Tfia® TSISH© ©M %ID March 1980 ^ of c 0a M^ / \$ \ Prepared for: * U.S. Department of Commerce 7 National Oceanic and Atmospheric Administration ^ Outer Continental Shelf Environmental Assessment Program c % 1* S ^TES O* * Cover Photo Species is a copepodid of Temora longicornis . A indicates oil droplets in the gut. B indicates an oil drop on feeding appendage. Courtesy of Sif Johansson , Ask'6 Laboratory . THE THESIS OIL SPILL Report of the first year scientific study (October 26, 1977 to December 1978) Editors: John J. Kineman Outer Continental Shelf Environmental Assessment Program Ragnar Elmgren Department of Zoology, University of Stockholm Sture Hansson Asko Laboratory, University of Stockholm Contributing Authors: Askb Laboratory, Department of Zoology and Botanical Institute University of Stockholm, Sweden Gunnar Aneer Ragnar Elmgren Maria Foberg Bjorn Guterstom Sture Hansson Sif Johansson Hans Kautsky Ulf Larsson Anders Lindhe Sture Nellbring Brita Sundelin Lars Westin Swedish Water and Air Pollution Research Institute (IVL) , Sweden Olle Linden Mats Notini Energy Resources Company, Inc. , United States Paul Boehm Judith Barak David Fiest Adria Elskus NOAA/OCSEAP Spilled Oil Research Team, United States Robert C. Clark, Jr. John Kineman MARCH 1980 ^BORtf* ivl A COOPERATIVE INTERNATIONAL INVESTIGATION BY ASKO LABORATORY, UNIVERSITY OF STOCKHOLM, SWEDEN SWEDISH WATER AND AIR POLLUTION RESEARCH INSTITUTE (iVL) STUDSVIK, SWEDEN I rfSSSSSBfen U.S. DEPARTMENT OF COMMERCE OFFICE OF MARINE POLLUTION ASSESSMENT NATIONAL OCEANIC AND ATMOSPHERIC ADMINISTRATION BOULDER, COLORADO, U.S.A. DISCLAIMER Mention of a commercial company or product does not constitute an endorsement by National Oceanic and Atmospheric Administration. Use for publicity or advertising purposes of information from this publication concerning propriety products or the tests of such products is not authorized. 11 TABLE OF CONTENTS Editors' Note Abstract ......... ix Chapter 1 Executive Summary 3 1 1 Development, early history and cleanup of the spill (Olle Linden and John Kineman) 3 1 2 The scientific study and participating institutions . 12 1 3 Scientific personnel participating in the study . 15 1 4 Acknowledgements . 16 1 5 Funding . . 19 1 6 Lessons learned for the management of oil spill investigations . 20 (Olle Linden, Ragnar Elmgren, Lars Westin and John Kineman) 1 7 References . 26 Chapter 2 Perspective ........ . 29 2 1 Baltic perspective (Ragnar Elmgren, Lars Westin and Olle Linden) . 29 2 2 Alaskan perspective ...... (Ragnar Elmgren and John Kineman) • . 33 2 3 References ........ . 38 Chapter 3 Scientific Summary and General Discussion . 43 (Olle Linden, Ragnar Elmgren, Lars Westin and John Kineman) 3 1 Background . 43 3 2 Discussion of results and major conclusions . 44 3 3 Recommendations for spill research contingency plans . . 50 3 4 References . 56 Chapter 4 Impact of Oil on the Pelagic Ecosystem .... (Sif Johansson) . 61 4 1 Introduction ........ . . . 61 4 2 Material and methods 4.2.1 Sampling stations 4.2.2 Methods 4.2.2.1 Phytoplankton 4.2.2.2 Primary production 4.2.2.3 Bacteria . 4.2.2.4 Zooplankton 4.2.2.5 Sedimentation . 62 . 62 . 62 . 62 . 64 . 64 . 64 . 65 4 .3 Results 4.3.1 Phytoplankton .... . 65 . 65 4.3.1.1 Phytoplankton biomass . 65 4.3.1.2 Phytoplankton species conn >osi1 :ion . 65 Page vii in 4.3.2 Primary production . 4.3.3 Bacteria . 4 r3 . 4 Zooplankton 4.3.5 Sedimentation ^: 4 4 Discussion ...... 4 5 Conclusions ...... 4 6 Acknowl edgements . 4 7 References ...... Chapter 5 5.1 5.2 5.3 5.4 5.5 5.6 Chapter 6 6.1 6.2 6.3 6.4 6.5 6.6 6.7 6.8 NOAA Acute Phase Experiments on Pelagic and Surface Oil (John Kineman, 5.1, 5.3-5.5; Robert C Clark, Jr. , 5.2) Experiment design .... Hydrocarbon analysis associated with the Tsesis oil spill Water samples with a sterile bag sampler ..... Water column data from two scuba-obtained glass jar samples 5.4.1 Results of analysis 5.4.2 Discussion .......... 5.4.3 Conclusion .......... Aerial mapping of visible floating oil ..... References . . 67 67 70 70 74 77 77 79 83 83 84 87 90 90 90 92 93 93 Impact of Oil on Deep Soft Bottoms ....... 97 (Ragnar Elmgren, Sture Hansson, Ulf Larsson and Brita Sundelin) Introduction 6.1.1 6.1.2 Methods 6.2.1 6.2.2 6.2.3 Results 6.3.1 6.3.2 Background .... The spill area . Sediment sampling Macrofauna .... 6.2.2.1 Macrofauna sampling 6.2.2.2 Reproduction of Pontoporeia af finis Meiofauna .... Sediment samples ..... Macrofauna ....... 6.3.2.1 Macrofauna community response 6.3.2.2 Pontoporeia af finis reproduction Meiofauna ....... 6.3.3 Discussion .... 6.4.1 Sediment samples 6.4.2 Macrofauna . 6.4.2.1 Macrofauna community response 6.4.2.2 Pontoporeia aff inis reproduction 6.4.3 Meiofauna Summary and conclusions Suggested further studies Acknowledgement References 97 97 98 98 98 100 100 100 101 101 101 102 102 110 110 117 117 118 118 120 121 122 123 123 124 IV Chapter 7 7.1 7.2 7.3 7.4 Chapter 8 .1 8.2 Impact of Oil on the Littoral Ecosystem Introduction ...... (Mats Notini) Effects on Fucus macrofauna (Mats Notini) 7.2.1 Introduction .... 7.2.2 Materials and methods 7.2.3 Preliminary results 7.2.4 Discussion ..... 7.2.5 References ..... Effects on the phytal ecosystem (Hans Kautsky) 7.3.1 7.3.2 Materials and methods ..... Results 7.3.2.1 Field observations 7.3.2.2 Calculations from collected data 7.3.2.3 Comparison with data from oil ana of Mytilus edulis 7.3.3 Discussion 7.3.4 References ....... In situ respiration of three littoral communities near the Tsesis oil spill ..... (Bjorn Guterstom) 7.4.1 Introduction ...... 7.4.2 Methods 7.4.3 Results and discussion .... 7.4.4 References Impact of Oil on the Supralittoral Zone Damage on shore vegetation ..... (Anders Lindhe) 8.1.1 Introduction 8.1.2 Methods 8.1.3 Results: Acute effects on vegetation 8.1.4 Some comments on the clean-up methods 8.1.5 Concluding remarks . lysi: Effe (Mar 8.2. 8.2. 8.2. 8.2. 8.2. 8.2. 8.2. 8.2. cts on the supralittoral fauna ia Foberg) Introduction Materials and methods Description of the stations Results 8.2.4.1 Pitfall traps 8.2.4.2 Quantitative sampl Discussion .... Conclusions Acknowledgements References .... es 129 129 130 130 131 131 139 144 146 146 146 146 148 156 156 160 161 161 162 162 165 169 169 169 169 170 170 172 173 173 173 178 178 178 181 185 188 188 188 v Chapter 9 9.1 9.2 9.3 9.4 9.5 Chapter 10 10.1 10.2 10.3 10.4 Chapter 11 11.1 11.2 11.3 11.4 11.5 Appendix 1 Appendix 2 Impact of Oil on the Local Fish Fauna ...... 193 (Sture Nellbring, Sture Hansson, Gunnar Aneer and Lars Westin) Introduction . Material and methods Results Discussion References Introduction ............ (Olle Linden) Respiration measurements on the mussel Mytilus edulis (Sture Hansson) 10.2.1 Materials and methods ....... 10.2.2 Results 10.2.3 Conclusions and discussion ...... 10.2.4 References Measurements of byssus formation by the mussel Mytilus edulis (Olle Linden and Maria Foberg) Burrowing behavior in the clam Macoma balthica (Olle Linden) The Analytical Chemistry of Mytilus edulis, Macoma balthica , Sediment Trap and Surface Sediment Samples .... (Paul D. Boehm, Judith Barak, David Fiest and Adria Elskus) Introduction Methods 11.2.1 11.2.2 Results 11.3.1 Sampling Sample analysis 11 11 11 Mytilus edulis . 11.3.1.1 Aliphatic hydrocarbons 11.3.1.2 Aromatic hydrocarbons Sediment traps . Surface sediments Macoma balthica Discussion .... References .... Investigations Scenes from the Tsesis oil spill 193 193 195 200 201 Laboratory Studies Carried Out in Connection with the Spill . 205 205 205 205 206 209 209 210 213 219 219 221 221 221 225 225 225 233 239 251 257 264 272 277 285 VI EDITORS* NOTE Support for the editing and publication of this report was provided by the Outer Continental Shelf Environmental Assessment Program (OCSEAP). Since most of the scientifically important results of the Tsesis study will appear in journal articles, with a more thorough review process, the current document is intended primarily for use by government scien- tists and managers who are involved in environmental issues of oil and gas development, and specifically in decisions and legislation related to oil pollution. It was with the above purpose in mind that this scientific report was organized. It begins with a condensed description of the spill and the study that followed and becomes progressively more detailed. Three levels of detail are represented by the Abstract, Chapter 3, and the Research Reports. Chapter 1, dealing with matters of executive and management concern, is highlighted by section 1.6, which reviews the mistakes and successes and makes recommendations for the future manage- ment of spill follow-up studies and environmental protection. Chapter 2 was added to address such questions as "so what?", or in general the potential overall significance of the incident, by attempting to put it into ecological perspective. The study was heavily supported by OCSEAP with the objective of putting the incident into perspective for Alaska as well as the Baltic, where the spill took place. Since OCSEAP is a goal-directed program, with basic research applying to the environmental assessment of Alaska, the validity of the spill research concept is judged by the value of such connections. Chapter 3 integrates and highlights the most important scientific results and major conclusions of the overall study and places them in a position for quick reference. Although it was not possible to include a separate section detail- ing the daily clean-up operations performed by the Swedish Coast Guard, such information is summarized in Section 1.2. It is impossible to separate the effects of an oil spill from the mitigating or sometimes damaging effects of the cleanup. vn Overall editing of the report was done in numerous locations in Sweden, the United States, and Kenya. Detailed editing of the research reports (Chapters 4-11) was performed mainly in Sweden, although many additions and revisions were made to address comments from the U.S. editor and reviewers of the first draft. Chapter 5 was written and edited in the United States. The report was produced by team effort, with information integrated from various sources. Individual contribu- tions could not always be credited separately. The editors are responsible for any omissions, ambiguities and technical errors and, in sections without named authors, also for factual errors and conclusions, whereas the authors of Chapters 4-11 are responsible for the scientific content of their reports. For the purpose of identifying any such problems, feedback would be appreciated from anyone who is inclined to comment . Address comments to: John J. Kineman Attn: % Editor (Rx4) NOAA-OMPA-OCSEAP 1790 30th Street Boulder, Colorado 80303 303-499-1000 ext. 6531 or: Sture Hansson Asko Laboratory University of Stockholm Box 6801 S-113 68 Stockholm Sweden Tel. 08/340860 0156/22260 (Asko field laboratory) or, concerning sections 4-11, to individual authors (addresses in section 1.3) v:.n ABSTRACT On October 26, 1977 , the Soviet tanker Tsesis struck a rock in the fairway while inbound through the archipelago off Sodertalje, Sweden (northern Baltic proper) . During the next few days she released about 1100 tons of oil, mostly a No. 5 fuel oil, but also some bunker oil. Due to a quick and efficient response by the Swedish Coast Guard, and unusually favourable circumstances, the clean-up operations recovered most of the oil, leaving only about 400 tons in the enclosed archi- 2 pelago, visibly oiling an area of 34 km . Within a day after the grounding, a cooperative international scientific investigation was launched to cover important aspects of the spilled oil's ecological effects on plankton, benthos, fish, littoral and supralittoral communities, as well as the chemical and biochemical processes of weathering, bioaccumulation and depuration. The area was one which had been relatively well studied in the past as the site of several marine biological programs. The spill study was initiated rapidly enough to have some littoral sites surveyed immediately before impact by oil--an example of the "ideal" oil spill study. The direct effects on the plankton included decreased zooplankton biomasses in the immediate vicinity of the tanker. Within days, in- creased phytoplankton biomass and primary production, as well as bac- terial abundance, were noted in a larger area surrounding the tanker. Here zooplankton abundance and biomass were apparently not affected in spite of considerable contamination of zooplankton by oil droplets. No contamination or harmful effects on pelagic fish could be demonstrated. In a small bay near the spill site, oil concentrations of 60 |Jg/l were found in the water below a weathered oil slick. After one month, all parameters measured in the pelagic zone had returned to normal values. Damage to bird life and the littoral zone was alleviated by the season of the spill. The supralittoral zone showed little damage when surveyed the following summer. In the littoral zone, no effects on the algae were seen, but the fauna of the most heavily oiled coastline IX showed direct effects. Crustaceans were especially hard hit, but less than a year later, recovery was well underway. Oil analyses of mussels, Mytilus edulis , showed that oil had reached a larger area than was visibly impacted. Extremely high oil levels were found in Mytilus after the spill, and a year later Tsesis oil was still evidenced in the mussels. Sediment traps deployed in the area collected material strongly contaminated with oil for the first weeks after the spill, demonstrating that oil was rapidly transported to the benthos. A minimum sedimentation of about 20 tons of oil was estimated. About two weeks after the spill, when the first samples were taken on deeper (30m) soft bottoms, oil impact was already extensive. At the most heavily affected station, motile macrofauna were greatly reduced, possibly through emigration. The sedentary species remained, and no increase in their mortality was demonstrated. Macoma balthica showed high oil levels over a large area, even at stations where no clear impact was shown by macrofauna community composition. All meiofauna groups, except the nematodes, were also reduced at the most affected station, and there was evidence of high mortality of ostracods. A few months after the spill, the few remaining gravid amphipod ( Pontoporeia af finis ) females at the most affected station showed an increased frequency of abnormal eggs. The recovery of the deep soft bottoms after oil damage proved slower than for other systems studied. About ten months after the spill, neither macro- nor meiofauna at this station showed any recovery, and oil levels in Macoma balthica were higher than immediately after the spill. In June, seven months after the spill, the herring showed less spawning and lower hatching success in the oiled area than in a refer- ence area. This might be due to factors other than oil and needs con- firmation. The chemical analysis by Gas Chromatography (GC) and Mass Spec- trometry (MS) showed rapid weathering of the Tsesis oil, and all oil found either in biota or in sediment trap material was altered in com- position, but identifiable as Tsesis oil. Depuration in Mytilus and Macoma showed more rapid elimination of the aliphatic than the aromatic fractions. The trimethylbenzenes were identified as a fraction particu- larly resistant to degradation. The oil analyses proved invaluable for correct interpretation of the ecological data. x CHAPTER 1: EXECUTIVE SUMMARY The contents of this report result from a number of unusual cir- cumstances, which provided a truly unique opportunity for the scientific investigation of a medium-sized oil spill in a marine environment. Although the disaster of a tanker grounding can hardly be called fortunate, its location, circumstances, and timing were fortuitous. The sections that follow will describe the unusual opportunity provided by the unfor- tunate grounding of the Tsesis , and will discuss in general some impor- tant lessons for the management of oil spills. 1.1 Development, Early History and Cleanup of the Spill (Olle Linden and John Kineman) At 11:05 a.m. on October 26, 1977, the Tsesis ran aground while entering the Sodertalje ship channel in the archipelago about 50 km south of Stockholm, Sweden. The position of the grounding was approxi- mately latitude 58° 49.7'N and longitude 17° 43.8'E (Fig. 1.1). The location is shown on the general chart of the area (Fig. 1.2) and on the detailed chart of the ship channel (Figs. 1.3). The 19,334 dwt. , 165-m long tanker, Tsesis , was owned by a Soviet government shipping company in Ventspils, Latvia, and was enroute from Ventspils with 17,575 tons of a medium grade fuel oil. The ship had unloaded part of the cargo at Stockholm and was bound for the industrial city of Sodertalje with a pilot when she struck the Karingklabben shoal, immediately southeast of Fifong Island, in the southern part of the ship channel. The submerged rock which the Tsesis struck was uncharted at 6.5 m depth, whereas the draft of the ship before grounding was estimated to be about 8.5 m. On grounding, the ship was badly damaged and seven cargo tanks were ripped open (Figure 1.4). The punctured tanks contained a total of 6400 tons of a straight run #5 fuel oil. This oil had a density of 0.9022 at 20 C and was derived from Russian crude oil. It apparently contained enough of the original light crude fractions to allow it to emulsify. The Atlantic Fig. 1.1 The Nordic Countries •%. see detailed map, fig. 1.2 100 200 I I L_ 300 km J ^ V.,.,r>fk c fflalmp ■ MW %#►*"<*>*" ■ ;.,.*..,,.. , . . I Uniform* r. r i k * r • Itilhm - ^' T V ,_""** . i . . . : A . ■»* i ■t\ ISO. **T , /". . ■'■■ ttpir* ' Ji.-i-ff* * «$t*fc*« > !fe: J73S>" dftfrlfe t \" " '-a\,Aa - A \ 3 r ,,*.>■>«>■■ w »VH«s ^-sjsr/- s. «... *: ^ , ^.4 «*>*- ■ ■ -'■ ■ z «,«, / K ~'_1 8 . » " * vw",,,. 8 '' <)'«?<•' .■■„ s » I >. '*- T *'"' V J ■• '■■■ ^Ws ■ w-r T^\"i * ^' r - . * w £ ' " •/ " " '""" I &ja..,\ 1 '" «rik» » ' * r r «,i>-.„i„,. y — ■" , ! \ ', " u " - / ^ tm„UrJ< • ^-~. '* r««» 51 SA ^o..i« :*^,»».» l\*i "^"^ '"' ' s ' « " "t^ ' _r *-• f'. n ' , '*. j '« «> « Vy, %SHffi^» s , ^ s9 J /| ffl & ^5^ °H» " 'AJ . j j, ' » » » ■ a "«■ "i § f ,' ' ,.iv /-, -"\ « X a |fM»kflM>?