquaculture. 60:133-141.Curds, C.R. 1973. The role of protozoa in the activated sludge process. Amer. Zool.13:161-169.Dale, T. 1987. Oil pollution and plankton dynamics. II. Abundance pattern of ciliatesinside and outside enclosures and the responses of ciliates to oil during the 1980Spring bloom in Lindaspollene, Norway. Sarsia. 72:197-202.DuPont, H.L. 1986. Consumption of raw shellfish - is the risk now unacceptable? N. Eng.J. Med. 314(11):707-708.Dye, A.H. 1979. Measurement of biological oxygen demand in sandy beaches. S. Afr. J.Zool. 14:55-60.Dye, A.H. 1983. A method for the quantitative estimation of bacteria from mangrovesediments. Estuar. Coast. Shelf Sci. 17:207-212.Egidius, E. 1984. Diseases of salmonids in aquaculture. Helgol. Meeres. 37:547-569.113Egidius, E., L.P. Hansen, B. Jonsson and G. Naevdal. 1991. Mutual impact of wild andcultured Atlantic salmon in Norway. J. Cons. int. Explor. Mer. 47:404-410.Eleftheriou, A. and N.A. Holme. 1984. Macrofauna Techniques. IN: Methods for the studyof marine benthos. Holme, N.A. and A.D. McIntyre (ed). Blackwell ScientificPublications. p:140-216.Ellery, W.N. and M.H. Schleyer. 1984. Comparison of Homogenization andUltrasonication as Techniques in Extracting Attached Sedimentary Bacteria. Mar.Ecol. Prog. Ser. 15:247-250.Enell, M. 1987. Environmental impact of cage fish farming with special reference tophosphorous and nitrogen loadings. J. Cons. int. Explor. Sea CM. F:44.Eng, C.T., J.N. Paw and F.Y. Guarin. 1989. The environmental impact of aquaculture andthe effects of pollution on coastal aquaculture development in Southeast Asia. Mar.Poll. Bull. 20(7):335-343.Enger, 0., B. Husevag and J. Goksoyr. 1989. Presence of the fish pathogen Vibriosalmonicida in fish farm sediments. Appl. Environ. Microb. 55(11):2815-2818.Eskin, R.A. and B.C. Coull. 1987. Seasonal and three year variability of meiobenthicnematode populations at two estuarine sites. Mar. Ecol. Prog. Ser. 41:295-303.Esteban, G., C. Tellez and L.M. Bautista. 1991. Dynamics of ciliated protozoacommunities in activated sludge process. Wat. Res. 25(8): 967-972.Fattom, A. and M. Shilo. 1984. Hydrophobicity as an adhesion mechanism of benthiccyanobacteria. Appl. Environ. Microbiol. 47(1): 135-143.Fauque, G., J. Legall and L.L. Barton. 1991. Sulfate reducing and sulfur reducingbacteria. IN: Shivelyand, J.M. and L.L. Barton (ed.) Variations in Autotrophic life.Academic Press. p:271-337.Federle, T.W., R.J. Livingston, L.E. Wolfe and D.C. White. 1986. A quantitativecomparison of microbial and community structure of estuarine sediments frommicrocosms and the field. Can. J. Microbiol. 32:319-325.Fenchel, T.M. 1967. The ecology of marine microbenthos. I. The quantitative importanceof ciliates as compared with metazoans in various types of sediments. Ophelia.4:121-137.Fenchel, T.M. 1969. The ecology of marine microbenthos. IV. Structure and function ofthe benthic ecosystem, its chemical and physical factors and the microfaunacommunities with special reference to the ciliated protozoa. Ophelia. 6:1-182.Fenchel, T.M. 1975. The quantitative importance of the benthic microfauna of an ArcticTundra Pond. Hydrobiologia. 46(4):445-464.Fenchel, T.M. 1978. The ecology of micro and meiobenthos. Ann. Rev. Ecol. Syst. 9:99-121.Fenchel, T.M. 1982. Ecology of heterotrophic microflagellates. IV. Quantitativeoccurrence and importance as bacterial consumers. Mar. Ecol. Prog. Ser. 9:35-42.114Fenchel, T.M. 1986. The ecology of heterotrophic microflagellates. Marshall, K.C. (ed).Adv. Microb. Ecol. 9:57-98.Fenchel, T.M. and B.B. Jorgensen. 1977. Detritus Food chains of aquatic ecosystems: Therole of bacteria. Alexander, M. (ed.) Adv. Microb. Ecol. 1:1-57.Fenchel, T.M., T. Perry and A. Thane. 1977. Anaerobiosis and symbiosis with bacteria infree-living ciliates. J. Protozool. 24:154-163.Fenchel, T. and B.J. Finlay. 1990. Anaerobic free living protozoa: growth efficiencies andthe structure of anaerobic communities. FEMS Microbiol. Ecolog. 74:269-275.Fernandez-Astorga, A., A. Muela, R. Cisterna, J. Iriberri and I. Barcina. 1992. Biotic andabiotic factors affecting plasmid transfer in Escherichia coli strains. Appl. Environ.Microbiol. 58(1):392-398.Findlay, S.E.G. 1982. Effect of detrital nutritional quality on population dynamics of amarine nematode (Dtplolaimella chitwoodi). Mar. Biol. 68:223-227.Findlay, R.H., M.B. Trexler, J.B. Guckert and D.C. White. 1990a. Laboratory study ofdisturbance in marine sediments: response of a microbial communityl. Mar. Ecol.Prog. Ser. 62:121-133.Findlay, R.H., M.B. Trexler and D.C. White. 1990b. Response of a benthic microbialcommunity to biotic disturbance. Mar. Ecol. Prog. Ser. 62:135-148.Finlay, B.J. 1978. Community production and respiration by ciliated protozoa in thebenthos of a small eutrophic loch. Freshwater Biology. 8:327-341.Finlay, B.J. 1980. Temporal and vertical distribution of ciliophoran communities in thebenthos of a small eutrophic loch with particular reference to the redox profile.Freshwater Biology. 