\ V ■ • ^ '" - ' ■ V Fig. 1.2 General chart of the area of Tsesis grounding, Fig. 1.3 Detailed chart of ship channel. ^r~-i i i ii ' i --h j i 1, i i i . i ii 1 1 — I ' ____ r _ i i i ii i 1 i ■ •■ ~n — i — --t.-l- "^i^i i ! u i I i BOW BC- Bunker C S -- Starboard c = Center p = Port = Tanks damaged on grounding Fie. 1.4 TSESIS TANK LAYOUT This increased its stability and made recovery more difficult because of the greater volume. Furthermore, the viscosity of the emulsion remained very low in a sea temperature of 8 C, so that it behaved like an un- emulsified oil and remained difficult to contain. In addition to the seven cargo tanks, the port and starboard bunker tanks were also damaged, leaking an unknown quantity of Bunker C. Oil leakage began immediately after grounding, and the oil was blown by westerly winds of 2 to 4 m/s (5 to 8 mph) . About \\ hours after the accident, the Coast Guard and pilots arrived and deployed containment booms. Within another four hours, special units from the Coast Guard arrived and deployed high seas booms on the leeward side of the ship. Early the next morning, October 27, crews began recovery of the oil, using suction pumps, skimmers, and other clean-up equipment (see footnote) , but shortage of support tanks for the deposition of the re- Footnote: The Swedish Coast Guard reports that the following material was used during the Tsesis clean-up operations: 1. Containment booms: High seas booms, Sea-pact 3 (1500 m) Bravur (200 m) Coastal booms Expandi (2500 m) Addition booms were deployed by local agencies. 2. Coast Guard ships: 7 oil combatting vessels 7 Coast Guard cutters 2 motor boats 4 work boats 3. Other ships 4 tugboats 4 barges 5 tankers 4. Oil recovery equipment: 2 Vikoma skimmers 8 Komara skimmers (hydraulic) 3 Vacuum pumps 1 Lockhead skimmer 3 belt skimmers adjusted to combatting vessels 1 Marco oil recovery unit 1 Petrolina belt skimmer 1 Oil recovery unit (screw type) 1 Framo-unit for "lightering" on-top 5. Other equipment: 8 containers 6. Reserve supply (not used): Approx. 1700 m containment booms, 2 Vikoma skimmers, 1 Framo unit, dispersants. Detailed information on the organization, control and evaluation of the clean-up techniques employed can be obtained from Captain Sven Uhler, Swedish Coast Guard Service, Fack, S-103 10, Stockholm, Sweden. covered oil delayed the work. Shortly after noon, the Soviet tanker Talsy, a sister ship to Tsesis , arrived to assist in "lightering"; however, this ship was refused permission to off-load Tsesis oil because it had not been cleaned properly, and the danger of explosion was high. Swedish authorities ordered barges, which arrived later that day, allow- ing the recovery of oil inside the high seas boom to continue. Throughout the period, wherever floating oil was being transported, trajectories were mainly dependent on wind and topography. This greatly simplified tracking the surface oil, especially since the winds remained steadily from the west until October 28. The part of the oil that had already escaped before any booms arrived hit the shores of Toro, east of Svardsf jarden, during the after- noon of October 27th (see map, Fig. 4.9). Periodically the wind became strong enough to cause complete failure of containments, even when the booms were in place. Additional booms were deployed along shorelines, in sounds and in other areas where recovery of oil might be successful. In some cases the oil was directed into bays where it could be collected from the shores or from the sea. In all, about 6,600 m of containment booms were used. In the early morning of October 28 a Swedish tanker arrived, pro- viding greater storage capacity than the barges and, therefore, allowing the recovery of oil to continue. The wind still complicated the work as it turned southwest and increased in velocity to 4.5 to 5.8 m/s (10 to 13 mph) . This was responsible for complete failure of even the near- shore booms, and the change in wind direction caused beach pollution farther north along the shores of Liso Island (see photos, Appendix 2). During the following days, the recovery of oil continued both in the proximity of the ship and in bays and sounds throughout the area. A few hours after noon on October 31st, Tsesis was towed off the rock where it had been firmly grounded for 5 days. She was then anchored in Svardsf jarden (see Fig. 4.9 for map) until November 3, when she was finally towed from the area for extensive repairs in Stockholm. The winds remained steady from the south through November 1st. About three weeks after the grounding, the Coast Guard reported no more floating oil in the area. By that date, a total amount of approx- 3 imately 1700 m of liquid material, including oil, water, debris, etc., had been recovered. Taking into account that samples of the spilled oil contained 18 to 76% water, the Coast Guard estimated that 600 to 700 tons of whole oil were actually recovered. It is difficult to determine the quantity of the remaining oil because accurate tank soundings were not available. However, a reasonable estimate could be made that the total volume of the spill was somewhat larger than 1000 tons (Table 1.1). This implies that about 400 tons were left in the environment after the initial clean-up operation. Some of this oil could be found adhering to rocks, stones, and mud beaches along 18km of shoreline. The area of impact is shown in Figure 4.10. 10 Table 1.1 The quantity of oil in the damaged tanks Tank 3 l Quantity of oil (m ) Bunker C tanks Starboard bow Port bow unknown unknown Cargo tanks Starboard 1 Starboard 2 Center 1 Center 2 Center 3 Port 1 Port 2 1162 1419 91 1087 58 1200 1400 m =6.29 barrel = 0.902 metric ton (20° C) 11 1.2 The Scientific Study and Participating Institutions The circumstances of the Tsesis incident contributed to the even- tual involvement of three institutions in the short and long term studies that followed. These were the Swedish Asko Laboratory, The Swedish Water and Air Pollution Research Institute (IVL) , and the U.S. National Oceanic and Atmospheric Administration (NOAA) ; the involvement of each is described below. The Tsesis oil spill was unusual in that the grounding was in the study area of an existing marine science program. This meant that scientists were already on-scene with extensive historical knowledge of the system, logistics and facilities were in place and ready, and a sampling program was in progress, requiring only modification to suit the circumstances and needs of the oil spill study. The Asko Laboratory, supported with personnel from the University of Stockholm, is situated on the Island of Askon only 4 km from the site of the grounding. The main project carried out by the laboratory since 1971, sponsored by the Swedish Natural Science Research Council, has been "Dynamics and Energy Flow in the Baltic Ecosystem." The project's objective has been to create a complete ecological model of the Baltic where the flows of energy and matter (representing basic processes such as inflows of light energy and nutrients, feeding, sedimentation and mineral cycling) tie together the various storages of plants, animals and chemical matter. Another research project, in progress by the Asko Laboratory, was con- cerned with the ecological effects of the waste water from a sewage plant. This plant is located about 25 km north of Asko in the fjord "Himmerf jarden" . It was found that the fjord and the archipelago out- side it, with its algal fringe, can act as a filter or "sink", retaining some of the nutrients and releasing a "cleaner" water into the open Baltic. This project originated in 1972, sponsored in part by the Swedish Environment Protection Board. The oil spill from the tanker Tsesis occurred on the boundary where these two well-documented investi- gation areas overlap. 12 It was also natural that a cooperation between the Asko Laboratory and the Swedish Water and Air Pollution Research Institute (IVL) would follow the Tsesis grounding—the Asko Laboratory had the necessary intimate ecological background knowledge of the area and IVL had the needed experience in oil spill research. The physical proximity and close personal connections between the two institutions helped even more. IVL is an environmental research institute, financed jointly by the Swedish government and industry. Its oil pollution unit is located at Studsvik on the mainland, not far from Asko. Researchers at IVL have conducted oil pollution research since 1971 and have investigated spills worldwide as part of a special U.N.F.A.O. response team. On the morning after the spill, when the Asko team arrived on scene to start the investi- gation, Mats Notini from IVL was included and played an important role in the planning and conduct of the research. Later, Dr. Olle Linden, then on leave from IVL, functioned for two months as project leader for the Asko Laboratory investigation. He continued to play an active and helpful part in the investigation, even after his return to IVL, serving as liaison officer between the Swedish investigators and NOAA. The unlikely participant in the Tsesis investigation was the United States Department of Commerce, National Oceanic and Atmospheric Adminis- tration (NOAA) . Within this organization at the time of the Tsesis spill, the Outer Continental Shelf Environmental Assessment Program (OCSEAP) , funded by the Bureau of Land Management (BLM) , maintained an oil spill response project in Boulder, Colorado, under the management of David Kennedy. The mission of this Spilled Oil Research (SOR) Team was to take scientific advantage of oil spills having a significant research potential. The main objectives of the team at the time of the Tsesis incident were to study trajectories of oil spills (under the scientific leadership of Dr. Jerry Gait) and chemical fates in the water column (chief scientist, Dr. James Mattson) . In addition, Lt. John Kineman of the NOAA Corps was given the task of investigating the "addition of a biological component" to the response team, since the sponsoring organ- ization (OCSEAP) was interested in encouraging research on the bio- logical effects of acute oil spills. The ultimate purpose of the SOR project within OCSEAP was to relate, by analogy, the study results from 13 actual oil spills to theoretical problems in the environmental assess- ment of Alaska. These interests, and the unique situation created by the Tsesis incident, formed the basis for international cooperation between Sweden and the United States. As a matter of routine, the Tsesis incident was evaluated by the SOR Team for research potential. It was concluded that enough oil would remain in the enclosed area for chemical studies of the water column, and that the proximity and par- ticipation of the Swedish institutions mentioned would afford an excel- lent opportunity to investigate the feasibility of various biological studies as part of a coordinated program. Discussions, via conference phone and in meetings, ensued to refine the experiment design. These discussions included Dr. Wilmot Hess (Director of the Environmental Research Laboratories of NOAA) , Dr. Rudy Engelmann (Director of OCSEAP) , Dr. Douglas Wolfe (Deputy Director of OCSEAP), David Kennedy, Elaine Chan, Dr. Jerry Gait, Dr. James Mattson, Dr. Peter Grose, and Lt. John Kineman from the U.S. and Dr. Bengt-Owe Jansson in Sweden. On the day following the spill, the Asko Laboratory (Dr. Jansson) received a request from the OCSEA Program, through the U.S. State Department, for permission to send a team of oil spill scientists to study the spill. The Asko Laboratory welcomed this offer, and the response coordinator, Lt. John Kineman, with two chemists, Dr. James Mattson and Robert C. Clark Jr. (MS), participated in the short-term phase of the spill study. David Kennedy coordinated the direction, communication and logistics for the NOAA team from the project office in Boulder, Colorado. Lt. Kineman, as biological coordinator for the OCSEAP SOR program and later as acting project manager, coordinated the U.S. part of the Tsesis long-term study and participated in evaluation meetings in Sweden and the U.S. The major element of this coordination was the funding of hydrocarbon analyses performed under contract to OCSEAP by Dr. Paul Boehm (principal inves- tigator) of Energy Resources Company Inc. (ERCO). Because of the par- ticipation of NOAA, the Tsesis study included advanced hydrocarbon analyses of biota and sediments. Without these analyses the ecological investigations made by the Swedes would have lost much of their value and in some cases would even have been impossible to interpret correctly. 14 1.3 List of scientific personnel with major participation in the Tsesis oil spill study Asko Laboratory Dr. Lars Westin 1) Dr. Ragnar Elmgren 1) 2) Dr Mia Foberg B.S. Gunnar Aneer 2) Dr. Bjorn Guterstam 1) 1) Sture Hansson B.S. Sif Johansson B.S. 1) Hans Kautsky B.S. 1) Ulf Larsson B.S 1) Sture Nellbring B.S Brita Sundelin B.S 1) 1) NOAA Lt. John Kineman 5) 6) Dr. James Mattson Robert C. Clark Jr. M.S. 7) Sweden: USA: 5) 6) 1) Asko Laboratory University of Stockholm Box 6801 S-113 86 STOCKHOLM Sweden 2) Department of Zoology Universty of Stockholm Box 6801 S-113 86 STOCKHOLM Sweden 3) Swedish Water and Air Pollution Research Institute Studsvik S-611 01 NYKOPING Sweden 4) Botanical Institute University of Stockholm S-106 91 STOCKHOLM 7) IVL Dr. Olle Linden 3) Mats Notini B.S. 3) Botanical Institute Anders Lindhe B.S 4) ERCO Dr. Paul Boehm Judith Barak B.S David Fiest M.S. Adria Elskus B.S ) Spilled Oil Research Team Enivronmental Research Laboratories National Oceanic and Atmospheric Admin U.S. Dept. of Commerce BOULDER, Colorado 80303 USA Marine Assessment Office National Oceanic and Atmospheric Administration 3300 Whitehaven Street, NW WASHINGTON, DC 20008 USA Environmental Conservation Div. Northwest & Alaska Fisheries Cent. National Marine Fisheries Service National Oceanic and Atmospheric Administration 2725 Montlake Boulevard East SEATTLE, Washington 98112 USA 8) Energy Resources Company Inc. 185 Alewife Brook Parkway CAMBRIDGE, Massachusetts 02138 USA 15 1.4 Acknowledgements The support of all funding authorities (See Section 1.5) is grate- fully acknowledged. During the acute phase of the spill, the Research Laboratory (U-lab) of the Swedish Environment Protection Board also supported the investigation through the offices of Gunnar Guzikovski and Paul Anderson, and the Swedish Coast Guard provided valuable assistance throughout the investigation. The hard work of the technical and administrative staff of the Asko Laboratory was a prime requisite for a successful study. The first draft was typed by Kerstin Ernquist, Maureen Moir, Jane Sjolander and Almaz Terrefe. Bibi Mayerhofer and Lis Klove-Bjorklund helped with many of the illustrations. Leif Lundgren and Ulf Aneer helped in the field with collection of samples. Carl-Henrik Bagger and Bernt Abrahamsson commanded R/V Aurelia. Inger Hafdell and Ann-Britt Holm sorted macro- fauna and Kerstin Rigneus phytoplankton samples. The team of Asko scientists, led by Dr. Bengt-Owe Jansson (Director of Asko Laboratory), who participated in the initial investigation and study plan design included, apart from the authors of this report (listed in section 1.3), the following persons: Dr. Sven Ankar, Hans Cederwall, $ke Hagstrom, Ruth Hobro, Nils Kautsky, Inger Wallentinus and Fredrik Wulff. During this initial phase the U.S. NOAA/OCSEAP SOR team provided invaluable guidance concerning both oil spills in general and sampling for hydro- carbons in particular. The administrative staff -of OCSEAP also assisted with administrative help and typing interim reports, contracts for the chemical analysis, communications, and drafts of the report. These people include Wanda Power, Rosalie Redmond, Kay Jentsch, Norene Easton, Rosa Echard, and Mary Venis. Susan Rothschild typed the corrections and final copy. Final editing and coordinated proofing was done by Marian Cord, who supervised the report through the stages of publication. At several points in the process of applying for funds to continue the chemical backup of the study, and in the process of producing this document, program descriptions were written and circulated to reviewers for comment. These reviews and comments were very helpful in evaluating 16 the studies and priorities for funding the chemical backup. Subsequent reviews contributed significantly to the current draft of this report. The help of the following reviewers is gratefully acknowledged (reviewers of this report are marked with an asterisk) : Dr. Jack Anderson, Rattelle N.W. Dr. Donald Malins, Environmental Conservation Division, Northwest Marine Fisheries Service, NOAA Dr. Charles T. Krebs , St. Mary's College Dr. George J. Mueller, Institute for Marine Science, University of Alaska Dr. Steve LeGore, NALCO Environmental Sciences Dr. Dennis Stainken, U.S. Environmental Protection Agency Dr. John Teal, Woods Hole Oceanographic Institute Dr. Kiell I. Johannessen, The Norwegian Marine Pollution and Monitoring Program Dr. Don Westlake, University of Alberta »Dr. Douglas Wolfe, OCSEAP, NOAA -Dr. John Calder, OCSEAP, NOAA -Dr. David Nyquist, OCSEAP, NOAA "Dr. Howard Feder, University of Alaska "Dr. John Farrington, Woods Hole Oceanographic Institution -Dr. Harold Hodgins , NMFS , NOAA -MS. Robert C. Clark, Jr., NMFS, NOAA -Dr. John Augenfeld, Battelle N.W. In addition to other acknowledgements mentioned above, the U.S. NOAA/OCSEAP SOR Team wishes to express appreciation to the following offices and personnel for crucial assistance during the activation phase of the team's response: The Associated Press wire desk in Denver, Colorado, cooperated in providing initial and subsequent news releases; Mr. Tuner, working with the On-Scene Coordinator in Sweden, provided information and facilitated contact with Asko Laboratory; Dr. ftke Hagstrom, of Asko Laboratory, provided the first thorough description of scientific activities on-scene and facilitated further coordination; Dr. Bengt-Owe Jansson, gave continuous information on spill conditions, initial 17 approval to work with the Asko Scientists, enthusiastic support for U.S. participation, and a warm reception to the U.S. Team upon arrival in Sweden; Mr. Kinter of the Environmental Office of the U.S. State depart- ment provided needed assistance in obtaining official clearances for U.S. participation; Mr. L.-E. Ortegren, Science Advisor at the Swedish Embassy in Washington, assisted with official clearances; CDR Roland Engdahl , Swedish Coast Guard, On-scene Coordinator at the Tsesis incident, gave willing approval for the presence of U.S. scientists; Dr. Wilmot Hess, Director of NOAA's Environmental Research Laboratories, approved OCSEAP participation in the Tsesis study; CDR Richard Alderman, of the NOAA office of International Affairs, assisted in obtaining clearance in spite of an oversight in initiating contact with his office; Dr. D. Lee Alverson, Director of the Northwest and Alaska Fisheries Center, and Dr. Donald Malins agreed to provide a chemist to join the OCSEAP team; Dr. Bill MacLeod of NWAFS Analytical Facility, made preparations to participate in the SOR Team response and later assisted in oil chemistry analysis; Dr. Thomas Austin, Director of the Center for Experimental and Data Analysis, approved the participation of Dr. James Mattson; Mr. James Murphy of the State Department Passport Office in Washington, D.C. expedited a passport for one participant; and finally, all personnel at the Asko Laboratory provided a pleasant atmosphere to work and live in. Special thanks go to Kerstin Ernquist, who helped with supplies and living arrangements and to Carl-Herman Bagger, captain of the research vessel Aurelia . Finally, thanks must go to Dr. Rudolf Engelmann (Director, OCSEAP) and Dr. Douglas Wolfe (Deputy Director, OCSEAP) who supported both the original concept of research response to oil spills and the coordinated effort reported herein. 18 1.5 Funding Initial financial support of the Asko Laboratory study was provided by the Swedish Coast Guard and later by the Swedish Environment Protection Board, which guaranteed resources for investigation of the acute phase for a duration of one month and then requested a program for further investigation of the long-term effects. A proposal based on existing knowledge of the ecological conditions in the area was submitted and tentatively accepted; however, no funds could be found. Consequently, the investigation ceased for almost six months except for an investi- gation of the Fucus fauna, carried out by IVL on a separate grant (see section 7.2) and some unfinanced sampling activities by interested and public-minded Asko scientists. Later, the Asko Laboratory was asked to submit a new proposal to the Swedish Environment Protection Board because new funds had become available. By then, the original proposal had become obsolete, and a pilot study had to be carried out to provide the necessary information for planning a continued study of those parts of the ecosystem that still showed evidence of persistent oil impact. The chemical back-up funded by OCSEAP was dependent on the biological program, and so also required stepwise approval. In all, the Asko Laboratory investigation has been financed through grants from the Swedish Environment Protection Board (Grants No. 7-548/77 and 7-548/78, a total of 260,000 Swedish Crowns), the Swedish Coast Guard (50,000 Swedish Crowns) and Sven and Dagmar Salens Stiftelse (110,000 Swedish Crowns). The IVL Fucus fauna study (Section 7.2) was supported through a separate Swedish Environment Protection Board grant to Mats Notini (75,000 Swedish Crowns) and an equivalent matching grant from IVL. IVL has also directly supported this study through Dr. Linden's continued participation after the original Asko grant was expended. Finally, the NOAA-OCSEA Program has supported the study through its own personnel and by funding the oil analyses. The total NOAA investment was approximately 70,000 U.S. Dollars, which were BLM funds administered by OCSEAP. The combined investments totaled approximately 900,000 Swedish Crowns or 200,000 U.S. Dollars. 19 1.6 Lessons learned for the management of oil spill investigations (Olle Linden, Ragnar Elmgren, Lars Westin, and John Kineman) Oil spills are no longer thought of as unlikely chance events, but rather as predictable results of the transport of oil by sea. New dialogue has been suggested to reflect this change in outlook, specifi- cally to begin discussing the "management" of oil spills. This empha- sizes the recognition that future spills will occur and implies the need for an organized approach to the conduct of events that follow. The Tsesis oil spill study was primarily an ad hoc planning effort, except for the fortunate existence of an ongoing marine science program. It was also a unique experience on the part of the local scientists involved--a typical situation for significant oil spills that have occurred to date. If the eventual transition is to be made from ad hoc planning to contingency management of events (including research) follow- ing the predictable and possibly larger future spills, a determination of important lessons to be learned from each incident is necessary. A number of these lessons, made apparent during the Tsesis study, are summarized for the purpose of improving future management of environ- mental investigations of oil spills. 1. Funding sources need to be identified in advance . In retrospect, it is obvious that the uncertainty of funding for this investiga- tion seriously hampered its successful realization. For a consid- erable and probably very important period of time, the major part of the studies was halted due to lack of financial support. Funding sources should be established immediately. This can be done, for example, by setting up a fund financed by a tax on im- ported oil (cp. the U.S. "super fund" legislation) from which institutions and individuals could seek restitution for oil spill damages to natural resources and ecosystems. This fund would also be used for the support of research following oil spills to deter- mine the extent of the damage. Such a plan has received much discussion in the United States and a similar method of financing has also been suggested in Sweden (MIST, 1979). 20 Furthermore, the level of funds allocated for the scientific study of an oil spill must not be based solely on gross factors such as the size of the spill or the degree of publicity, as has often been the case. The Tsesis study has shown that many factors are important in determining the seriousness of even small to medium pollution incidents. These include topography, water depth, oceanographic conditions, season, life stages, type of oil, degree of oil trapping, effectiveness of mitigative measures, species and communities insulted, and so on, forming a large matrix. The establishment of a reliable general fund will allow the initial scientific investigation to proceed quickly to evaluate elements of this matrix and to focus on the construction of a follow-up program that is the most scientifically competent. 2. Contingency research planning is needed . The Tsesis study was launched in a few days. Few of the local scientists involved had any experience with oil pollution studies. During the first days and weeks after the spill, a number of important processes took place rapidly in the affected area and had to be studied while in progress. Under such circumstances it is not surprising that important sampling at one site was neglected while scientists were occupied at other sites. In specific regions where future spills are likely, as with the Sodertalje strait and archipelago, local scientists must perform careful advance planning. The plan should list the different investigations to be carried out; materials and sampling equipment needed for the various investigations should be listed and procured. The plan should designate experienced person- nel and institutions capable of performing the studies; a project leader should be appointed and a center designated for the coordi- nation of the various investigations (see Pollack and Stolzenback, 1978). It is not reasonable, however, to expect spills to occur close to existing facilities, as with Tsesis ; this fact is the basis for the scientific support capability which was established in NOAA. A 21 properly coordinated physically responsive group can perform the initial scientific evaluation and construct the follow-up program. In the case of local contingency research plans in high risk areas (such as the Sodertalje shipping channel), details of the study can be laid out in advance and the study narrowed according to the events of the spill. For less pre-studied areas or for readiness in general, it is necessary to lay out a flexible scientific capa- bility. The initial investigation (quick response) will have to include some method for rapidly learning the important aspects of the environment unluckily chosen by the spill, and then determining the parameters that are most important to measure in that case. A plan that is flexible enough to fit a variety of situations, but rigid enough to comply with practised and proven methods, is the ultimate goal. The follow-up program should incorporate early, and to a maximum degree, the capabilities of local or available institu- tions . Furthermore, the approach described recognizes the importance of the short-term phase of the study. Results of seemingly similar incidents can be diverse, and the ability to predict or understand the extent of effects from a given incident will depend heavily on detailed knowledge of the initial conditions of the spill and the local environment. Also, it is always desirable to obtain an indication of pre-spill conditions; but the rapidity with which processes occur during the first few days (for example, the trans- port of Tsesis oil to the benthos) means that immediate sampling is often the only way. In any but the most ideal conditions (e.g., Tsesis ) a pretrained, quickly responsive scientific investigating team is needed. An additional benefit of such a response group is the established capability to evaluate a variety of acute pollution incidents . In the Tsesis incident, the need for a special response group was compensated by the fortuitous proximity of the Asko laboratory 22 and the existing marine science program. Still, the U.S. scien- tists, who responded for limited objectives only, re-learned the importance of precise and comprehensive pre-planning of experiments. Procedures that are developed enroute or on-scene are likely to be incompletely thought out. Yet, it has been the tendency of re- sponse programs to fail to adequately emphasize the necessary research and development of experiment designs and field procedures, For example, the possibility of problems in using plastic bag samplers (chapter 5) has been hypothesized for some time; yet, the procedural limitations of this sampling technique were not defined until after the loss of important data and considerable expense. Quick scientific response is a valid strategy, but only proven techniques and procedures should be used. Their development and documentation must be adequately supported in advance and strictly adhered to in the field. Further development is needed for sampling techniques . The largest gap in the field capability at the Tsesis site (and at other recent oil spills) was the lack of an adequate sea water sampler for low level hydrocarbon analysis. The best results were obtained by Scuba, using glass containers (protected during surface entry); but this procedure is not usually practicable and is often dangerous during the early stages of the spill when sampling is most impor- tant. Furthermore, the Tsesis investigation revealed that large quantities of oil can be sedimented through the water column. The use of sediment traps are therefore recommended in future studies (see Section 3.3.1). However, before this can be employed on a matter of course basis at oil spill studies, some further careful study should go into trap design and a determination made of trap- ping eff iciency--especially the possibility of overtrapping, which could lead to erroneous balance calculations. 23 Thirdly, because of the importance of the observed transport of oil to the benthos, it is apparent from the Tsesis study that a strong need exists for a specialized sediment (benthic) sampler which will properly represent the uppermost, thin, and often floccu- lent layer for hydrocarbon analysis. This was the final piece of the puzzle that could not be added in the case of Tsesis, because of the inadequacy of existing samplers for this purpose. It has been suggested and demonstrated repeatedly at oil spills, that the proper expenditure of funds on necessary research and development before the study can result in a great financial saving as well as make the difference between a successful experi- ment and a failure. These three high priority research and development projects should receive funding as soon as possible. The source of such funding is most logically a government program with legislated responsibility for environmental protection or research. In the United States possible candidates are the U.S. Coast Guard, U.S. Environmental Protection Agency, or NOAA. The scope of the scientific investigation must not be limited administratively . The effects of oil spills in various environ- ments will probably never be fully classifiable. Controlling factors are too numerous. Therefore, only a competent response team can properly evaluate the environmental significance of a particular spill and determine what kinds of studies are needed to detect and understand what may be happening to the system. The contingency plan must cover a broad spectrum of possibilities from which some will be selected at the time of the spill. Each spill study reveals new phenomena. A very significant finding of the Tsesis study was that a relatively small spill of a medium weight fuel oil can be acted upon to produce rapid transport of large quantities of oil to the benthos, where, out of sight and beyond mitigative measures, it can have significant effects for a 24 long time. The mechanism behind this transport is also important, as is the possibility that effects observed in the benthos (and elsewhere) can propagate through the system. It is clear that all aspects of the environment must be considered, with adequate concern for the less visible aspects of the spill. Attention cannot arbitrarily be directed toward commercial species, endangered species, or other aspects of the environment that may reflect changeable social values at the time. Social values determine what aspects are of greatest interest and are perhaps a significant focus of the study, but only the investigative team can determine which, sometimes subtle, parameters should be measured. Some examples illustrate this: o Often the effects on highly mobile species (however important commercially or aesthetically) are hard to demonstrate because of the difficulty of resampling the same individuals or knowing well enough what they have been exposed to. However, effects may be similar to those observable in sessile or less mobile species. This may have been the case with bottom fish as opposed to herring, o The characteristics of various organisms can make them good or poor indicators of various factors. The benthic macrofauna species, Macoma balthica, are the easiest to sample and analyze, and they are characteristically resistant to damage; this makes them excellent biological indicators of the level of "insult" to the benthic community because they stay alive long enough to repre- sent possibly high concentrations or repeated exposures (Shaw et al., 1976). Their use alone, however, would lead to the erroneous conclusion that effects were slight although concentrations were high. By looking also at the more sensitive meiofauna and the crustacean macrofauna, as was done in the Tsesis study, significant effects 25 could be demonstrated. Conversely, the level of insult would be hard to demonstrate on such sensitive organisms which tend to be killed or to emigrate. For the purpose of this report, the word "insult" is being used to indicate something different from "exposure", "effect", or "impact". Although the ambiguity of these terms will probably not be totally resolved by this definition, "insult" here should be thought of as the amount of harmful substance that penetrated an organism's natural, first line defenses (e.g. skin, shells, exoskeletons , etc.). This would include "body burden" and coating of sensitive areas, but not, for example, coating of shells. As a further example, water column concentrations may be said to "insult" filter feeders. "Effects" can be migration, death, metabolism changes, etc. "Impact" will be used to refer to the entire picture: exposure, insult, and effect. 1.7 References MIST, 1979. Ren tur - program for miljosakra sjotransporter. Betankande av kommitten for miljorisker vid sjotransporter. SOU 1979:43. Liber Forlag, Stockholm, 1-117 (In Swedish). Pollack, A.M. and K.D. Stolzenbach. 1978. Crisis Science: investigations in response to the Argo Merchant oil spill. MIT Sea Grant Program, Report No. MITSG 78-8, June 1978. Shaw, D.G., A.J. Paul, L.M. Cheek and H.M. Feder. 1976. Macoma balthica : An Indicator of Oil Pollution. Mar. Pollut. Bull. 7:29-31. 26 2. Perspective CHAPTER 2: PERSPECTIVE It is important that each pollution incident be viewed in the context of the environment in which it occurred. Only then can the significance of pollution incidents and the importance of studies be recognized. This chapter provides some environmental considerations which attempt to place the Tsesis incident into perspective. The Tsesis spill obviously has implications for the Baltic, where the spill took place; in addition, a discussion is provided which com- pares the situation with environmental conditions in Alaska, the focus of which was the underlying justification for U.S. support. 