10:15-34.Finaly, B.J. 1990. Ecology of free-living protozoa. Marshall, K.C. (ed). Adv. Microb.Ecolog. 11:1-36.Finlay, B.J., P. Bannister and J. Stewart. 1979. Temporal variation in benthic ciliates andthe application of association analysis. Freshwat. Biol. 9:45-53.Folke, C. and N. Kautsky. 1989. The role of ecosystems for a sutainable devlopment ofaquaculture. Ambio. 18(4): 234-243.Ford, R.J. 1984. Norwegian Salmon and trout farming. Mar. Fish. Rev. 46(3):44-47.Frid, C.L.J. and T.S. Mercer. 1989. Environmental monitoring of caged fish farming inmacrotidal environments. Mar. Poll. Bull. 20(8):379-383.Gee, J.M, R.M. Warwick, M. Schaanning, J.A. Berge and W.G. Ambrose, Jr. 1985.Effects of organic enrichment on meiofaunal abundance and community structureinsublittoral soft sediments. J. Exp. Mar. Biol. Ecol. 91:247-262.Gerlach, S.A. 1978. Food-chain relationships in subtidal silty sand marine sediments andthe role of meiofauna in stimulating bacterial productivity. Ocecologia. 33:55-69.115Gonzalez, H. and B. Biddanda. 1990. Microbial transformation of metazoan (Idoteagranulosa) faeces. Mar. Biol. 106:285-295.Gowen, R.J. and N.B. Bradbury. 1987. The ecological impact of salmonid farming incoastal waters: A Review. Oceanogr. Mar. Biol. Ann. Rev. 25:563-575.Grabacka, E. 1985. Ecology of some waters in the forest agricultural basin of the RiverBrynica near the Upper Silesian Industrial Region. 8. Ciliata in bottom sediments.Acta Hydrobiol. 27(4): 521-533.Grabacka, E. 1988. A regulated river ecosystem in a polluted section of the Upper Vistula.7. Bottom Ciliata. Acta Hydrobiol. 30(1/2):73-80.Grave, K., M. Engelstad, N.E. Soli and T. Hastein. 1990. Utilization of antibacterial drugsin Salmonid farming in Norway during 1980-1988. Aquaculture. 86:347-358.Gray, J.S., M. Aschan, M.R. Carr, K.R. Clarke, R.H. Green, T.H. Pearson, R.Rosenberg, R.M. Warwick. 1988. Analysis of community attributes of the benthicmacrofauna of Frierfjord/Langesundfjord and in a mesocosm experiment. Mar.Ecol. Prog. Ser. 46:151-165.Grondel, J.L., J.F.M. Nouws, M. de Jong, A.R. Schutte and F. Driessens. 1987.Pharmacokinetics and Tissue Distribution of Oxytetracycline in Carp, Cyprinuscarpio L., Following Different Routes of Administration. J. Fish Dis. 10: 153-163.Hajen, W. E. 1990. Evaluation offeedstuff digestibility in post-juvenile Chinook Salmon(Oncorhynchus tshawytscha) in seawater. M. Sc. Thesis. Dept. of Animal Science.University of British Columbia.Hall, P.O.J. and 0. Holby. 1986. Environmental impact of a marine fish cage culture. J.Cons. int. Explor. Sea. CM. F:46.Hall, P.O.J., L.G. Anderson, 0. Holby, S. Kollberg, M.O. Samuelsson. 1990. Chemicalfluxes and mass balances in a marine fish cage farm. I. Carbon. Mar. Ecol. Prog.Ser. 61:61-73.Hansen, J.A., D.M. Alongi, D.J.W. Moriarty and P.C. Pollard. 1987. The dynamics ofbenthic microbial communities at Davies Reef, central Great Barrier Reef. CoralReefs. 6:63-70.Hansen, L.S. and T.H. Blackburn. 1991. Aerobic and anaerobic mineralization of organicmaterial in marine sediment microcosms. Mar. Ecol. Prog. Ser. 75:283-291.Hanson, R.B. 1980. Measuring microbial activity to assess detrital decay and utilization.IN: Tenore, K.R. and B.C. Coull (ed.) Marine Benthic Dynamics. University ofSouth Carolina Press. p:347-358.Hargrave, B.T. 1972. Aerobic decomposition of sediment and detritus as a function ofparticle surface area and organic content. Limnol. Oceanogr. 17(4):583-596.Harrison, B.J. 1981. The biological determinants of the structure of harpacticoid copepodcommunities on an estuarine intertidal flat (Fraser River Delta, B. C.). Ph.D.Thesis. Dept. of Zoology, University of British Columbia (Canada). 400p.116Heip, C., M. Vincx and G. Vranken. 1985. The ecology of marine nematodes. Oceanogr.Mar. Biol. Ann. Rev. 23:399-489.Heip, C., R.M. Warwick, M.R. Carr, P.M.J. Herman, R. Huys, N. Smol and K. VanHolsbeke. 1988. Analysis of community attributes of the benthic meiofauna ofFrierfjord/Langesund ord. Mar. Ecol. Prog. Ser. 46:171-180.Henderson, J.P. and N.R. Bromage. 1988. Optimising the removal of suspended solidsfrom aquacultural effluents in settlement lakes. Aquacult. Engineer. 7:167-181.Hicks, G.R.F. and B.C. Coull. 1983. The ecology of marine meiobenthic harpacticoidcopepods. Oceanogr. Mar. Biol. Ann. Rev. 21:67-175.Hiraizumi, Y., M. Takahashi and H. Nishimura. 1979. Adsorption of polychlorinatedbiphenyls onto sea bed sediments, marine plankton, and other adsorbing agents.Environ. Sci. Technol. 13:580-584.Hockin, D.C. 1983. The effects of organic enrichment upon a community of meiobenthicharpacticoid copepods. Mar. Env. Res. 10:45-58.Holby, 0. and P.O.J. Hall. 1991. Chemical fluxes and mass balances in a marine fish cagefarm. II. Phosphorous. Mar. Ecol. Prog. Ser. 70:263-272.Hul, M. 1987. The effect of domestic sewage on the structure of the microbenthic ciliatecommunities in the Lyna River. Pol. Arch. Hydrobiol. 34(4): 567-578.Hul, M. 1988. Ciliate communities in the middle sector of the River Lyna (North EasternPoland) in conditions of nonpoint pollution inflow. Acta Hydrobiol. 