2.1 Baltic Perspective (Ragnar Elmgren, Lars Westin, Olle Linden) During the last decade a number of valid studies on various effects of oil spills in marine environments have been carried out. These studies have dealt with extremely large and spectacular spills as well as spills of a more normal size. In few instances, however, have more than a few aspects of a spill been studied, and often the concern has been with littoral systems or bird populations. Few studies have at- tempted to illustrate the total effect of an oil spill on all the var- ious parts of the ecosystem (COMS, 1978; Kiihnhold, 1978). The Tsesis study was an attempt to do so and also represents one of the few studies carried out in the Baltic or in brackish waters in general. The Baltic Sea is one of the largest areas of brackish water in the 2 world, with a surface area of 366,000 km . It is the largest extensive area of low but stable salinity, mostly within the B-mesohaline range (5-10 /oo S = 0.5 - 1% salinity). The fluctuations of the surface salinity are small in spite of occasional powerful injections of North Sea water, which greatly affect the salinity of the deeper Baltic waters A salinity gradient exists from north to south, from 2-3 /oo S, in the innermost Bothnian Bay, to about 15-20 °/oo S in Kiel Bay. Tides are almost entirely absent. 29 There is also a gradient in climate, from subarctic conditions and more than six months of ice cover in the coastal zone of the extreme north, to a more maritime climate with an average of only a month of coastal ice in the south. Salinity is the most important controlling factor for the ecology of the Baltic Sea. The organisms are present in an osmotic environment, which allows only a limited number of euryhaline marine and fresh-water organisms and a few brackish-water specialists to establish themselves in the inner Baltic. The high osmotic stress, therefore, results in a simple ecosystem with just a few dominant species. The classic generalized "Remane's curve" (Remane, 1934) shows a marked species minimum at 6-7 /oo S. For further information, the reader is referred to the following general reviews of the ecology of the Baltic Sea: Remane, 1934, 1940, 1958; SergerstrSle , 1957; Zenkevitch, 1963; Jansson, 1978. There has been an increase in tanker traffic in the Baltic Sea, both with respect to the frequency of ships in the area and to tonnage. From April 1976 the new buoyed-off fairway at Darss Sound at the en- trance of the Baltic Sea permits tankers of up to 160,000 tons (dw) to enter. The extreme risk of huge spillages when ships of this size enter the Baltic is obvious. The shallow and imperfectly sounded waters increase the risk of grounding. Also, the poor visibility due to prevail- ing weather conditions during most of the year, as well as the high fre- quency of ships sailing in the area, increases the risk of grounding and collision. Totally, at least 1,000 spills with a calculated volume of about 100,000 tons are reported each year (Engdahl , 1976). Accidents involving spillage of oil are permanent phenomena in the Baltic. In heavily congested areas, such as the entrance of the Sodertalje ship channel, spills are also very frequent. Less than 8 months before the grounding of Tsesis , another accident resulting in a spill of 100-200 tons of medium grade (No. 5) fuel oil occurred about 5 km north of the Tsesis rock, but most of the oil drifted north into Himmerf jarden, and the area most heavily oiled by Tsesis was only lightly touched (as shown also by the relatively low pre-Tsesis-spill hydrocarbon levels in Mytilus in this area, see section 11). About four months after the grounding of 30 the Tsesis, another tanker carrying 35,000 tons of oil grounded some 100 km to the northeast, also releasing substantial quantities of oil. This oil did not reach the area investigated in this report. The spill site was located on the Swedish coast of the northern Baltic, roughly 65 km south of Stockholm. The archipelago here is relatively narrow and open to the Baltic, with a number of small barren skerries. Depths in the area are variable with a mean of 25-30 meters. The surface salinity is quite stable at 6-6.5 /oo S, and increases by 1 /oo at 40 m depth. Surface temperatures vary from below C in winter to a maximum of about 21-22 C in hot summers. This area is therefore representative of mean Baltic conditions and fairly typical of coastal areas with respect to hydrography, hydrology, and topography. It should be recognized, therefore, that not only are the Tsesis results significant for understanding or predicting the effects of spills in the Baltic, but also the methods and capabilities refined during the study may be directly applicable to future Baltic spill studies. The need for performing holistic ecological studies on the effects of oil spills in the Baltic has been explained in section 1.6. The Baltic environment is unique, so that the environmental risks must be studied before significant widespread damage becomes a fact. If this is not accomplished before a "catastrophic" incident, or its equivalent in accumulated smaller ones, it will then be too late to benefit from many of the lessons. The Tsesis incident was the second largest spill in the Baltic Sea to date. Since the spill occurred in an archipelago characterized by thousands of islands, inlets, rocks, and shoals, a significant length of shoreline was impacted, while topography and hydrography contained the spill in a relatively small, shallow area. In areas such as this (typi- cal of the Baltic) low exchange rates, low energy wave conditions, and near zero tidal energy mean that the removal and degradation of oil is very slow. Low temperatures especially in the bottom layers, makes evaporation of the acutely toxic, volatile aromatic hydrocarbons very slow. Furthermore, there are indications that the low salinity may play a part in prolonging the residence time of the lighter aromatic compounds 31 (Linden, unpubl.). There is little doubt that the cold, stagnant bottom waters of the Baltic, with low oxygen levels due to partial eutrophication and a pronounced halocline, will slow the degradation of oil that reaches the bottom below the halocline and cause it to be retained longer in the sediments, In general, conditions could hardly be worse for repeated oil spills. The chemical analysis of Macoma balthica has supported the possibility that Tsesis oil has been deposited in reservoirs where it degrades slowly and is, at times, resuspended. There are indications that a slow accumulation of petroleum hydrocarbons in the sediments is already a wide-spread phenomenon in the Baltic (Rudling, 1976). It may be important to note, especially when making comparisons between the Tsesis experience and other oil spills, that the effect of the Tsesis spill on the marine environment was seasonally dependent. The spill occurred during what was probably the least critical of all time periods both from the point of view of the aquatic biota and from the point of view of human habitants (consequently there was less pres- sure for short term remedial and cosmetic activity) . 32 2.2 Alaskan perspective (Ragnar Elmgren and John Kineman) "The prediction or assessment of pollution effects on the basis of observations extrapolated from one environment to another is seldom supported by adequate data. Unfortunately, however, few data on pollution effects exist for most areas and species, which has led to the use of information from areas that may be dissimilar in critical respects." (Evans and Rice, 1974) The above quotation defines the basic fallibility of intersite comparisons, but serves as a reminder that, nevertheless, the pressing requirements for oil and gas development have made such comparisons necessary. It is therefore important that 1) study areas are selected which are most comparable to those being assessed for development; and 2) the comparisons are thoroughly analyzed. The Baltic shares some important biotic and abiotic oceanographic features with Alaskan coastal waters. The geographical extent of the Baltic, from about 54 N to about 66 N, coincides largely with that of Alaskan waters, resulting in similar annual pulses of incident solar radiation as a forcing function for pelagic primary production at com- parable latitudes. As in Alaskan waters, there are large differences in climate between the southern and northern Baltic. The northernmost Bothnian Bay features near-arctic conditions and is ice-covered for more than half the year in the archipelagos, whereas in the much milder climate of the southern Baltic near-coastal ice is generally present for only one or two months. Water temperatures below about 20 m depth in the Baltic Sea are also comparable to Alaska. Below this depth, near- arctic conditions prevail throughout the year (annual mean temperature about 2-5 C, peak temperature around 10 C during the late autumn storm mixing) . Many Baltic macrobenthic species have taxonomically closely related counterparts in Alaska. The similarity between Baltic and Alaskan macrobenthos is especially striking when comparisons are made with inshore samples from Prudhoe Bay in the Beaufort Sea (Feder, 1976), but some dominant species are also shared with fauna of the Bering Sea (Alton, 1974) and the coastal areas of Prince William Sound, an 33 embayment of the Gulf of Alaska (Feder and Paul, 1977). In many cases where the species are not identical or their identity is questionable, the genera are the same (see Table 2.2.1). It is well known that the species occurring in the Baltic are tolerant of lowered salinities, this is presumably true of many of their counterparts in Alaska, even though the Alaskan species usually live in full salinity sea water. While salinity is known to be a major factor controlling distribution in the Baltic, it apparently is not a major factor in Alaska, except in isolated cases. However, there are also important differences between the two regions which must be kept in mind when comparisons are made. The Baltic is nearly a-tidal, whereas the tidal range in most Alaskan waters is generally large. This gives the shorelines entirely different physical and biological characteristics. Also, due to the absence of tides and a relatively short fetch across the Baltic, turbulent mixing is lower. This results in the rapid warming of Baltic surface waters in late spring to early summer, with the thermocline forming at about 15-20 m depth. The surface layer generally reaches temperatures of 15-18 C in summer, but inshore, and in exceptional summers even offshore, tempera- tures may temporarily exceed 20 C. Although there are locations in Alaska where some of the physical conditions of the surface water are similar to those in the Baltic (for example, the narrow tidal range in the Beaufort Sea), community level comparisons will probably be best in the deeper (below 15-20 m) zones (the soft bottoms). The salinity characteristics of the two regions clearly differ, with the only analogous areas in Alaska being at the head of estuaries and fjords adjacent to glaciers, where melt-water and runoff will at times create mesohaline conditions similar to the Baltic. However, the Baltic is unlike most estuarine areas of lowered salinity (including those in Alaska), because the salinities vary little during the year (generally less than ±1 o/oo S). This allows some species of marine origin to penetrate to surprisingly low salinities. A final differentiating feature of the Baltic is the salinity stratification of the Baltic proper (also observed in some bays of the 34 Table 2.2.1 Species and genera shared between the northern Baltic Sea and the coastal waters of Alaska. The list is far from complete. Group Plants (Flora) : Phaeophyta : Chlorophyta : Animals (Fauna) Priapulida : Polychaeta : Mollusca Crustacea : Ostracoda Mysidacea Isopoda Amphipoda Decapoda Bryozoa : Fish: Saltic Sea Fucus vesiculosus Cladophora glomerata Pygospio elegans Terebellides stroemi Nereis diversicolor Macoma balthica My a arenaria Mytilus edulis Paracyprideis fennica Heterocyprideis sorbyana Mysis relicta Saduria entomon Pontoporeia af finis Pontoporeia femorata Gammarus zaddachi Gammarus locusta Crangon crangon Electra crustulenta (syn. Membranipora crustulenta ) Coregonus nasus Coregonus lavaretus Alaska same same Halicryptus spinulosus same same same N. zonata same Macoma calcarea Macoma nasuta same same Most probably circum- polar species also present in Alaska Mysis spp . same same same same same C. spp. Membranipora spp . same Coregonus autumnalis Coregonus sardinella Myoxocephalus quadricornis same Gadus morrhua G. macrocephalus Platichtys flesus P. stellatus Clupea harengus membras Clupea harengus pallasi 35 Aleutian Islands), which leads to stagnation and oxygen deficiency at depths below 60-70 m and a consequent severe reduction of the bottom fauna below this zone. The area exposed to Tsesis oil did not include depths below 60 m, however. The ecological importance of the above differences cannot be com- pletely evaluated at present, due to lack of experimental data; however, as in the case of studies in controlled ecosystems, some differences may be desirable for the conduct of specific studies. A comparison of significant stresses (Table 2.2.2), for example, suggests that some stresses commonly imposed on marine communities and organisms in Alaska, are either negligible or held constant in the Baltic. This implies a more "controlled" situation in the Baltic and may help to isolate the effects of the spill from other perturbations. On the other hand, as shown in Table 2.2.3, the major differences combine to suggest a longer retention time and slower degradation of oil at the Tsesis site than at most places in Alaska. Although there are important differences in physical conditions between the Baltic and Alaskan coastal waters, the similarities mentioned initially still indicate that oil pollution studies in the two areas have important problems of common interest, such as: 1. Oil degradation in waters of very low temperature, especially studies of oil degradation in deeper (below 20 m) Baltic soft bottoms; these should be highly relevant for Alaskan environ- mental impact studies. 2. Community level effects in the deep soft bottoms. 3. Vulnerability of shared species (Table 2.2.1), some of which have been shown to be exceptionally susceptible to oil pollution, e.g., Pontoporeia femorata (Rohrbacher-Carls , 1978). 4. Problems connected with oil spills in ice-covered areas. 5. Problems connected with oil spills in osmotically stressed environments . 6. Problems connected with oil spills in environments with a high number of islands (fjords, bays, etc.) that create long coast- lines in a small area. 36 Table 2.2.2 Comparison of significant stress factors Stress Factor Alaska Baltic (spill area) 1. Brackish salinity Occurs in isolated areas Relatively uniform through- stressful for both such as bays within out local areas; constant fresh and saltwater Prince William Sound and stress for most organisms, organisms) varies greatly with time and depth (3-4m fresh water lens forms from glacial and snow melt and run-off) 2. Salinity fluctuations Significant in isolated Small: allows even areas: seasonal and de- stressed organisms to pendent on mixing exist beyond their normal range. 3. Low temperature Significant seasonal Relatively constant with of bottom water variation inshore, but time and location. not in deeper waters 4. Sediment load Seasonally high from Low glacial runoff and ad- jacent to major river systems, such as the Yukon and Copper Rivers 5. Other pollutants Low Significant Table 2.2.3 Factors affecting the elimination of oil from the environment Factor Alaska Baltic (spill area) Oxygen deficiency of Rare (only in isolated Significant below halocline bottom water bays in Aleutians) very thin oxidized layer, sometimes totally anaerobic conditions Tide and wave energy High along outer coastal Low (Combined) areas (but variable with area; low in the Arctic). Turnover of water mass Frequent, typically Less frequent: pronounced well mixed thermocline and halocline Other factors (bacteria, differences not well known photo-oxidation, penetrability of bottom sediments, etc.) 37 The two large regions compared here are geographically extensive and internally very heterogeneous. For these reasons detailed compari- sons with the Tsesis study should preferably be made for individual regions within Alaskan waters, with due regard to the specific features of the region. The above highly generalized comparisons were intended mainly to indicate that more detailed comparisons can be meaningful. 2.3 References Alton, M.S. 1974. Bering Sea benthos as a food resource for demersal fish populations. In: Hood, D.W. and Kelley, E.J. (eds.): Oceanography of the Bering Sea. Institute of Marine Science, University of Alaska, Fairbanks, 257-277. COMS 1978. In the Wake of the Argo Merchant. Proceedings of a symposium held January 11-13, 1978, at the Center for Ocean Managments Studies, University of Rhode Island, 1-181. Engdahl, R. 1976. Commodore R. Engdahl, Swedish Coast Guard Service. Private communication. Evans, D.R. and S. Rice. 1974. Effects of oil on marine ecosystems: A review for administrators and policy makers. Fish. Bull. 72:625-638 Feder, H.M. 1976. Benthic Biological Studies. In: Feder, H.M. , Shaw, D.G. and Naidu, A.S. (eds.): The Arctic Coastal Environment of Alaska. Vol. 1. The nearshore marine environment in Prudhoe Bay, Alaska. Institute of Marine Science, University of Alaska, Fairbanks, IMS Report No. 76-1 (March 1976), 53-161. Feder, H.M. and A.J. Paul. 1978. Biological cruises of the R/V Acona in Prince William Sound, Alaska from 1970-1973. Institute of Marine Science, Univ. of Alaska, Fairbanks, IMS Report R 77-14, Sea Grant Report 76-6, 1-76. Jansson, B.O. 1978. The Baltic - a systems analysis of a semi-enclosed sea. In: Advances in Oceanography, H. Charnock and G. Deacon (eds.), Plenum Publ. Corp., 131-183. Kiihnhold, W.W. 1978. Impact of the Argo Merchant oil spill on macro- benthic and pelagic organisms. AIBS conf. Assessment Ecological Impacts of oil spill. Keystone, CO, 152-179. Remane, A. 1934. Die Brackwasserfauna . Verh. dt. Zool. Ges., 34-74. Remane , A. 1940. Einfiihrung in die zoologische Okologie der Nord- und Ostsee. Tierw. Nord. u. Ostsee la, 1-238. 38 Remane, A. 1958. Okologie des Brackwassers . Die Binnengewasser 22: 1-216. Rohrbacher-Carls , M.R. 1978. Some sublethal effects of oiled sediments on a population of the marine amphipod Pontoporeia femorata (Kr^yer, 1842). Unpubl . M S thesis, Dalhousie University, April 1978, 1-92. Rudling, L. 1976. Oil pollution in the Baltic Sea. A chemical analytical search for monitoring methods. Statens NaturvSrdsverk . SNV PM 783, 1-80 SegerstrSle, S. 1957. Baltic Sea. Mem. Geol. Soc. Am. 67: 757-800. Zenkevitch, L.A. 1963. Biology of the seas of the U.S.S.R., (transl. from Russian by S. Botcharskaya) . London: Allen and Unwin, 1-955. 39 3. Scientific Summary and General Discussion CHAPTER 3: SCIENTIFIC SUMMARY AND GENERAL DISCUSSION (Olle Linden, Ragnar Elmgren, Lars Westin and John Kineman) 3.1 Background Few field studies of oil spills in the Baltic Sea are available. The Palva accident in 1969 was the object of a study at Abo University (Pelkonen and Tulkki , 1972). One year after the spill, no observable damage was found in the coastal zone, but it should be noted that the accident occurred at the very end of the archipelago of Finland's south- western coast, a highly exposed area. The only previous study of the long-term effects of an oil spill in the area exposed to Tsesis oil (Notini, 1977) has been a 5-year study of the recovery following a spill at Gastviken, Musko. The bay was severely contaminated with 400 tons of medium and heavy heating oil from the grounded Irini in the month of October 1970. A widespread recoloni- zation of the littoral fauna occurred quickly in the first summer after the oil spill. The less rapidly moving organisms, such as mussels and snails, reestablished themselves during the following years. Four years after the oil spill, no effects could be discerned in the ecological system of the Fucus belt zone in the bay, but the oil which was incor- porated into the sediments of the bottom remained and was not signifi- cantly degraded. The oil that remained on the beaches had been dras- tically altered. The rapid recuperation of the bay is explained in part by the large effort made to remove the oil by mechanical means, by the small size of the bay, by the good water exchange properties, and by the closeness of undisturbed areas which could serve as sources for recoloni- zation. The studies mentioned above appear to illustrate the importance of the relationship between energy in the system, oil degradation, and ecological recovery. This relationship has been described convincingly by several authors, including E.H. Owens (1978) and E. Gundlach et al. (1978). Quantitatively, the largest oil catastrophes have all occurred in areas of high energy, for instance, Metula 1974 (Straughan, 1977), Santa Barbara 1969 (Straughan, 1971), Torrey Canyon 1967 (Southward and 43 Southward, 1978), Monte Urquiola 1976 (Wennergren, personal communi- cation), and Amoco Cadiz 1978 (Hess, 1978). The recuperative ability of communities in the affected coastal areas is surprising. The smaller spills in protected environments, such as Florida 1969 in Buzzards Bay, Massachusetts (Sanders, 1978), often have the stronger locally acute as well as long-term effects. Well documented examples of this principle exist for isolated parts of a system, such as salt marshes (Krebs and Burns, 1977; Baker et al., 1977) and sediments in areas with low water- exchange rates (Mayo et al., 1978; Teal et al., 1978). In such areas, the aromatic hydrocarbon compounds have been shown to be most persistent (Vandermeulen and Gordon, 1976). The Irini study suggested, quite logically, that an important relationship exists between recovery rate and the proximity of source areas for re-colonization. This implies that the areal extent and patchiness of defaunation are also important factors in determining the rate of recovery. Spill studies to date have not shed much light on this aspect. 3.2 Discussion of results and major conclusions In the supra-littoral zone, faunal and floral surveys were carried out about 8 months after the accident. In general these studies showed no obvious remaining effects either on the invertebrate fauna of the beaches or on the various land plant species along the shores. These results, at least with respect to the flora, are in accordance with the findings in other studies (e.g. Baker, 1971). It is interesting to note that the only remaining damage in this area had obviously been caused during the clean-up operations (for example, deep wheel-tracks from dump trucks). One important factor in explaining the absence of effects in the supra-littoral zone might be that the oil spill occurred at a season characterized by extremely low standing stocks and production rates of animals and plants in the system. The studies of the impact of the oil on the littoral zone indicates substantial acute damage to all macrofauna species of the Fucus belt. 44 Based on field observations by divers, the disappearance of the Fucus belt macrofauna was due to death and/or narcosis. Along the shores of Toro and parts of Liso, mass mortality occurred among littoral crusta- ceans. After 12 months, considerable recovery had taken place, presum- ably through immigration of amphipods and isopods from refuge areas. Isopods of the genus Iaera were, however, an exception. In several places, no recovery at all of Iaera could be observed. This could probably be explained by the fact that these animals are extremely thigmotactic , and are never found swimming in the free water. The re- covery process of the littoral zone faunal composition was one of both migration and multiplication of surviving individuals. The gammarids were, for instance, totally lacking at a contaminated station in Novem- ber. In June, they showed high abundance in all belts, but only 8% of the samples contained adult individuals (corresponding frequency for reference stations was 75-80%) . The isopods Idothea remained in the area in low frequency in November, but reproduced in June and were subsequently found in higher frequency, especially in the Cladophora belt. The molluscs of the littoral zone do not appear to have suffered drastic mortality. Instead, bivalves and gastropods seem to have re- entered the algal zones after some time of immobilization on the bottom. Even if the recovery of the littoral fauna is incomplete, the results obtained so far would indicate that a complete recovery of all fauna representatives can be predicted within 2 to 3 years. In addition to results of the Tsesis investigations, this statement is supported by the findings of Notini (1978) when studying the recovery of an oil polluted bay in the Stockholm archipelago. Here, the recovery took 3 to 4 years after the spill, which occurred under much the same conditions as the Tsesis spill. The metabolism studies of the Fucus community did not reveal any definite effects on community metabolism in the affected area. This may have been due to the considerable variation of the background values. 45 Comprehensive studies on the long-term effects of oil in littoral communities indicate that the damage can be very great and can persist for prolonged periods of time. The thorough work of Southward and Southward (1978) on the effects of the Torrey Canyon oil spill of 1967 shows that serious effects, manifested as absence of several faunal species and abnormal and excessive growth of algae, were evident even after 8 years. After 10 years, the situation was still not normal with respect to the diversity of fauna. Evidence from Baja California during the wreck of the Tampico Maru indicates that species were still absent and the balance of the littoral community was not restored 16 years after the accident (North, 1973). The most important differences between these spills and the Tsesis spill, with respect to the effects in the littoral zone, are mainly results of the quantity and composition of the spilled oil and the technique used during the clean-up operations. The observations after the Torrey Canyon spill indicate that the longest lasting effects could be observed at the heavily oiled rocks, which received repeated appli- cations of toxic dispersants (Southward and Southward, 1978). The severe effects during the wreck of the Tampico Maru were caused by the large amounts of highly toxic diesel oil (North, 1973). Also, at the Tsesis site, the biological effects of the spill were minimized by low biological activity and low temperature and by the decision not to use oil dispersants. The oil spill happened at the start of a season with low activity in the littoral zone. The low temperature and ice cover in January and February minimized plant and animal activity, and many individuals died in accordance with the natural seasonal pattern. During these 3 to 4 months of low activity, before the spring growth started, some of the remaining oil in the littoral zone may have been washed out, and the toxic fractions may have been diluted. Thus, the effect was less than it might have been if the oil spill had occurred in the beginning of the growing season in April, May or June. The accumulation of petroleum hydrocarbons in littoral bivalves (Mytilus) is a very rapid phenomenon. During large spills in coastal areas, bivalves frequently demonstrate such accumulation (Blumer et al., 46 1970; Straughan, 1977; and others). The analysis of the tissues of blue mussels showed extremely high concentrations (up to 3% of fresh weight) of petroleum hydrocarbons to be present the first months after the spill. It is noteworthy that the Tsesis oil was found to be present in mussel tissues in the entire area of Svardsf jarden. Thus the contamina- ted area was much larger than initially assumed, based on visual observa- tions . The depuration of oil from contaminated bivalves is a process governed by several factors, such as size of the initial spill, com- position of the oil, the temperature, filtration rates and filtration behavior and physiological state of the animal (Fossato and Canzonier, 1976; DiSalvo et al., 1975; Stegeman, 1974; Stegeman and Teal, 1973). Studies have shown that acutely acquired petroleum is fairly rapidly released (Fossato and Canzonier, 1976; Anderson, 1975; Kanter, 1974; Lee et al., 1972; and others). In contrast, chronically accumulated hydrocarbons are retained for comparatively longer periods of time (Boehm and Quinn, 1978; DiSalvo and Guard, 1975). The Tsesis oil was released gradually, although perhaps not com- pletely, from the Mytilus tissues, during the year following the spill. This agrees with observations after the West Falmouth oil spill, that shellfish from the spill area continued to show a fuel oil hydrocarbon pattern for several years (Blumer et al., 1970; Blumer and Sass, 1972; Teal and Farrington, 1976). The pattern of depuration of the Tsesis oil from Mytilus indicated a somewhat more rapid release of the aliphatic fraction compared to the aromatic hydrocarbons. This is in general accordance with the findings of Blumer et al. (1970) and Stegeman and Teal (1973). In general, the chemical analysis of oil in blue mussels has proven to be a good measure of the extent of the littoral insult resulting from the Tsesis spill. Weathered Tsesis oil sedimented to the bottom in quantity within days after the spill, as shown by sediment trap data. These together with the oil analysis of Macoma balthica from a number of stations, show that the exposed area was much larger than originally suspected, includ- ing areas which were considered clean by visual inspection or biological sampling. The extent of the area where visible slicks were observed is 47 clearly not a reliable indicator of the total area affected. Within this affected area (shown by oil analysis), benthic macrofauna community responses varied from a slight and temporary decrease in abundance of motile species, especially Pontoporeia femorata , to a drastic and last- ing reduction in amphipods (Pontoporeia af finis and P. femorata ) . The disappearance was possibly caused by emigration from the contaminated area, since very few dead specimens were found. In laboratory experi- ments, Pontoporeia af finis has been shown to actively avoid oil con- taminated sediments. However, even at the most affected station, the sedentary macrofauna species remained with no observed increases in mortality. The few remaining gravid amphipods showed a statistically significant increase in the number of abnormal eggs at the most affected station. Of the meiofauna, all taxonomic groups except the nematodes showed abnormally low abundance at the most affected station and there was evidence of greatly increased mortality of ostracods following the spill. As late as August and September the following year, 10 months after the spill occurred, there was still no sign of recovery of the affected groups of macro- and meiofauna. The oil content of Ma coma balthica , which had initially shown some decrease, increased again in August to values even higher than the highest recorded earlier, one month after the spill. Whether this increase is a real increase due to a new influx of oil, presumably mobilized from shallower sediments, or just an expres- sion of considerable patchiness in the oil distribution, is uncertain. It supports the hypothesis that the breakdown of oil in such sediments is very slow, and that a prolonged period may be needed for full recovery. Because of the long life-cycle (two years) and non-migrating be- havior of ostracods, the effects of the oil on the soft bottoms will remain for at least two years. Most other meiofauna groups have a much shorter life-cycle and can recolonize the bottoms shortly after the oil is degraded. The amphipods, P. affinis and P. femorata have a more mobile lifestyle and are expected to invade the area very quickly after the oil is degraded. 48 The pelagic study covered only the acute phase of the spill, essen- tially the first month after the grounding. Valid oil concentration results were obtained from only two water column samples during this period. These two samples suggested concentrations of around 60 |Jg/l of micro-dispersed oil droplets under a weathered and fully emulsified part of the Tsesis oil slick. The phytoplankton biomass generally increased in the affected area, but it could not be shown whether this was due to increased growth rates or to decreased zooplankton grazing. Correspond- ingly, primary production was elevated. Phytoplankton species composi- tion showed little change. Zooplankton were found to be heavily contami- nated (internally and externally) by oil (50% of the specimens were visibly contaminated with oil droplets in the first week, 20% after three weeks); although no reduction was found in the standing stock, either for zooplankton or ciliates, except in the immediate vicinity of the ship. Bacterial abundance increased in the spill area, but here, too, it was inconclusive as to whether this was due to decreased grazing pressure or to increased growth rates. The most interesting result of the pelagic study was the high oil content (up to 0.7%) found in sediment trap material during the first two weeks after the spill. This was true also for sediment traps de- ployed in an area windward of the tanker, where no oil slicks were ever observed and which had been thought to be little affected by the spill. Within a week the oil in the sediment traps was significantly altered by weathering. The minimum extent of the impacted area can be estimated to 2 be 42 km from oil analysis of sedimented material and of Mytilus and Macoma . By extrapolating from the sediment trap data it is possible to give a minimum estimate of 19 tons of oil reaching the sediment. An amount of 40 tons, which is not unlikely, would be 10% of the estimated total amount remaining after oil recovery operations had ceased. One month after the spill all parameters measured in the pelagic system were essentially normal (phytoplankton, zooplankton, bacteria, oil content of sedimenting material). This indicates the relatively short duration of oil impact from a moderate spill in the pelagic system. Acute effects on the fish could not be demonstrated. 49 The fish species occurring in the area show both local and long distance migrations. Pelagic species (herring, sprat, cod, etc.) show the longest migrations. Littoral species and demersal fish from deep soft bottoms show local migrations (Westin, unpubl.). These are sea- sonal and largely a change of biotope. Thus the season governs the composition of the fish fauna in any given biotope. The largest fish concentrations during the season of the Tsesis spill occurred in the pelagic system, where large concentrations of herring are normally found in the general area of the spill (Aneer et al., 1978). However, echosounding surveys in the area failed to demon- strate avoidance behavior by the herring and analyses showed no oil contamination of locally caught herring. Possible effects of oil on the local fish fauna are thus likely to be mainly indirect, long-term, and difficult to detect. The impoverished littoral and benthic fauna in the impacted area obviously results in decreased food availability. The oil contamination of bivalves such as Mytilus edulis and Macoma balthica is likely to cause oil contamination of fish utilizing these food items (e.g. the flounder, Platichtys flesus, which preys on Macoma in this area). The most important effect of the spill on fish is likely to be found in species or life stages that are subject to long-term exposure. The herring in the Asko area spawn on exposed, fairly shallow bottoms, and an investigation of the main June spawning, half a year after the spill, showed both lower frequency of spawning and lower hatching rate in the spill area, but not a significant increase in the number of malformed larvae. The oil spill is not, however, the only possible explanation for these results. 