30(3/4): 353-366.Husevag, B., B.T. Lunestad, P.J. Johannessen, 0. Enger and O.B. Samuelsen. 1991.Simultaneous occurrence of Vibrio salmonicida and antibiotic resistant bacteria insediments at abandoned aquaculture sites. J. Fish Dis. 14: 14:631-640.Hustvedt, S.O., T. Storebakken and R. Salte. 1991. Does Oral Administration of Oxolinicacid or Oxytetracycline Affect Feed Intake of Rainbow Trout? Aquaculture. 92:109-113.Inglis, V. and R.H. Richards. 1991. The in vitro susceptibility of Aeromonas salmonicidaand other fish-pathogenic bacteria to 29 antimicrobial agents. J. Fish Dis. 14:641-650.Iwama, G.K. 1991. Interactions between aquaculture and the environment. Critic. Rev.Environ. Control. 21(2):177-216.Jacobsen, M.D. 1989. Withdrawal Times of Freshwater Rainbow Trout, Salmo gairdneriRichardson, after Treatment with Oxolinic Acid, Oxytetracycline and Trimetoprim.J. Fish Dis. 12:29-36.Jacobsen, P. and L. Berglind. 1988. Persistence of oxytetracycline in sediments from fishfarms. Aquaculture. 70:365-370.Jaffe, R. 1991. Fate of hydrophobic organic pollutants in the aquatic environment: Areview. Environment. Pollut. 69:237-257.117Jensen, P. 1987. Feeding ecology of free-living aquatic nematodes. Mar. Ecol. Prog. Ser.35:187-196.Johnson, M. 1990. Major and Minor metal loading of surface sediments in VancouverHarbour, Second Narrows and Port Moody. B.Sc. Dept. Oceanography. Universityof British Columbia (Canada).Joint, I.R., J.M. Gee and R.M Warwick. 1982. Determination of fine-scale verticaldistribution of microbes and meiofauna in an intertidal sediment. Mar. Biol. 72:157.Jones, J.G. 1979. Microbial activity in lake sediments with particular reference to electrodepotential gradients. J. Gen. Microbiol. 115: 19-26.Jones, G.E. 1980. Biogeochemical succession of bacterial activities in aquatic sediments.Microbiology. pp:348-349.Jorgensen, B.B. 1977a. Distribution of colorless sulfur bacteria (Beggiatoa spp.) in acoastal marine sediment. Mar. Biol. 41:19-28.Jorgensen, B.B. 1977b. The sulfur cycle of a coastal marine sediment. Limnol. Oceanog.22(5):814832.Jorgensen, B.B. 1980. Seasonal oxygen depletion in the bottom waters of a Danish fjordand its effect on the benthic community. Oikos. 34:68-76.Jorgensen, B.B. 1989. Sulfate reduction in marine sediments from the Baltic Sea-North SeaTransition. Ophelia. 31(1):1-15.Jorgensen, B.B. and T. Fenchel. 1974. The sulfur cycle of a marine sediment modelsystem. Mar. Biol. 24:189-201.Jorgensen, B.B. and N.P. Revsbech. 1983. Colorless sulfur bacteria, Beggiatoa spp. andThiovulum spp., in 02 and H2S microgradients. Appl. Environ. Microbiol.45(4): 1261-1270.Jorgensen, B.B. and N.P. Revsbech. 1989. Oxygen uptake, bacterial distribution, andcarbon-nitrogen-sulfur cycling in sediments from the Baltic Sea - North Seatransition. Ophelia. 31(1):29-49.Jorgenson, B.B., M.Bang, T.H. Blackburn. 1990. Anaerobic mineralization in marinesediments from the Baltic sea-North Sea transition. Mar. Ecol. Prog. Ser. 59:39-54.Jumars, P.A., D.L. Penry, J.A. Baross, M.J. Perry and B.W. Frost. 1989. Closing themicrobial loop: Dissolved carbon pathway to heterotrophic bacteria from incompleteingestion, digestion and absorption in animals. Deep sea Res. 36(4):483-495.Kaspar, H.F., P.A. Gillespie, I.C. Boyer and A.L. MacKenzie. 1985. Effects of musselaquaculture on the nitrogen cycle and benthic communities in Kenepuru Sound,Marlborough Sounds, New Zealand. Mar. Biol. 85:127-136.Kaspar, H.F., G.H. Hall and A.J. Holland. 1988. Effects of sea cage salmon farming onsediment nitrification and dissimilatory nitrate reductions. Aquaculture. 70:333-344.118Kelly, J.R. and S.W. Nixon. 1984. Experimental studies of the effect of organic depositionon the metabolism of a coastal marine bottom community. Mar. Ecol. Prog. Ser.17:157-169.Kelly, M.T. and A. Dinuzzo. 1985. Uptake and clearance of Vibrio vulnificus from GulfCoast Oyster (Crassostrea virginica). Appl. Environ. Microbiol. 50(6):1548-1549.Kelly, M.T. and E.M.D. Stroh. 1988. Occurrence of Vibrionaceae in natural and cultivatedoyster populations in the Pacific Northwest. Diagn. Microbiol. Infect. Dis. 9:1-5.Kemp, P.F. 1987. Potential impact on bacteria of grazing by a macrofaunal deposit feeder,and the fate of bacterial production. Mar. Ecol. Prog. Ser. 36:151-161.Kemp, P.F. 1988. Bacterivory by benthic ciliates: significance as a carbon source andimpact on sediment bacteria. Mar. Ecol. Prog. Ser. 49:163-169.Kemp, P.F. 1990. The fate of benthic bacterial production. Rev. Aquat. Sci. 2(1):109-124.Kinner, N.E. and C.R. Curds. 1987. Development of protozoan and metazoan communitiesin rotating biological contactor biofilms. Wat. Res. 21(4):481-490.Kirby, H. 1950. Materials and methods in the study of protozoa. University of CaliforniaPress. 72p.Korman, J. 1989. Enriching effects of salmon farms in British Columbian coastal watersand the influence of flushing and seasonality. M.Sc. Thesis. Dept. Oceanography,University of British Columbia. 94 pp.Kosciuszko, H. and M. Prajer. 