3.3 Recommendations for spill research contingency plans The people involved in the Tsesis oil spill investigation have made a number of observations which may help in working out contingency research plans. These are: 50 1. The pelagic system : The most important finding in the pelagic study was the sedimentation of oil. Thus, sediment traps should be a standard instrument in the study of the environmental impact of oil spills, wherever their deployment and retrieval are physically possible. With deployment sooner (within 1-2 days after the spill) and a larger number of traps, it would have been possible to make a much better estimate of the amount of oil reaching the benthos--a measurement of crucial importance in estimating the extent and duration of the impact of oil on the ecosystem (also see section 1.6). Another crucial measurement in the pelagic zone is the composition and concentration of oil in the water column. Where such measurements can be made and tied to other observations, for example, the presence of oil droplets in association with zooplankton (Chapter 4) , important information on the mechanism of oil transport through the water column can be produced. Due to the problems of sampling, however, it is not yet practicable to survey large areas of the water for hydrocarbon analysis in a way that is statistically adequate. If a proper sampler can be devised (see section 1.6), its use will be of greatest value in conjunction with plankton, bacteria, littoral, and sedimentation studies to establish level of exposure as a basis for interpreting other phenomena. The study of plankton showed increases in bacterial and phytoplank- ton biomass, but it failed to elucidate the mechanism behind the increase. Higher sampling frequency, especially at the reference stations, might have helped, as might more "frequency of dividing cells" (FDC), measure- ments of bacterial growth rates. Clearly defined bacterial composition studies would certainly improve the assumption that bacterial increases are due to oleophilic species. Perhaps an equally important experiment would be to analyze the feeding rate of zooplankton from the area. This might demonstrate whether oil reaches the benthos mostly after ingestion by zooplankton and incorporation into fecal pellets or by direct adsorp- tion to detrital and mineral seston particles, which then sediment. The relation between dispersed particles, suspended sediments, and zooplankton feeding behavior should be investigated thoroughly. 51 The greatest single problem in the pelagic studies was the accurate prediction of the natural variability to be encountered and a subsequent statistical design that would adequately remove this mask. For example, the FDC measurements provide a good example of when significant differ- ences may indeed have existed between polluted and reference stations, but sample frequency was too low to detect it. It is true that very often field logistics may preclude a truely adequate statistical design; in such cases it would be advisable to evaluate whether or not a lesser effort will produce any usable results. Conducting fewer experiments with better statistical designs may be more advantageous. These com- ments also apply to areas other than the pelagic zone. Statistical design for field experiments is thoroughly discussed in a recent work by Moore and McLaughlin (1978). In most oil spills, direct impact on the pelagic system will be of short duration (less than a month). The most important measurements in the pelagic zone, therefore, will be those helping to define its impor- tance as a medium that transports oil to other less resilient systems, such as the benthos or the littoral zone. However, this is not meant to understate the importance of effects on pelagic larvae and other plankton, which might be of long term or trophic significance depending on the circumstances of the spill. 2. The littoral system : The net-bag method for sampling in the Fucus belt has proven to be an efficient method of surveying the effects on the littoral macrofauna. Studies of the effect of oil on littoral and supra-littoral communities, particularly on plants, are warranted as a means of evaluating the damage caused by the spill. These zones are especially sensitive to various methods of cleanup, and so studies should be included which assess the effectiveness or additional damage of cleanup methods. In situ (e.g., suspended) bio-assays were not attempted during the Tsesis study, although they have been recommended at times (MITR£, 1978)-. It is believed that in cases where direct studies of organisms in place are possible, such as Tsesis , fewer doubts are inherent than with bio-assays; thus studies of the existing system should be con- sidered first. Furthermore, when samples and transects can be done 52 post-spill but pre-exposure , as was done after the Tsesis grounding, the study is of course, most ideal. This requires rapid response and ac- curate trajectory predictions. 3. The deep soft botto ms: Since the Tsesis study showed that oil can reach the benthos within days, immediate sampling is necessary as an indication of pre-spill conditions (see section 1.6). The amphipods and the ostracods, both crustacean groups, were especially sensitive to oil. The amphipods probably emigrated from the affected area, while most of the ostracods probably died. Repeated sampling soon after the spill would be necessary to confirm such conclusions. The oil analysis proved invaluable in the interpretation of the biological responses, and for this the sediment trap data and the Ma coma balthica samples were particularly useful. Ma coma have been suggested as ideal indicators of oil pollution and the Tsesis study supports this idea. The most reliable indicator of the insult was the contamination of Macoma by oil. The second most reliable were changes in macro- and meiofauna community composition. Whereas Macoma are well suited for registering an oil insult to the area, other species are far more sen- sitive for registering ecological effects. Therefore, a well balanced (ideally holistic) study of the macro- and meiofauna is required to give a totally accurate picture. The oil distribution seems to have been quite patchy, and at least a rough mapping of the affected area would have been highly desirable. It may not be feasible, however, to perform such a mapping in a statis- tically reliable way. For this reason, the importance of patchiness and possible "refugia" is difficult to study, although it is likely that recovery rates depend strongly on the proximity of source areas for recolonization. Areal extent of the affected area can be established by sampling a grid of stations, taking one sample per station. Such a survey would require expensive analyses, but would, perhaps, permit greater economy in later analyses by identifying the most desirable sites for the more detailed studies. The location of areas of poor degradation should receive attention as possible reservoirs for pollutants . 53 The benthic zone in general, from the standpoint of damage and recovery, seems to merit much greater attention in future oil spill studies than it has characteristically received in the past. 4. The fish fauna: The seasonal influence on the migratory be- havior of the fish makes the effects of an oil spill dependent on the season of the year. Adult fish can possibly avoid high concentrations of oil; therefore, the greatest effects are to be expected largely in those earliest life stages that are subjected to prolonged oil exposure. Most fish species of economic importance along the Baltic coast spawn in the spring and early summer. An oil spill just before or during this period will have a maximal effect on reproduction. Investi- gations during this period should be concentrated on spring-spawning herring, where differences in hatching success are relatively easy to investigate. During other seasons, effects should be sought primarily in species with slow roe development and/or spawning on deeper soft bottoms, where oil effects probably persist for a longer period of time. Also, the accumulation of oil in food animals, especially on deep soft bottoms, may result in bioaccumulation in fish feeding on those food items. This should be studied in connection with future spills. No acute fish mortality was observed following the spill. A large mortality of small fish (probably gobies) in the littoral was, however, observed about a month after the spill by local residents. This implies that observations of fish should cover a longer period than the acute phase of the spill. 5. Integration of disciplines and studies : It is absolutely necessary to have high caliber analytical chemistry available on an interactive basis with the biological studies. Samples in the Tsesis investigation were collected and preserved from the beginning, with attention to minimizing extraneous contamination or introduction of other ambiguities, to preserve the integrity of the samples for later analysis. The storage of samples in anticipation of the funding for their analysis proved to be an excellent investment. The chemistry eventually showed the nature of the insult, which was necessary to properly interpret biological effects. 54 However, chemical analysis alone is not sufficient. Chemistry by itself can indicate levels of insult, but the high cost of analysis necessitates the establishment of analytical priorities. When biological studies are conducted, highest analytical priority must be given to those samples which can best elucidate observed biological phenomena. In the Tsesis investigation, having the biologists perform sampling for chemical analysis worked well. Instructing biologists on specific chemical sampling procedures to avoid contamination or other ambiquities was not a problem, and this seems to be the best way to ensure that the chemistry and biology are fully integrated in time, location, and purpose Although it is not a serious disadvantage to have a large set of samples to choose from, the analysis of samples carefully chosen to support a specific study appears to be more cost effective than a broad survey approach where time and space correlations with the biological sampling are hard to ensure. It seems that, even though chemical analysis is essential, the biology program should dictate the priorities. If the biology program cannot be complete (due to circumstances, logistics, or other reasons) it should at least be as cohesive as possi- ble. Many conclusions of the Tsesis investigation required supporting results from a number of studies. This requires careful planning. 6. Implementation : Spills do not usually occur in convenient locations. The problems of responding to an accidental spill in most cases, Tsesis excepted, necessitate a quick responding investigative team to assess the situation and select appropriate follow-up studies (see section 1.6). In the case of local contingency research plans in high risk areas (such as the Sodertalje shipping channel), details of the study can be laid out in advance and the study narrowed according to the events of the spill. For less pre-studied areas or for readiness in general, it is necessary to lay out a flexible scientific capability. The initial investigation (quick response) will have to include some method for rapidly learning the important aspects of the environment unluckily chosen by the spill, and then determining the parameters that are most important to measure in that case. A plan that is flexible enough to fit a variety of situations, but rigid enough to comply with practised and proven methods, is the ultimate goal. 55 3 . 4 References Anderson, J.W. 1975. Laboratory studies on the effects of oil on marine organisms: an overview. Publ. American Petroleum Institute 4249. Washington D.C. Aneer, G., A. Lindquist and L. Westin. 1978. Winter concentrations of Baltic herring ( Clupea harengus var membras L.). Cont . Asko Lab. Univ. Stockholm! 21: 1-19. Baker, J.M. 1971. The ecological effects of oil pollution on littoral communities. Proceedings of a symposium. Edited by E.B. Cowell. London, Institute of Petroleum. Baker, J., J. Addy, B. Dicks, S. Hainsworth, D. Levell, G. Crapp and S. Ottway. 1977. Ecological effects of marine oil pollution. Rapp. P. -v. Reun. Cons. int. Explor. Mer. 171, 196-201. Blumer, M., Sass, J., Souza, G., Sanders, H.L., Grassle, J.F., and Hampson, G.R. 1970a. The West Falmouth oil spill. Woods Hole Oceanographic Institution. Ref. 70-44. Woods Hole, Mass. Blumer, M. and Sass, J. 1972. Indigenous and petroleum derived hydro- carbons in a polluted sediment. Mar. Pollut. Bull. 3:92-3. Boehm, P.D. and J.G. Quinn. 1978. The persistence of chemically accumulated hydrocarbons in the hard shell clam, Mercenaria mercenaria . Mar. Biol. 44:227-233. COMS. 1978. In the wake of the Argo Merchant. Proc. Symp . Jan. 11-14, 1978 Center for Ocean Management Studies, Univ. Rhode Island, Kingston, R.I. USA. 1-181. DiSalvo, L.H. and H.E. Guard, 1975. Hydrocarbons associated with suspended particulate matter in San Francisco Bay waters. In Proc. Conf. Prevention Control Oil Pollution. Publ. American Petroleum Institute. Washington, D.C, 173-196. DiSalvo, L.H., H.E. Guard, and L. Hunter. 1975. Tissue hydrocarbon burden of mussels as potential monitor of environmental hydrocarbon insult. Ev. Sci. and Tech. 9:247-251. Fossato, V.U. , and W.J. Canzonier. 1976. Hydrocarbon uptake and loss by the mussel Mytilus edulis . Mar. Biol. 36:243-250. Gundlach et al. 1978. Some guidelines for oil-spill control in coastal environments: based on field studies of four oil spills. Presented at ASTM symposium on chemical dispersants for the control of oil spills, Williamsburg, Va . October 1977. In press. 56 Hess, W.N. (ed.). 1978. The Amoco Cadiz oil spill. A preliminary scientific report. NOAA/EPA Spec. Rep. 1-283. Kanter, R. 1974. Susceptibility to crude oil with respect to size, season, and geographic location in Mytilus californianus (Bivalvia). Univeristy of Southern California Sea Grant Report, USC-SG-4-74, 43 pp Krebs, C.T. and K.A. Burns. 1977. Long-term effects of an oil spill on populations of the salt-marsh crab Uca pugnax. Science 197: 484-487. Lee, R.F., Sauerheber, R., and Benson, A. A. 1972. Petroleum hydrocarbons: uptake and discharge by the marine mussel Mytilus edulis . Science, Washington, D.C., 177:344-6. Mayo, D.W. , D.S. Page, J. Cooley, E. Sorensen, F. Bradley, E.S. Gillfillan and S.A. Hanson. 1978. Weathering characteristics of petroleum hydrocarbons deposited in fine clay marine sediments, Searsport, Maine. J. Fish. Res. Board Can. 35: 552-562. MITRE 1978. Oil Spill Workshop: results of the Region I workshop on oil spill ecological damage assessment. MITRE Technical Report: MTR-7843 July 1978. Moore and McLaughlin. 1978. Design of field experiments to determine the ecological effects of petroleum in intertidal ecosystems. Prepared for NOAA and EPA. Resource Management Associates Report 6200, April, 1978. North, W.J. 1973. Position paper on effects of acute oil spills. Background papers for a workshop on inputs, fates, and effects of petroleum in the marine environment, National Academy of Science, Washington, D.C., 745-765. Notini, M. 1978. Recovery of a littoral community in the Northern Baltic following an oil spill. J. Fish. Res. Bd. Canada, 35:745-753. Owens, E.H. 1978. Mechanical dispersal of oil stranded in the littoral zone. J. Fish. Res. Board Can. 35: 563-572. Pelkonen, K. and P. Tulkki. 1972. The Palva oil tanker disaster in the Finnish SW archipelago. III. The littoral fauna of the oil polluted area. Aqua Fennica 1972, 129-141. Sanders, H.L. 1978. Florida oil spill impact on the Buzzards Bay benthic fauna: West Falmouth. J. Fish. Res. Board Can. 35: 717-730. Southward, A.J. and E.C. Southward, 1978. Recolonization of rocky shores in Cornwall after the use of toxic dispersants to clean up the Torrey Canyon spill. J. Fish. Res. Board. Can. 35:682-706. 57 Stegeman, J.J. 1974. Hydrocarbons in shellfish chronically exposed to low levels of fuel oil. In Pollution and Physiology of Marine Organisms. Academic Press, New York, 1974. Stegeman, J.J. and Teal, J.T. 1973. Accumulation, release and retention of petroleum hydrocarbons by the oyster Crassostrea virginica. Mar. Bio. , 22:37-44. Straughan, D. (ed.). 1971. Biological and oceanographical survey of the Santa Barbara Channel Oil Spill, 1969-1970. Vol. 1. Biology and Bacteriology. Sea Grant Publ. No. 2. Allen Hancock Foundation, University of Southern California, Los Angeles. Straughan, D. 1977. Biological survey of intertidal areas in the straits of Magellen in January 1975, five months after the Metula oil spill. In Fate and Effects of Petroleum Hydrocarbons in Marine Ecosystems and Organisms. Ed.: D.A. Wolfe, Pergamon Press, New York, USA, 247-260. Teal, J.M. and J.W. Farrington, 1976. Accumulation, release and retention of petroleum hydrocarbon by the oyster. Crossostrea virginica . Mar. Biol. 22:37-44. Teal, J.M., K. Burns and J. Farrington., 1978 . Analyses of aromatic hydro- carbons in intertidal sediments resulting from two spills of No. 2 fuel oil in Buzzards Bay, Massachusetts. J. Fish. Res. Board Can. 35: 510-520. Vandermeulen, J.H. and D.C. Gordon Jr. 1976. Reentry of 5-year-old stranded bunker C fuel oil from a low-energy beach into the water, sediments and biota of Chedabucto Bay, Nova Scotia. J. Fish. Res. Board Can. 33: 2002-2010. 4. Impact of Oil on the Pelagic Ecosystem CHAPTER 4: IMPACT OF OIL ON THE PELAGIC ECOSYSTEM (Sif Johansson) 4.1 Introduction Hardly any study has verified severe oil-induced effects on the pelagic ecosystem following an actual spill (Sanborn 1977, Michael 1977, Kiihnhold 1978) . Existing knowledge is derived mostly from laboratory experiments using unrealistically high concentrations of oil. In such experiments, phytoplankton clearly react in a very diverse manner. Even within the same taxonomic group two species can react in a totally opposite manner (Prouse et al., 1976; Winters et al., 1976; Dennington et al., 1975; Mironov, 1968; and others). Depending on the composition and physiological status of the phytoplankton community and the type of oil used, primary productivity can be inhibited, stimulated or affected not at all when exposed to realistic concentrations (H'siao et al., 1978; M Bender et al., 1977; Gordon and Prouse, 1973). Comparatively few investigations have dealt with effects on primary production, biomass and species composition at the same time. In large plastic bag (CEPEX) experiments, Lee et al. (1977) found that the rela- tive dominance of a diatom Cerataulina bergonii , had decreased from 50-95% in control bags, to only 10% in oil contaminated bags in 8 days. The decrease was followed by an increase of microflagellates and, de- spite a lowered total biomass, primary production increased. This was explained to be a result of higher growth rate among microflagellates. Conover (1971) following the wreck of the Arrow reported oil con- taminated zooplankton in situ . Droplets of oil had been ingested or adhered to the feeding appendages. Though as much as 10% of the total oil spill was estimated to be associated with zooplankton, no apparent effects could be detected. Parker and Watson (1969) observed ingestion of oil by zooplankton, both experimentally and in situ and also found no harmful effects. Lee et al. (1977) in enclosed experiments showed oil-induced minor alterations, i.e., an increased abundance of ciliates and rotifers. A 61 lower growth rate for copepods was also indicated, but the total stand- ing stock remained unchanged. 4.2 Material and methods 4.2.1 Sampling stations Two stations in the spill area (IV and V, see map, Fig. 4.1) were selected for intense monitoring during the month following the spill. The area in which station IV was located was directly exposed to surface oil slicks only once, when a minor slick, which had been re- leased from the tanker on removal from the area on November 3, drifted quickly by. In the area northeast of the grounding site, at station V, large oil slicks occurred during the first week after the accident. Even at the end of the investigation period minor oil slicks were seen drifting in this area. Immediately after the grounding, zooplankton net hauls were taken at a station close to the grounding site (station II). Bacteria were also sampled once at stations II and III (October 31). Two stations, I and VI, enclosing the spill area, are regularly sampled by the Asko Laboratory. The research programmes include all parameters sampled in the spill area and thus data from these stations could be utilized for reference. Lack of resources limited the sam- pling, which unfortunately hampered statistical analysis of the results. The salinity in the impacted area varied between 6.6 and 7.0 /oo S in the surface water and between 6.6 and 7.9 /oo S at 20 m depth. The temperature in the water gradually decreased, from 8.4 C to 6.4 C at the surface and from 8.4 C to 6.9 C at 20 m depth, from November 1 to November 17. 4.2.2 Methods 4.2.2.1 Phytoplankton Phytoplankton were collected with a plastic tube 20 m long, care- fully avoiding surface oil films. All samples were preserved with Lugol's solution containing acetic acid. Phytoplankton were counted 62 MORKO (sTATIONV TORO STATION VI y \ *S» 12 3 4 5km Fig. 4.1 Position of stations II, III, iv, V and the reference stations I and VI. 63 using the Utermohl (1958) technique. Small cells (<10 pm) , with 1-4 flagellas, belonging to Chrysophycea , Cryptophycea and Chlorophycea we i counted collectively and are referred to as monads. 4.2.2.2 Primary production Water samples were collected, while carefully avoiding surface oil films, and incubated for four hours at 0, 1, 2, 4, 6, 8, 10, 15, and 20 m depth. Dark bottles were incubated at and 20 m depth. Measure- ments were carried out as described by Larsson and Hagstrb'm (1979). 14 Four pCi of carrier free NaH C0„ were added to all bottles. After incubation, 10 ml of the samples were transferred to scintillation vials, and counted in Instagel (Packard Instruments) in an Intertech- nique SL 40 liquid scintillation counter. The uptake of carbon was calculated according to Gargas (1975). 4.2.2.3 Bacteria Bacteria were sampled at a depth of 2 m using a sterile Niskin water sampler. At reference station VI the number of bacteria was determined in an integrated water sample (equal aliquots of water from 0, 5, 10, 15 and 20 m were pooled - a standard procedure in the routine programmes at these stations (see Hagstrom et al., 1979). The bacteria were preserved in formaldehyde containing acridine orange and counted in an epif luorescence microscope, as described by Hagstrom et al. (1979). They also described the method used for determining the frequency of dividing cells (FDC). 4.2.2.4 Zooplankton Zooplankton was sampled by vertical net hauls from bottom to sur- face, using a UNESCO WP-2 net with 90 pm mesh size. Special efforts were made to avoid contamination from surface oil films. The samples were preserved in 4% formaldehyde buffered with hexamine. Counting and species determination were performed using an inverted microscope. 6U 4.2.2.5 Sedimentation Sediment traps, a PVC cross with 20 cylindrical glass tubes (0 26 mm, length 200 mm), were positioned at a depth of 20 m. The sedimented matter was divided in two parts; one was used for deter- mination of dry weight and the other was transferred to a glass jar (cleaned with hexane) and stored deep frozen until analysed for oil. 4.3 Results 4.3.1 Phytoplankton 4.3.1.1 Phytoplankton biomass At both reference stations, total phytoplankton biomass remained _2 fairly constant, around 600 rag C m , throughout the period (Fig. 4.2). In the impacted area phytoplankton biomasses were mostly more than a factor of two higher in the two weeks following the spill, but gradually decreased and approached the level at the reference stations by the end of November. There is no significant difference between the two stations (IV and V) in the impacted area (p = 0.32, rank sum test according to Dixon and Massey 1969:345), but valid statistical comparisons between the impacted area and the reference stations cannot be made, due to the long time spread and few measurements at the latter stations (see Fig. 4.3). 4.3.1.2 Phytoplankton species composition At all stations monads constituted 75 to 90% of the total biomass and diatoms 10 to 15% (dominating species: Coscinodiscus granii and Skeletonema costatum ) throughout the investigated month. The remaining few percent were peridineans ( Gymnodinium sp. and Gyrodinium sp.). A few species belonging to Cyanophycea and Chlorophycea were also present, but their contribution never exceeded one percent. This is not an abnormal phytoplankton composition for autumn in this area (Hobro, in press) . 65 22 o Eh < PL, o Eh >n ex. u I I I ' I I I I I L_J 1 1 1 1 1 1 1 25* I 1 5 7 J I L JL J L J 1 1 L OCTOBER Fig, 9 11 NOVEMBER -2 14 17 4.2 Phytoplankton biomass, gC m " , at stations IV ( — ), V (o o) and the reference stations VI ( C3 ) and I ( mm ) . Date of grounding is narked with an arrow 23 300 H O i— i P 1 Q 'O O Pd CM P-i 1 B >i V, u S 00 H e P!j Cu 200 100 INSOLATION cal cm d -2000 1600 -1200 800 -400 -0 J L 25 j i i i » * i |_ J I I L J I I I L i Fig. OCTOBER 4.3 9 11 14 NOVEMBER J I I L_J I 'III 17 23 —2 — 1 Primary production, mg C m '" d , at stations IV (• ), V (o — o) , and reference stations VI (o ) and I (m )• Insolation values (dotted line) are scaled at the right, Date of grounding is marked with an arrow. 4.3.2 Primary production The reference stations showed the normal autumnal decrease in primary production, compared to unpublished Asko Laboratory data from six previous years, with values at the end of November only one fourth of those at the end of October (Fig. 4.3). The primary production in the impacted area tended to be higher compared to values from the refer- ence stations. Fluctuations were induced by changing light conditions, (dotted line in Fig. 4.3), as indicated by the rather clear correlation (linear regression, r = 0.45, N = 17, all stations combined) between insolation and primary production per biomass (Fig. 4.3 and 4.4). Sta- tion V in the most contaminated area, normally had higher primary pro- duction than station IV, and the difference borders on statistical significance (p = 0.06, sign test, Dixon and Massey, 1969: 335-340). 4.3.3 Bacteria The total number of bacteria was higher at the contaminated sta- tions than at the reference stations (Fig. 4.5). Five days after the grounding there were about three times as many bacteria at stations II and III (1.15 x 10 ml" 1 ) compared to station VI (0.35 x 10 ml" 1 ). In November, the difference was a factor of about two. Hagstrom et al. (1979) reported the standard deviation of bacterial counts to be 9 ± 4%, which supports the reality of the observed differences. The difference in sampling strategy is unlikely to influence the results markedly, since bacteria are rather uniformly distributed in the water column during this time of the year (Hagstrom et al., 1979). In a comparison with data from the same time of year from both station VI (1977 and 1978) and two stations (1977) in the more eutrophicated area north of station I, the bacterial counts from the oiled area stand out as unusually high (Hagstrom et al., 1979; Larsson and Hagstrom, pers. comm. ) . The measurements of the frequency of dividing cells (FDC) showed no clear differences between stations (Table 4.1). 67 zn en o H PQ H U ID P O pci P-. I PI Pi Pm 0.5 OCTOBER Fig. 4.4 The production per unit of biomass at stations IV ( - V (o o) , and reference stations VI ( C3 ) and I ( ■ ) Date of grounding is marked with an arrow. ), vO O u 0) 6 3 (3 Pi W H u d 01 T3 QJ w d d o E CO i-H CO +J o 4-1 X) d CO *v S-l Qj 4-> ■U CO E T3 /— > 0) T3 4-> O d •H QJ M E OJ •H ft TO qj u w QJ Oh <4H o X) d 4-> CO si oo >» •H CO 0) X> 3 Sh >> QJ $-i Oh 13 w d QJ •H >H 4-> r-H QJ •H E o QJ <+H U O CO d +-> a 1 d w d o Vh -1 QJ ft CM H O CM E "O CU 00 00 £ t— II •H T3 O CM I • E X) OJ 00 x> QJ w •H O 9 >h m x> o tH •I U 00 d o oo a n. X> CM 5-1 OJ E 4-> 4-1 4-> CO IS E >> • 5-1 x> XS aj CO oo X O •H S-l QJ ft d o •H +J CO ■M CO o ^o ch r-~ O Cn 00 \Q LO r-« vo cni CN CO 00 00 O 00 O O CO r-H CM LOCOCO CTi ^O cn a\ co i-H h* cm en vD CM CO r-H O r- o> cm r— i u o^ r-H U Ci r— 1 u > > QJ > > QJ > > QJ o o Q o O P o o Q 125 1 55 i 55 1 55 1 1 55 55 1 1 r-H o\ i— l CM STl r—i CM On r-H >>> >>> >>> ooo ooo ooo 555555 555555 555555 u QJ +J ft CO si u o u P=H QJ d r-H CO > d CO QJ E QJ Xi 4-1 00 d •H W d X QJ 4H CO r-H d o l—i CO u •H O X QJ 4-> d cu E •rH X QJ W 4H O 4-> d d o oo d Sh QJ 4-> 4-> CO E X QJ 4-> d QJ E •H t3 • QJ > W 4H d O CO 4-> > d HH d O 03 E d CO O •H d 4-i O CO 4-> CO w 4-» CO 4-1 Q co 72 613 20 -a CM 1 s B0 bO 4a hJ M o 20 Q W H S5 W BB S H Q W en -40 20 1 - " station V - i station II - - station IV - - OCTOBER 9 17 NOVEMBER 14 21 DECEMBER -2 -1 Fig. 4.8 Sedimented amount of oil, mg m d at stations V, II and IV. Calculated as explained in Table 4.2. 73 4.4 Discussion The pelagic system is first to suffer from an oil spill. Depending on local geography, distance from shores, windstress, etc., the exposure of the pelagic system to oil will change over a period of time. Soon after release from the tanker the oil starts to change both chemically and physically, e.g., volatile fractions evaporate and soluble fractions enter the water phase. The rate of this process decreases rapidly with time. Combined with a continuous dilution this weathering will limit the period of detectable ecological effects. The results from this study show only minor effects on the pelagic system. Changes in species composition could not be detected either in the phytoplankton or the zooplankton community. The phytoplankton biomass almost certainly increased in the affected area. This may have been due to decreased zooplankton grazing or increased growth rate. The existing data on productivity per unit biomass indicated a normal ratio, and the fact that zooplankton was found to be heavily contaminated with oil (50% of specimens with visible oil droplets in the first week, 20% after three weeks), mainly on the feeding appendages, makes decreased zooplankton grazing the more probable explanation. Similar results were found after the blow-out on the Ekofisk Bravo platform (Lannergren, 1978). No significant differences in zooplankton composition or biomass could be detected, except at station II immediately after the spill, when two measurements from separate days (Oct. 28 and 29) showed a drastically lowered biomass. Gyllenberg and Lundqvist (1976) have shown that zooplankton, when exposed to oil, either try to escape or enter a state of "narcosis". Either of these mechanisms could explain the very low biomasses near the wreck. Bacterial abundance increased in the contaminated area. It is however, impossible to judge whether this was a consequence of increased growth rate or decreased grazing. An effort was made to estimate bac- terial growth rate from the frequencey of dividing cells (Table 4.1), but no clear differences were found between polluted and unpolluted areas . 74 Perhaps the most interesting results from the pelagial study were obtained from the sediment traps. During the first period of sediment trapping (Nov. 1-9), very high amounts of oil (up to 0.7%) were recorded in the sedimented matter. This was also true for station IV situated about 2.5 km windward of the tanker. Unfortunately, no traps were positioned before November 1st, which leaves the period just after the spill, with potentially high sedimentation of oil, uncovered. This is likely to be a larger source of uncertainty in the following calcula- tions, than the methodological uncertainties inherent in all attempts at measuring sedimentation. Between Nov. 9 and Nov. 17 there was still a substantial oil content in sedimented matter at stations II and V, whereas at station IV practically no oil was found. From Nov. 17 on- ward, oil was still found in sedimented matter from station V, probably as a consequence of release from the shores due to waves and cleaning operations . Several mechanisms can facilitate sedimentation of oil, e.g., weathering of the oil leading to increased density, adsorption of oil to particles, or ingestion by zooplankton. Conover (1971) found oil incorpor- ated in zooplankton fecal pellets, a mechanism which, through the rela- tively high sinking rates of fecal pellets, will accelerate sedimenta- tion. In the present case it is, however, unlikely that zooplankton itself or the production of fecal pellets significantly contributed to the sedimentation of oil. At this time of year zooplankton biomasses are low, near the yearly minimum. A more probable path is sedimentation through adsorption to detritus particles, since seston levels in the area are normally high in the autumn, due to wind-induced re-suspension of bottom sediment. Only two days before grounding, the area was sub- ject to fairly strong southwest winds. Wind speeds of up to 10-14 m/s also occurred several times during the acute phase of the spill. The sediment trap data make it possible to estimate roughly the amount of oil leaving the water phase through sedimentation. The af- fected area has been estimated as the area inside a line connecting the outermost points where oil has been found, either through visual observa- 2 tion (Fig. 4.9) or by direct measurements of oil. A total area of 42 km 75 ■" • ,' '" 'H Jb>ur -V - 7/ vVHl^v AY '. J*" ■ ir r- — \Vj ^t^^-^iv Kubii \ ut -^ »/.HirfJ i . ■•a.^?_ A ^ ■'H. A > ASKON , '^ •■ zi» U-W-. v\S^ 7 - .---i'tr^:M ^'^ v HA' '.:r4^^ 1 13 J. -5 ' :_s...,™ Figure 5.5.1 Aerial reconnaissance on November 1, 1979; wind from the south. Areas with heavy visible oil concentrations are indicated by dots. Light sheen is not indicated. 94 6. Impact of Oil on Deep Soft Bottoms CHAPTER 6: IMPACT OF OIL ON DEEP SOFT BOTTOMS (Ragnar Elmgren, Sture Hansson, Ulf Larsson and Brita Sundelin) 6.1 Introduction 6.1.1 Background Since the Torrey Canyon catastrophe in 1967, millions of dollars have been invested in research on the biological effects of oil pollu- tion of the seas. Innumerable scientific papers, summarized in many books and reviews (e.g. GESAMP 1977, Mclntyre and Whittle 1977, Cowell 1977), have resulted. Most of these studies have, however, been con- cerned either with the surface layer of the sea, where plankton, fish and fish eggs as well as sea birds may be affected by a spill, or with the intertidal zone, where stranded oil may cause extensive destruction of the natural communities. Relatively little attention has been given to the effects of oil on subtidal benthos communities, even if a few good field studies exist (e.g., Addy et al., 1979; the West Falmouth oil spill study, Sanders, 1978). These few studies are all, however, concerned only with the benthic macrofauna, while the smaller meiofauna is totally ignored, even though its importance in energy flow terms is often similar to that of the macrofauna. This general picture is also valid for the Baltic Sea. Furthermore the Baltic ecosystem has so many unique features, that the usefulness of studies from tidal and more fully marine areas for risk evaluation concerning various types of pollutants in the Baltic is questionable. Almost the only study of oil impact on a Baltic soft bottom community is that by Leppakoski and Lindstrom (1978), and treats continuous oil pollution from a refinery, rather than an acute spill situation. The small study of the Palva spill by Mustonen and Tulkki (1972) should perhaps also be mentioned, even though it is rather inconclusive. This comparative dearth of information is the more unfortunate, as studies from the intertidal zone show that recovery from an oil spill is slowest in fine sediment environments, where oil may persist virtually 97 unchanged in the deeper, oxygen-free layers for at least five to ten years (Krebs and Burns, 1977). This persistent oil continues to present a hazard to the biological community, preventing its return to pre-spill status, and constituting a potential source of slow, continuous oil leakage to surrounding areas (Vandermeulen and Gordon, 1976). 