1990. Effect of municipal and industrial pollution on thebiological and chemical quality of the water in the upper and middle courses of theriver Biala Przemsza (southern Poland). Acta Hydrobiol. 32(1/2):13-26.Kozloff, E.N. 1987. Marine Invertebrates of the Pacific Northwest. University ofWashington Press. 511 pp.Kristensen, E., M. Holmer and N. Bussarawit. 1991. Benthic metabolism and sulfatereduction in a southeast Asian mangrove swamp. Mar. Ecol. Prog. Ser. 73:93-103.Krone, R.B. 1978. Aggregations of Suspended Particles in Estuaries. IN: EstuarineTransport Processes. Kjerfve, B. (ed.). University of S. Carolina Press. p:177-190.Kuenen, J.G. 1989. Comparative ecophysiology of the nonphototrophic sulfide oxidizingbacteria. p:349-365. IN: Cohen, Y. and E. Rosenberg (ed.). Microbial mats.American Society of Microbiology.Kuenen, J.G., L.A. Robertson and H. van Gemerden. 1985. Microbial interactions amongaerobic and anaerobic sulfur-oxidizing bacteria. Marshall, K.C. (ed). Adv. Microb.Ecol. 8:1-54.Kuuppo-Leinikki, P. 1990. Protozoan grazing on planktonic bacteria and its impact onbacterial population. Mar. Ecol. Prog. Ser. 63:227-238.Kwandrans, J. 1988. A regulated river ecosystem in a polluted section of the UpperVistula. 6. Communities of sessile algae. Acta Hydrobiol. 30(1/2):61-71.119Lackey, J.B. 1961. Bottom sampling and environmental niches. Limnol. Oceanogr. 6:271-279.Lasserre, P. 1990. Marine Microcosms: Small scale controlled ecosystems. IN: Lalli, C.M.(ed.) Enclosed experimental marine ecosystems: A review and recommendations.Coast. Estuar. Studies. 37. Springer-Verlag.Legall, J. and G. Fauque. 1988. Dissimilatory reduction of sulfur compounds. IN:Zehnder, A.J.B. (ed.) Biology of anaerobic microorganisms. John Wiley & Sons. p:587-640.Lee, J.J., E.B.Small, D.H.Lynn and E.C. Bovee. 1985. Some techniques for collecting,cultivating and observing protozoa. IN: Lee, J.J., S.H. Hutner and E.C. Bovee(ed.) An Illustrated guide to the Protozoa. Allen Prees, Inc.Levings, C.D. (1992). Using coastal habitats for pen rearing salmon in British Columbia:status of current knowledge. J. Appl. Aqua. (In Press).Levy, S.B. 1984. Resistance to the tetracyclines. IN: Bryan, L.E. (ed.) Antimicrobial drugresistance. Academic Press Inc. p:192-241.Levy, S.B. 1988. Tetracycline resistance determinants are widespread. ASM News. 54(8):418-421.Lighthart, B. 1969. Planktonic and benthic bacteriovorous protozoa at eleven stations inPuget Sound and adjacent Pacific Ocean. J. Fish. Res. Bd. Can. 26:299-304.Lopez, G.R. and J.S. Levinton. 1987. Ecology of deposit feeding animals in marinesediments. Quart. Rev. 62(3):235-260.Lumb, C.M. 1989. Self pollution by Scottish Salmon Farms. Mar. Poll. Bull. 20(8):375-379.Lumb, C.M. and S.L. Fowler. 1989. Assessing the benthic impact of fish farming.McManus, J. and M. Elliott (ed). Develop. Estuar. Coast. Study Techniq. EBSA 17Symposium. p:75-78.Lunestad, B.T. and J. Goksoyr. 1990. Reduction in the antibacterial effect ofoxytetracycline in sea water by complex formation with magnesium and calcium.Dis. Aquat. Org . 9:67-72.Madoni, P. and P.F. Ghetti. 1981. The structure of ciliated protozoa communities inbiological sewage treatment plants. Hydrobiologia. 83:207-215.Mallory, L.M., C.S. Yuk, L.N. Liang and M. Alexander. 1983. Alternative prey: Amechanism for elimination of bacterial species by protozoa. Appl. Environ. Microb.46(5): 1073-1079.Mann, K.H. 1988. Production and use of detritus in various freshwater, estuarine, andcoastal marine ecosystems. Limnol. Oceanogr. 33(4):910-930.Manru, G., F. Weisong and S. Yunfen. 1988. Ecological study on protozoa in the sedimentof the Three-Gorges area of the Changjiang river. Chin. J. Oceanol. Limnol.6(3):272-280.120Marshall, K.C., R. Stout and R. Mitchell. 1971. Mechanism of the initial events in thesorption of marine bacteria to surfaces. J. Gen. Microbiol. 68:337-348.Marty, D. 1981. Distribution of different anaerobic bacteria in Arabian Sea sediments.Mar. Biol. 63:277.Mattson, J. and 0. Linden. 1983. Benthic macrofauna succession under mussels, Mytilusedulis, cultured on hanging long lines. Sarsia. 68:97-102.McIntyre, A.D. and R.M. Warwick. 1984. Meiofauna techniques. IN: Holme, N.A. andA.D. McIntyre (ed). Methods for the study of marine benthos. 2nd ed. BlackwellScientific Publications. IBP handbook 16. p:217-244.McPhearson, R.M., A. DePaolo, S.R. Zwyno, M.L. Motes Jr. and A.M. Guarino. 1991.Antibiotic resistance in Gram-negative bacteria from cultured catfish and aquacultureponds. Aquaculture. 99:203-211.Meyer-Reil, L.A. and A.Faubel. 1980. Uptake of organic matter by meiofauna organismsand interrelationships with bacteria. Mar. Ecol. Prog. Ser. 3:251-256.Mir, J., M. Martinez-Alonso, I. Esteve and R. Guerrero. 1991. Vertical stratification andmicrobial assemblage of a microbial mat in the Ebro Delta (Spain). FEMSMicrobiol. Letters. 86:59-68.Montagna, P.A. 1982. Sampling design and enumeration statistics for bacteria extractedfrom marine sediments. Appl. Environ. Microbiol. 43(6):1366-1372.Montagna, P.A. 1984. In situ measurement of meiobenthic grazing rates on sedimentbacteria and edaphic diatoms. Mar. Ecol. Prog. Ser. 18:119-130.Montagna, P.A., B.C. Coull, T.L. Herring, B.W. Dudley. 1983. The relationship betweenabundances of meiofauna and their suspected microbial food (Diatoms and Bacteria).Est. Coast. Shelf. Sci. 17:381-394.Moore, D.C. and G.K. Rodger. 1991. Recovery of a sewage sludge dumping ground. II.Macrobenthic community. Mar. Ecol. Prog. Ser. 75:301-308.Moriarty, D.J.W. 1986. Measurement of bacterial growth rates in aquatic systems fromrates of nucleic acid synthesis. Marshall, K.C. (ed). Adv. Microb. Ecol. 9:2445-191.Munro, A.L.S. 1990. Salmon Farming. Fish. Res. 10:151-161.Montagnes, D.J.S. and D.H. Lynn. 1987. A quantitative protargol stain (QPS) for ciliates:method description and test of its quantitative nature. Mar. Microb. Food Webs.2(2):83-93.Nedwell, D.B. 1982. The cycling of sulphur in marine and freshwater sediments. IN:Nedwell, D.B. and C.M. Brown (ed.) Sediment microbiology. Academic Press.p:73-106.Nedwell, D.B. 1984. The input and mineralization of organic carbon in anerobic aquaticsediments. Marshall, K.C. (ed). Adv. Microb. Ecol. 7:93-124.121Nedwell, D.B. and P.A. Lawson. 1990. Degradation of digested sewage sludge in marinesediment-water model systems, and fate of metals. Mar. Poll. Bull. 21(2):87-91.Nelson, D.C. and R.W. Castenholz. 1981. Use of reduced sulfur compounds by Beggiatoaspp. J.Bacteriol. 147(1):140-154.Nelson, D.C., B.B. Jorgensen and N.P. Revsbech. 1986. Growth pattern and yield of achemoautotrophic Beggiatoa spp. in oxygen-sulfide microgradients. Appl. Environ.Microbiol. 52(2):225-233.Neu, H.C. and A.D.S. Caldwell. 1978. (Eds.) Problems of Antibiotic Therapy. AcademicPress. 87 pp.Newell, S.Y and R.D. Fallon. 1982. Bacterial productivity in the water column andsediments of the Georgia (U.S.A.) coastal zone: Estimates via direct counting andparallel measurement of thymidine incorporation. Microb. Ecol. 8:33-46.Nilsson, J.R. 1989. Tetrahymena in cytotoxicology: with special reference to effects ofheavy metals and selected drugs. Europ. J. Protistol. 25:2-25.Nilsson, P., B. Jonsson, I.L. Swanberg and K. Sundback. 1991. Response of a marineshallow water sediment system to an increased load of inorganic nutrients. Mar.Ecol. Prog. Ser. 71:275-290.Novitsky, J.A. and D.M. Karl. 1986. Characterization of microbial activity in the surfacelayers of a coastal sub-tropical sediment. Mar. Ecol. Prog. Ser. 28:49.Nygaard, K., B.T. Lunestad, H. Hektoen, J.A. Berge and V. Hormazabal. 1992.Resistance to oxytetracycline, oxolinic acid and furazolidone in bacteria from marinesediments. Aquaculture. (In Press).O'Connor, B.D., J. Costelloe, B.F. Keegan and D.C. Rhoads. 1989. The use of REMOTStechnology in monitoring coastal enrichment resulting from mariculture. Mar. Poll.Bull. 20(8):384-390.Oka, H., H. Matsumoto, K. Uno, K. Harada, S. Kadowaki and M. Suzuki. 1985.Improvement of Chemical Analysis of Antibiotics. VIII. Application of PrepackedC 18 Cartridge for the Analysis of Tetracycline Residues in Animal Liver. J.Chromat. 325:265-274.Oka, H., Y. Ikai, N. Kawamura, M. Yamada, K. Harada, S. Ito and M. Suzuki. 1989.Photodecomposition products of Tetracycline in aqueous solution. J. Agric. Foodchem. 37(1):226-231.Onji, Y., M. Uno and K. Tanigawa. 1984. Liquid Chromatographic Determination ofTetracycline Residues in Meat and Fish. J. Assoc. Off Anal. Chem. 67(6):1135.Oviatt, C.A., J.G. Quinn, J.T. Maughan, J.T. Ellis, B.K. Sullivan, J.N. Gearing, P.J.Gearing, C.D. Hunt, P.A. Sampou and J.S. Latimer. 1987. Fate and effects ofsewage sludge in the coastal marine environment: a mesocosm experiment. Mar.Ecol. Prog. Ser. 41:187-203.Parker, J.G. 1981. Ciliated protozoa of the polluted Tees Estuary. Estuar. Coast. Shelf Sci.12: 337-340.122Parsons, T.R., Y. Maita and C.M. Lalli. 1984a. A manual of chemical and biologicalmethods for seawater analysis. Pergamon Press. 174 pp.Parsons, T.R., M. Takahashi and B. Hargrave. 1984b. Biological oceanographic processes.3rd ed. Pergamon Press. 330 pp.Parsons, T.R., B.E. Rokeby, C.M. Lalli and C.D. Levings. 1990. Experiments on theeffect of salmon farm wastes on plankton ecology. Bull. Plank. Soc. Jap. 37(1):49-57.Patterson, D.J. and T. Fenchel. 1990. Massisteria marina Larsen & Patterson 1990, awidespread and abundant bacterivorous protis associated with marine detritus. Mar.Ecol. Prog. Ser. 62:11-19.Pearson, T.H. 1975. The benthic ecology of Loch Linhe and Loch Eil, a sea loch systemon the west coast of Scotland. IV. Changes in the benthic fauna attributable toorganic enrichment. J. Exp. Mar. Biol. Ecol. 20:1-41.Pearson, T.H. 1982. The Loch Eil Project: Assessment and synthesis of a discussion ofcertain biological questions arising from a study of the organic pollution ofsediments. J. Exp. Mar. Biol. Ecol. 57:93-124.Pearson, T.H. and R. Rosenberg. 1978. Macrobenthic succession in relation to organicenrichment and polloution of the marine environment. Oceanogr. Mar. Biol. Ann.Rev. 16:229-311.Pearson, T.H. and S.O. Stanley. 1979. Comparative measurement of the redox potential ofmarine sediments as a rapid means of assessing the effect of organic pollution. Mar.Biol. 53:371-379.Pearson, T.H., G. Duncan and J. Nuttall. 1982. The Loch Eil Project: populationfluctuations in the macrobenthos. J. Exp. Mar. Biol. Ecol. 56:305-321.Phillips, M.J., M.C.M. Beveridge and L.G. Ross. 1985. The environmental impact ofsalmonid cage culture on inland fisheries: present status and future trends. J. FishBiol. (Suppl.A). 27:123-137.Pilson, M.E.Q. 1990. Application of mesocosms for solving problems in pollutionresearch. IN: Lalli, C.M. (ed.) Enclosed experimental marine ecosystems: A reviewand recommendations. Coast. Estuar. Studies. 37. Springer-Verlag. p:155-168.Pike, E.B. and C.R. Curds. 1971. The microbial ecology of the activated sludge process.IN: Sykes, G. and F.A. Skinner (ed.) Microbial aspects of pollution. AcademicPress. p:123-147.Plante, C.J., P.A. Jumars and J.A. Baross. 1989. Rapid Bacterial growth in the hindgut ofa marine deposit feeder. Microb. Ecol. 18:29-44.Postgate, J.R. 1979. The sulphate reducing bateria. Cambridge Univeristy Press. 151 pp.Powell, E. 1989. Oxygen, sulfide and diffusion: Why thiobiotic meiofauna must be sulfideinsensitive first order respirers. J. Mar. Res. 47:887-932.Pratt, J.R. and J. Cairns, Jr. 1985. Functional groups in the protozoa: roles in differingecosystems. J. Protozool. 32(3):415-423123Pratt, J.R., R. Horwitz and J. Cairns, Jr. 1987. Protozoan communities of the Flint River-Lake Blackshear ecosystem (Georgia, U.S.A.). Hydrobiologia. 148:159-174.Pritchard, P.H. and A.W. Bourqin. 1984. The use of microcosms for evaluation ofinteractions between pollutants and microorganisms. Adv. in Microb. Ecol. 7:133-215.Qureshi, A.A. and M.A. Qureshi. 1992. Multiple antibiotic resistant fecal coliforms in rawsewage. Water, Air and Soil Poll. 61:47-56.Raffaelli, D.G. 1987. The behaviour of the nematode/copepod ratio in organic pollutionstudies. Mar. Env. Res. 23:135-152.Raffaelli, D.G. and C.F. Mason. 1981. Pollution monitoring with meiofauna, using theratio of nematodes to copepods. Mar. Poll. Bull. 12(5):158-162.Reape, T.J. and A.M. Burnell. 1991. Dauer larva recovery in the nematode Caenorhabditiselegans. II. The effect of inhibitors of protein synthesis on recovery, growth andpharyngeal pumping. Comp. Biochem. Physiol. 98B(2/3):245-252.Reimers, C.E., S. Kalhorn, S.R. Emerson and K.H. Nealson. 1984. Oxygen consumptionrates in pelagic sediments from the Central Pacific: first estimates frommicroelectrode profiles. Geochimica et Cosmochimica Acta. 48:903-910.Revsbech, N.P. and B.B. Jorgensen. 1986. Microelectrodes: Their use in microbialecology. Marshall, K.C. (ed). Adv. Microb. Ecol. 9:293-351.Revsbech, N.P., P.B. Christensen and L.P. Nielsen. 1989. Microelectrode analysis ofphotosynthetic and respiratory processes in microbial mats. p:153-162. IN: Y.Cohen and E. Rosenberg (ed.) Microbial Mats. American Society of Microbiology.Ricker, W.E. 1984. Computation and Uses of Central Trend Lines. Can. J. Zool. 62:1897-1905.Riddle, M.J., D.M. Alongi, P.K. Dayton, J.A. Hansen and D.W. Klumpp. 1990. Detritalpathways in a coral reef lagoon. I. Macrofaunal biomass and estimates ofproduction. Mar. Biol. 104:109-118.Ridler, N.B. 1984. Socioeconomic aspects of sea cage salmon farming in the maritimes.Can. J. Fish. Aquat. Sci. 41:1490-1495.Rieper, M. 1985. Some lower food web organisms in the nutrition of marine harpacticoidcopepods: an experimental study. Helgol. Meeres. 39:357-366.Ritz, D.A., M.E. Lewis and M. Shen. 1989. Response to organic enrichment of infaunalmacrobenthic communities under salmonid seacages. Mar. Biol. 103:211-214.Rogstad, A., V. Hormazabal, O.F. Ellingsen and K.E. Rasmussen. 1991. PharmacokineticStudy of Oxytetracycline in Fish. I. Absorption, Distribution and Accumulation inRainbow Trout in Freshwater. Aquaculture. 96: 219-226.Rosenberg, M. and S. Kjelleberg. 1986. Hydrophobic interactions: Role in bacterialadhesion. Marshall, K.C. (ed.) Adv. Microb. Ecol. 9:353-393.124Salte, R. 1982. Oxytetracycline Residues in Rainbow Trout fed a Commercial MedicatedFeed. Acta vet. Scand. 23:150-152.Salte, R. and K. Liestol. 1983. Drug Withdrawal from Farmed Fish: Depletion ofOxytetracycline, Sulfadiazine, and Trimethophrim from Muscular Tissue ofRainbow Trout (Salmo gairdneri). Acta vet. Scand. 24:418-430.Sampou, P. and C.A. Oviatt. 1991. Seasonal patterns of sedimentary carbon and anaerobicrespiration along a simulated eutrophication gradient. Mar. Ecol. Prog. Ser. 72:271-282.Samuelsen, 0.B., V. Torsvik, P.K.Hansen, K. Pittman and A. Ervik. 1988. Organic wasteand antibiotics from aquaculture. J. Cons. int. Explor. Mer. C.M. 1988/F:14 14pp.Samuelsen, O.B. 1989. Degradation of oxytetracycline in seawater at Two differentTemperatures and Light Intensities, and the persistence of Oxytetracycline in thesediment from a fish farm. Aquaculture. 83:7-16.Samuelsen 0.B., E. Solheim and B.T. Lunestad. 1991. Fate and microbiological effects offurazolidone in a marine aquaculture sediment. The Sci. Tot. Environ. 108:275-283.Samuelsen, O.B., B.T. Lunestad, B. Husevag, T. Holleland and A. Ervik. 1992. Residuesof oxolinic acid in wild fauna following medication in fish farms. Dis. Aquat. Org .(In press).Sanders, R.W. 1987. Transfer efficiencies of microbial carbon to higher trophic levels: theheterotrophic flagellate - crustacean link. EOS. 68(50):1705.Scheraga, M., M. Meskill and C.D. Litchfield. 1979. Analysis of methods for thequantitative recovery of bacteria sorbed onto marine sediments. IN: Methodology ofbiomass determinations and microbial activities in sediments. Litchfield, C.D. andP.L. Seyfried (ed.) ASTM STP 673. p:21-39.Schwinghamer, P. 1981. Extraction of living meiofauna from marine sediments bycentrifugation in a silica sol-sorbitol mixture. Can. J. Fish. Aquat. Sci., 38:476-478.Schwinghamer, P. 1988. Influence of pollution along a natural gradient and in a mesocosmexperiment on sediment microbial numbers and biomass. Mar. Ecol. Prog. Ser.46:193-197.Seymour, E.A. and A. Bergheim. 1991. Towards a reduction of pollution from intensiveaquaculture with reference to the farming of salmonids in Norway. Aquacult.Engineer. 10:73-88.Shilo, M. 1989. The unique characteristics of benthic cyanobacteria. IN: Microbial Mats:Physiological ecology of benthic microbial communities. Cohen, Yl. and E.Rosenberg (ed). American Society for Microbiology. Wahington, DC. p:207-213.Sindermann, C.J. 1984. Disease in marine aquaculture. Helgol. Meeres. 37:505-532.Skyring, G.W. and J. Bauld. 1990. Microbial mats in Australian coastal environments.Marshall, K.C. (ed). Adv. Microb. Ecol. 11:461-498.125Sladecek, V. 1973. System of water quality from the biological point of view. Arch.Hydriobiol. Beih. 7:1-218.Small, E.B. 1973. A study of ciliate protozoa from a small polluted stream in East-CentralIllinois. Amer. Zool. 13:225-230.Smock, L.A. and K.L. Harlowe. 1983. Utilization and processing of freshwater wetlandmacrophytes by the detritivore Asellus forbesi. Ecology. 64(6):1556-1565.Stanley, S.O., J.W. Leftley, A. Lightfoot, N. Robertson, I.M. Stanley and I. Vance. 1981.The Loch Eil Project: sediment chemistry, sedimentation and the chemistry of theoverlying water in Loch Eil. J. Exp. Mar. Biol. Ecol. 55:299-313.Stirling, H. P. and T. Dey. 1990. Impact of intensive cage fish farming on thephytoplankton and periphyton of a Scottish freshwater loch. Hydrobiol. 190:193-214.Stoecker, D.K. and J.M. Capuzzo. 1990. Predation on protozoa: its importance tozooplankton. J. Plank. Res. 12(5):891-908.Stoessel, F. 1989. On the ecology of ciliates in riverwaters: The evaluation of water qualityvia ciliates and filamentous bacteria. Aquatic Sci. 51(3):235-248.Stout, J.D. 1981. The role of protozoa in nutrient cycling and energy flow. Adv. Microb.Ecol. 4:1-50.Strohi, W.R. and J.M. Larkin. 1978. Enumeration, isolation, and characterization ofBeggiatoa from freshwater sediments. Appl. Environ. Microbiol. 36(5):755-770.Sundback, K., V. Enoksson, W. Graneli and K. Pettersson. 1991. Influence of sublittoralmicrophytobenthos on the oxygen and nutrient flux between sediment and water: alaboratory continuous flow study. Mar. Ecol. Prog. Ser. 74:263-279.Svealv, T. 1991. Strategies and technologies in offshore farming. Fisheries Research.10:329-349.Tenore, K.R. and B.C. Coull. 1980. (ed.). Marine Benthic Dynamics. University of SouthCarolina Press. 457 pp.Tenore, K.R., L. Cammen, S.E.G. Findlay and N. Phillips. 1982. Perspectives of researchon detritus: do factor controlling the availability of detritus to macroconsumersdepend on its source? J. Mar. Res. 40:473-490.Terzaghi, E. and M. O'Hara. 1990. Microbial Plasticity: The relevance to microbialecology. Marshall, K.C. (ed). Adv. Microb. Ecolog. 11:431-460.The Supelco Reporter. 1985. Isocratic Analysis of Tetracycline Antibiotics Using aSupelcosil LC-18-DB HPLC Column. IV (1):3.Thode-Andersen, S. and B.B. Jorgensen. 1989. Sulfate reduction and the formation of S 35labeled FeS, FeS2 and S° in coastal marine sediments. Limnol. Oceanogr.34(5): 793-806.Tietjen, J.H. 1980. Microbial-meiofaunal interrelationships: A review. Aq. Microb.Ecolog. pp:335-338.126Toerien, D.F., A. Gerber, L.H. Lotter and T.E. Cloete. 1990. Enhanced biologicalphosphorous removal in activated sludge systems. Marshall, K.C. (ed). Adv.Microb. Ecolog. 11:173-230.Toranzo, A.E., P. Combarro, M.L. Lemos and J.L. Barja. 1984. Plasmid coding fortransferable drug resistance in bacteria isolated from cultured Rainbow Trout. Appl.Environ. Microbiol. 48(4): 872-877.Torsvik, V.L., R. Sorheim and J.Goksoyr. 1988. Antibiotic resistance of bacteria from fishfarm sediments. J. Cons. int. Explor. Mer. C.M.1988/F:10 9pp.Tremaine, S.C. and A.L. Mills. 1987. Inadequacy of the eucaryote inhibitor cyclohexamidein studies of protozoan grazing on bacteria at the freshwater-sediment interface.Appl. Environ. Microbiol. 53(8): 1969-1972.Trevors, J.T., T. Barkay and A.W. Bourquin. 1986. Gene transfer among bacteria in soiland aquatic environments: a review. Can. J. Microbiol. 33:191-198.Tsutsumi, H. 1987. Population dynamics of Capitella capitata (polychaeta) in anorganically polluted cove. Mar. Ecol. Prog. Ser. 36:139-149.Tsutsumi, H. 1990. Population persistence of Capitella spp. on a mud flat subject toenvironmental disturbance by organic enrichment. Mar. Ecol. Prog. Ser. 63:147-156.Tsutsumi, H., S. Fukunaga, N. Fujita and M. Sumida. 1990. Relationship between growthof Capitella spp. and organic enrichment of the sediment. Mar. Ecol. Prog. Ser.63:157-162.Uhlig, G., H. Thiel and J.S. Gray. 1973. The quantitative separation of meiofauna: Acomparison of methods. Helg. wiss. Meeres. 25:173-195.Underwood, A.J. and C.H. Peterson. 1988. Towards an ecological framework forinvestigating pollution. Mar. Ecol. Prog. Ser. 46:227-234.Van Loosdrecht, M.C.M., J. Lyklema, W. Norde and A.J.B. Zehnder. 1989. BacterialAdhesion: A Physicochemical Approach. Microb. Ecol. 17:1-15.Velji, M.I. and L.J. Albright. 1986. Microscopic enumeration of attached marine bacteriaof seawater, marine sediment, fecal matter, and kelp blade samples followingpyrophosphate and ultrasound treatments. Can. J. Microbiol. 32:121-126.Venrick, E.L. 1971. The statistics of subsampling. Limnol. Oceanogr. 16(5):811-818.Vismann, B. 1991. Sulfide tolerance: physiological mechanisms and ecologicalimplications. Ophelia. 34(1): 1-27.Warwick, R.M., T.H. Pearson and A. Ruswahyumi. 1987. Detection of pollution effectson marine macrobenthos: further evaluation of the species abundance/biomassmethod. Mar. Biol. 95:193-200.Warwick, R.M., M.R. Carr, K.R. Clarke, J.M. Gee and R.H. Green. 1988. A mesocosmexperiment on the effects of hydrocarbon and copper pollution on a sublittoral soft-sediment meiobenthic community. Mar. Ecol. Prog. Ser. 46:181-191.127Warwick, R.M., H.M. Platt, K.R. Clarke, J. Agard and J. Gobin. 1990. Analysis ofmacrobenthic and meiobenthic community structure in relation to pollution anddisturbance in Hamilton Harbour, Bermuda. J.Exp. Mar. Biol. Ecol. 138:119-142.Weglenska, T., L. Bownik-Dylinska, J. Ejsmont-Karabin and I. Spodniewska. 1987.Plankton structure and dynamics, phosphorous and nitrogen regeneration byzooplankton in Lake Glebokie polluted by aquaculture. Ekol. Pol. 35(1): 173-208.Weisse, T. and U. Scheffel-Moser. 1991. Uncoupling the microbial loop: growth andgrazing loss rates of bacteria and heterotrophic nanoflagellates in the North Atlantic.Mar. Ecol. Prog. Ser. 71:195-205.Weise, W. and G. Rheinheimer. 1978. Scanning electron microscopy and epifluorescenceinvestigation of bacterial colonization of marine sand sediments. Microb. Ecol.4:175-188.Weston, D.P. 1986. The environmental effects of floating mariculture in Pugest Sound.Washington Department of Fisheres. 148 pp.Weston, D.P. 1990. Quantitative examination of macrobenthic community changes alongan organic enrichment gradient. Mar. Ecol. Prog. Ser. 61:233-244.White, E.R., M.A. Carroll, J.E. Zarembo and A.D. Bender. 1975. Reverse Phase HighSpeed Liquid Chromatography of Antibiotics. J. Antibiotics. 28(3):205-214.Whitehurst, I.T. and B.I. Lindsey. 1990. The impact of organic enrichment on the benthicmacroinvertebrate communities of a lowland river. Wat. Res. 24(5): 625-630.Whitfield, M. 1969. Eh as an operational parameter in estuarine studies. Limnol. Oceanogr.14:547-558.Widbom, B. and R. Elmgren. 1988. Response of benthic meiofauna to nutrient enrichmentof experimental marine ecosystems. Mar. Ecol. Prog. Ser. 42:257-268.Widdell, F. 1988. Microbiology and ecology of sulfate and sulfur reducing bacteria. IN:Zehnder, A.J.B. (ed.) Biology of anaerobic microorganisms. John Wiley & Sons.p:469-586.Wiggins, B.A. and M. Alexander. 1988. Role of protozoa in microbial acclimation formineralization of organic chemicals in sewage. Can. J. Microbiol. 34:661-666.Wiggins, B.A., S.H. Jones and M. Alexander. 1987. Explanations for the acclimationperiod preceding the mineralization of organic chemicals in aquatic environments.Appl. Environ. Microbiol. 53(4):791-796.Windsor, M.L. and P. Hutchinson. 1990. The potential interactions between salmonaquaculture and the wild stocks - a review. Fish. Res. 10:163-176.Wirsen, C.O. and H.W. Jannasch. 1978. Physiological and morphological observations onThiovulum sp. J. Bacteriol. 136(2):765-774.Wyatt, C.E. and T.H. Pearson. 1982. The Loch Eil Project: population characteristics ofciliate protozoans from organically enriched sea loch sediments. J. Exp. Mar. Biol.Ecol. 56:279-303.128Ye, Li-Xun, D.A. Ritz, G.E. Fenton and M.E. Lewis. 1991. Tracing the influence onsediments of organic waste from a salmonid farm using stable isotope analysis. J.Exp. Mar. Biol. Ecol. 145:161-174.Zar, J.H. 1984. Biostatistical Analysis. 2nd ed. Prentice-Hall, Inc. 718pp.Zehnder, A.J.B. and W. Stumm. 1988. Geochemistry and biogeochemistry of anaerobichabitats. IN: Zehnder, A.J.B. (ed.) Biology of anaerobic microorganisms. JohnWiley & Sons Inc. p:1-38.ZoBell, C.E. 1946. Studies on redox potential of marine sediments. Bull. Am. Ass. Petrol.Geolog. 30(4):477-513.129

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0086466/manifest

Comment

Related Items