6.1.2 The spill area The Tsesis spill occurred in an area where benthic data have been collected since 1972 by Ulf Larsson, Asko Laboratory, in connection with a study of the environmental impact of the large modern sewage plant "Himmerf jardsverket" . Pre-spill macrofauna data were already available, while meiofauna samples are stored and will be sorted later if funding becomes available. Furthermore, the benthos of the Asko area has been treated in detail in a number of earlier papers (Cederwall, 1977, 1978; Ankar, 1977; Ankar and Elmgren, 1976; Elmgren, 1976) so that the back- ground knowledge for the study of the oil spill was exceptionally good. Stations 20 and 21 (see map, Fig. 6.1) were generally sampled for macrofauna once a year in October-November (i.e. at about the time of the spill) starting in 1972. The depth of the stations were: station 15: 44-45 m, station 20: 32-33 m, station 21: 28-29 m, and the bottom substrate, mud at all the stations. Stations 20 and 21 are located in an area, "Svardsf jarden" , which is cut off from exchange of deep water with the open Baltic by a sill of about 20 m depth between Asko and Toro, whereas Station 15 is outside this sill. 6.2 Methods 6.2.1 Sediment sampling Cores for oil analysis of surface sediments were collected using either a modified Kajak core sampler (Kajak et al., 1965) of 80 mm internal diameter (10 Nov. 1977, station 1 and 2; 17-31 August 1978, station 14, 15, 20, 20C, 20D) or the Asko corer, also used for meiofauna sampling (inner diameter 22 mm, used for all other samples). The top 2 cm of the sediment were extruded into hexane-washed glass bottles and kept deep-frozen until analyzed. On 4 December 1978, further cores 98 ^<^7 r,3 -"a i f — > mosk6 ( I / y>% \ ^ r „.Sa£> .^x ASK6LA3?> ^ J ai as: S c ^ ^ J V,\ <* »?A • "* % 33 77 10 V T r 7J >~k / ■a i \ Mf&Zf £km '*-> Skarimge \ •/ 7 Skdrlm^ei-ik ■■* '-* ** fa ^yi I 2/ "it w 1 viken 69- »S^p ) ; La vi ^:> Rav " y1 S VJ . ) I \ ~-^&*.r.i.:Ys, ■ 34 '»j •' is siyjSkoffabma. • j Bynanjtorp \ v \:i .ft,;,, \ jyc /"7i "\ ,,> .iV^IWY^ 1 fTonnas J.--/ "\ La. tf 1 'A » I 33 \ Si- Ju . - ^ rflt / m/ ^ soT^.j,i|. /7 *.s ]/<: w ^p> -TV ^ $ •-JUlberga 16 9 , » 'Ju.rtnkjw. 34 « 24 Grdnsdqr'l? 33 '"' ,. '^ 2* So r T la 14 33 * .\ppelgarn „, 3 « 20 c 2 2, , ' Syvtksgr 18 37 81 7J .' 77 2(9 *yirt* " a v r t - y ^^^/ zl » 7h" r- 87 /y a, )k\iop \ » 56> "'" 8,5 .\i I 74 /i Sankhallan w i T j » A 27 « 47 5 ■ I U « * V >, W ^Xors/i J3 ^75 if * 11 10 V6*.- TS h 13 c' 2 ' /? la 7,2 - K/l'itiflt 30 ' 10 7* Styrmannen ?jl ,,/ Bx.«.r|. ! ,,-A- ::W; J / ' _ V - r / : y f&rashi/ j-yaii ■ y Svardsb / ^. ^^"-v f.,./ a* ST? ; \ |>. / .w! A /- / ':r*f - 7/ S n . \J.indhobi>iai Va/ M / g .J ;/ ^-fSvardsd \ ~ St ' r «^A i y ,7 , 0** W < \S6Jruj ^_ ' 2« "' ilrhorni-t w 15 8\ ) \S-B'J aS 47 95. "j\ [ j ' VJ1 " 17 ** > % \ &/ b 5S5 22 9 "* TafeHadan Kfypi/r «/ v^y Lisoudde « "\j V*5 "'*[ / S« «.-7 M , ' **'i*l-;/ 7.<* An^§ /l9 73 « o r 7 H ^ Fifon( orvikah 1 f / x *--.„„ ** *. TLsteOwbnshl™ v. .2 (J" ««■■/»• '*% U 22ia 20 29 " ^7 X^ 7o ;/ I'i. 1-, Sarin. r '3o m # ' _.,-, 2 v 11 , 24 17. La. 30 ■y!7^» 7,e fj, ,'TH. XSringH. „ XBrtyhtf -- \/ Brannnn*^ " * 27 '■* 10 5 J0 ?2 .;?Uiv^° 3 ' 24 lllnn 35 31 U3 33 ^ 2 3.7 33 30 La. 33 W 23_ W 10 f .It"'''' « 25 3* J7 La. TSESIS 1U * rf^A 9F~AritL \93 -,^7 25 -c faa « l x *t AdnVsgr* V 23 2* 8 ■'. N * "V -3S 5fl ^2 2^X 8s S» n '* • ergholmen 7 * > I /"_/■ TaRskdr 72 I .HerrhanLra Y/" 1 7j'^v'\ Fi§ * 7#2,1 Location of stations used \ t \-4s. ** : in t ^ ie littoral studies \'\ f 77? % f ^2 %? I I L ' Km 1 ****-*" ,,, \ W )k>- ) Kofg,AJ, ** \ •■/ \WB>iav.r«Aft,--')^y ''"""" 4_ V t r, ; i" 5 ^ jf) rrr,k!5t.r/ n . . 'V -^ 3 4 5 KM r Ts ■^.7. ;;:.- J. ± j '£ If- 7-"^ \ LVA* .. : r Table 7.2.2.1 Fucus Bag Sampling Number of samples taken Station: Date of sampling A B C D 10.27.77 (5) X (5) X (5) X (5) X 11.02.77 (5) 11.09.77 (5) (5) (5) 2(0) 5(0) 5(4) 11.15.77 (5) 12.14.77 (5) (5) (5) (5) 5(0) (5) 5(0) 5.02.78 5(0) 5(3) (5) 5(0) 5(0) (5) 5(3) 6.20.78 5(1) 5(0) 5(0) 5(0) 8.28.78 5(0) 5(0) 5(0) 5(0) 5(0) 5(0) 5(0) 10.30.78 5(0) 5(3) (5) 5(0) 5(0) (5) pre-spill samples ( ) number of samples examined 133 only 8-10% of the pre-spill situation at stations A and C. Station B between these two stations was probably affected in the same way (Fig. 7.2.3.1). During the same period no such decrease occurred at the "reference" Station G. Station D, which was hit by the oil four days later, showed a possible decrease in magnitude of 40-50% although the variations in the samples were high. Samples from stations F and E have not been sorted yet, but observations during diving operations in the area indicated less acute damage compared to stations A, B and C. In spite of the initially heavy degree of oil pollution, a sig- nificant recolonization of the Fucus fauna had already started in the middle of December 1977 at Stations A and C. At Station C the total number of macrofauna specimens in October 1978 was of the same order as in October 1977 before the oil hit the station. At Station G also, the "reference" station, the situation in October 1977 and 1978, respective- ly, was very much the same. The process of recolonization seemed to be slower at Station B and possibly at Station D. The crustaceans Gammarus spp., Idotea spp . and Iaera spp . (Fig. 7.2.3.2, 7.2.3.3 and 7.2.3.4, respectively) were all drastically reduced by the oil. The specimens of the amphipod Gammarus spp. were sorted into two length classes, > 10 mm and < 10 mm (Fig. 7.2.3.2). The pre- liminary results show no differences in the effect of the oil on these two groups. In November 1977 Gammarus spp. was totally missing or significantly reduced at all examined stations hit by the oil (Stations A, B, C and D) compared to the pre-spill situation in October. But a recolonization had already started in December 1977 at stations A and D. The isopods Idotea spp. and Iaera spp. (Fig. 7.2.3.3 and 7.2.3.4) were not totally missing in the November samples, but individuals remaining were very few. At station B Iaera spp. was completely missing both in December and in May, half a year after the accident. Recolonization by Idotea spp. at the same station seemed likewise to be a slow process. However, in October 1978, one year after the Tsesis ran aground, a considerable recolonization of both species had taken place at station B 134 ■p x: tn ■H !h W U Cn o o n3 3 ■H > •H C 4-1 e p 1300 - r-S-i OCT NOV DEC MAY JUN n=1 AUG OCT 1000 - B 3r~i x i i x i NOV DEC MAY n=3 AU G OCT n=3 1000 i ir OCT 2000 D NOV DEC MAY JUN AUG OCT 1000 OCT NOV NOV DE :c MAY JUN AUG OCT - NOV DEC MAY AUG OCT 1000 f f- i OCT NOV DEC MAY n = 3 JUN AU 6 OCT Fig. 7.2.3.1 Total number of macrofauna specimens in Fucus samples first year following the spill. [ = 95% confidence interval for population means. 135 50 in ^"fl la 37 5 JUVENILE OCT NOV DEC MAY JUN AU G OCT -p Cn ■H OJ >i in T3 W P En Cn o o TJ •H > •H ■H m o >-i 10 mm = juveniles < 10 mm 136 total number D ioo -i +j X! CT> •H 0) <: >i W u PL, o o CD rd •H > ■H T3 C ■H M-l CD g z OCT 0.9 NOV DEC NOV DEC 250 2.2 OCT NOV 20 D OCT NOV NOV 40 - NOV 100 - T _X_ 1 | I -L j OCT NOV i ii ■ iflfflff.i. ... .i^ DEC DEC i DEC DEC MAY JUN n = 1 MAY n=3 MAY JUN MAY JUN MAY £ m MAY n=3 JUN AUG AU6 AUG AUG AUG AUG OCT 250 - B 7.7 2.2 10.8 OCT n=3 OCT OCT OCT OCT Fig. 7.2.3.3 Number of Idotea spp. in Fucus samples first year following the spill. j = 95% confidence interval for population means. 137 400 -. OCT NOV DEC MAY JUN n=1 AUG OCT 25 _ -p xi tn •H CD M T3 o Pm tP O O CD r0 ti ■H > •H c ■H 4-1 o S-l 0) X! B B ± N0V DEC MAY AUG OCT n=3 n=3 100 - , C r-K- ■ r^n OCT NOV DEC 500 D r^n OCT NOV NOV DEC NOV DEC OCT MAY JUN AUG OCT MAY JUN AUG OCT 400 - .-■ F MAY AUG OCT 200 - r3h ~r G i X 1 NOV DEC MAY n=4 n=3 Fig. 7.2.3.4 Number of Iaera spp JUN AUG OCT in Fucus samples first year following the spill. j = 95% confidence interval for population means. 138 The dominating bivalve in the area, Mytilus edulis , showed great fluctuations in the Fucus samples, both within a sample set and between the sets (Fig. 7.2.3.5). However, a decrease in density figures in November compared to October due to oil was found. During the same period no such decrease occurred at Station G. The effects are there but closer analysis of the collected data is needed. When the Mytilus edulis from stations C and G are sorted into two groups (> 5 mm and < 5 mm), comparison of these size groups strongly indicates a mortality of small individuals (< 5 mm) at Station C (Table 7.2.3.1). The abundance of Theodoxus fluviatilis, the dominating gastropod in the area, also decreased significantly during the first months following the spill (Fig. 7.2.3.6), but at station C recolonizat ion was very slow. In October 1978 the mean density figures were still significantly lower than in October 1977. 7.2.4 Discussion Unfortunately, sorting and examination of Fucus samples is very time consuming. At present 49% of the collected Fucus samples have been inspected. In spite of this the available data presented here strongly indicate drastic effects on the Fucus macrofauna in the area. The abundance of all macrofauna species, with the possible exception of the barnacles Balanus improvisus, decreased during the acute phase at sta- tions affected by the oil. Comparison with the reference station G on the island of Fifong shows no such decrease there during the same period. The degree and duration of the damage varied at the different stations. From the results available, station B and C were probably the stations most affected, followed by station A. Station D showed a smaller degree of acute effect and a preliminary look at the samples from station E shows the same type of results. This may be explained by the fact that the oil reached these locations 5 to 10 days after the oil spill occurred. Therefore, this oil had been weathered for a longer period of time compared to that at stations B, C and A. Both these stations (D and E) are situated in shallow bays with low water exchange. They are the most protected of the stations studied here. Therefore the 139 200 - 100 - A OCT NOV DEC MAY JUN AUG OCT -p A tn ■H 0) >^ CO tJ> o o S-l cu CO H 03 3 •H > •H c •H 4-1 U CD Si e 100 B T r [*] j- NOV DEC MAY AUG OCT n=3 n=3 -p ^^ 100 - .. ~p C l I «4» . 0( :t NOV DEC MAY JUN AUG OCT 000 - T D 1 OCT NOV NOV DEC MAY JUN AU G OCT 1000 - NOV DEC MAY AUG OCT 400 - ^^ m G ■*■ ^ OCT Fig. 7.2.3 NOV n=4 DEC MAY n=3 JUN AUG OCT 5 Number of Mytilus edulis in Fucus samples first year following the spill. f = 95% confidence interval for population means. 140 Table 7.2.3.1 The Percental Distribution of Mytilus edulis in 2 Size Classes at 2 Stations Date Station C Station G > 5 mm < 5 mm 10.27.77 71.6 28.4 11.09.77 99.2 0.8 12.14.77 100 05.02.78 85.6 14.4 10.30.78 90.8 9.2 5 mm < 5 mm 44.2 55.8 38.9 61. 1 X 81.6 18.0 18.4 82.0 xx XX based on 4 Fucus samples based on 3 Fucus samples 141 5C0 - i -p Xi •H CD u en d o Cm CP o o U 0) (13 •H > •H c •H m o 0) X! A 2.0 OCT NOV DEC 0.8 MAY T JUN n=1 AUG OCT 50 - B • _L NOV DEC 250 - OCT NOV DEC OCT NOV NOV DEC 50 - NOV DEC 250 I OCT NOV n=4 DEC MAY n=3 AU6 OCT n=3 T 1 0.6 HH r^n MAY JUN AUG OCT >0 - -3- — — * D -*i MAY JUN AUG OCT i MAY AUG OCT 3.8 MAY n=3 JUN AUG OCT Fig. 7.2.3.6 Number of Theodoxus f luviatilis in Fucus samples first year following the spill. '; = 95% confidence interval for population means. 142 long-range effects may be of the same magnitude as at the other stations, as the residence time of the oil in the littoral system is dependent on the energy in the system (Owens 1978). The relatively fast recolonization of Station A may be explained by the fact that this station was exposed to oil during a comparatively short period of time. Due to a change in wind direction on October 28 the drift of oil towards shore stopped. Thus, this station was exposed to the drifting oil spill for less than 24 hrs. In December 1977 a significant increase of most faunal species had already occurred at station A. Only one genus, Iaera , did not increase during that period. This may be explained by the low mobility of these small animals. At Station B, in the center of the polluted area, recolonization of this species had not started in May 1978. Crustaceans are without doubt sensitive to oil pollution (Notini and Hagstrom 1974; Linden, 1976; Notini, 1978) and many of them must have died during the acute phase of the Tsesis oil spill. The results from this study indicate that the recolonization of these species to a great extent is due to a horizontal migration from the unaffected areas. The early recolonization of the molluscs was, however, largely dependent on the vertical recolonization by surviving individuals. These were narcotized during the acute phase of the spill. By wave action they were swept down from the Fucus plants to the bottom. Later, the surviving individuals recovered and reentered the Fucus. As indi- cated in Table 7.2.3.1, the mortality among small Mytilus edulis seemed to be higher than that among larger individuals. Similar observations have been made in previous studies (Notini, 1978). Bivalves have been found useful for monitoring petroleum input since they reflect the concentration and relative amounts of different hydrocarbons in the water (Lee, 1977). They are able to bioaccumulate , but cannot metabolize hydrocarbons in their tissues. Thus, the Mytilus samples from the different stations in this study are of great interest since they make it possible to estimate dose and response data in situ as well as the background oil contamination before the spill. These 143 preliminary data indicate Station C to have had the highest concen- tration of oil in water during the acute phase. But the samples also indicate relatively high amounts of aromatic hydrocarbons in the tissues of Mytilus from Station G. Together with data from sediment traps and Ma coma balthica , this shows that Tsesis oil must be considered widely spread in the area. Station G can thus not be considered as a "non- affected" station, even though the samples of the Fuc us fauna showed no clear reduction of the fauna. The data presented here clearly demonstrates drastic effects on the animal life of the Fucus community. On the other hand recolonization at several stations started soon after the acute phase. 7.2.5 References Blumer, M. , Sanders, H.L., Grassle, J.F., and Sass, J. 1971. A small oil spill. Environment 13(2):1-12. Dybern, B.I., Ackefors, H., and Elmgren, R. (Eds.). 1976. Recommenda- tions on methods for marine biological studies in the Baltic sea. Baltic Marine Biologists Publ. 1: 1-98. Guterstam, B. 1977. An in situ study of the primary production and metabolism of a Baltic Fucus vesiculosus community. In: Keegan, B.F., P. O'Ceidigh and P.J.S. Boaden (Eds.): Biology of Benthic Organisms. Pergamon Press, Oxford, 311-319. Haage, P. 1969. BlSstSngsbaltet. Zoologisk Revy 31:21-22. (In Swedish). Jansson, A.M. 1974. Community structure modelling and simulation of the Cladophora ecosystem in the Baltic sea. Contrib. Asko Lab. Univ. Stockholm 5:1-129. Jansson, B.O. and Wulff, F. 1977. Ecosystem analysis of a shallow sound in the northern Baltic - A joint study by the Asko group. Contrib. Asko Lab., Univ. Stockholm 18:1-160. Lee, R.F. 1977. Accumulation and turnover of petroleum hydrocarbons in marine organisms. In: Wolfe, D.A. (ed.). Fate and effects of petroleum hydrocarbons in marine organisms and ecosystems. Pergamon Press. New York, 60-70. Linden, 0. 1976. Effects of oil on the amphipod Gamma rus oceanicus . Environ Pollut. 10:239-250. 144 Michael, A.D. 1977. The effects of petroleum hydrocarbons on marine populations and communities. In: Wolfe, D.A. (ed.) Fate and effects of petroleum hydrocarbons in marine organisms and eco- systems. Pergamon Press, New York. 129-137. Notini, M. and Hagstrom, A. 1974. Effects of oils on Baltic littoral community as studied in an outdoor model test-system. In: Marine Pollution Monitoring (Petroleum) Symposium and workshop. Gaithersburg, Maryland, NBS Spec. Publ. 409, 251-254. Notini, M. 1978. Long-term effect of an oil spill on Fucus Macrofauna in a small Baltic bay. J. Fish. Res. Board Can. 35 (5 ): 745-753 . North, W.J., Neushul , M. Jr., and Clendennins, K.A. 1964. Successive Biological changes observed in a marine cove exposed to a large spillage of oil. Symposium Commission International Exploration Scientifique Mer Mediterranee , Monaco, 335-353. Owens, E.H. 1978. Mechanical dispersal of oil studies in the littoral zone. J. Fish. Res. Board Can. 35 (5) : 563-572 . Southward, A.J. and Southward, E.C. 1978. Recolonization of rocky shores in Cornwall after the use of toxic dispersants to clean up the Torrey Canyon spill. J. Fish. Res. Board. Can. 35 (5) :682-706 . 145 7.3 Effects on the phytal ecosystem (H. Kautsky) 7.3.1 Materials and Methods Seven stations were chosen to represent the entire area contam- inated by the oil spill (Fig. 7.2.1). Most stations were placed in coves as it was expected that these would hold the oil for a longer time and the pollution effect would be higher and easier to detect. In some cases the oil was, in fact, forced into the coves with the help of spill-booms. All seven stations, including a reference station (G) with no visible oil spill, were sampled in November 1977 and resampled in June 1978. From the June sampling only 3 stations have been sorted (B, D and G) . Sampling was done by SCUBA divers, using the technique described in Dybern et al. (1976) and by Jansson and Kautsky (1977). The vegetation coverage and the Mytilus edulis coverage were estimated visually. The vertical extent of each identified belt was noted. Within each zone, quantitative samples were taken at random. In November, the sampling sites were marked with bricks; in June, random sampling was made in the vicinity of these bricks in order to facilitate comparison between November and June samples. Square frames with a side of 15, 20 or 50 cm were used depending on the kind of vegetation dominant in each belt. The samples in November were sorted and dried shortly after collection; in June the samples were frozen and sorted later. The samples were analyzed for species composition, abundance and 2 biomass. The results are given in individuals per m or in g dry weight 2 per m . 7.3.2 Results 7.3.2.1 Field observations November: At the most contaminated stations B, C and D (see Fig. 7.3.1) no free-swimming animals could be observed. At location B an oil smear was identified on Mytilus edulis down to a depth of 4 m. 146 CO-T •TT— CO •*» y ^^X CO 1 ?- '*■* 1 v\ 4-1 CO •i,i*" I - ' - ,'_ x \ co e I , ' " - j - 1 • v 1 u_ o CO CO -H />•?' - x s ,"- ~ 1 - v / ' S C H ' ' l~ \- \/ •H Oi T3 Zlz~ Pu cO cO -<^~ E co ''/ x " ,7ci~ CO CD r' ^ N v \ ' - \ CO CJ [ - ~\ . '::/ ju ■ « c B co co *\Z*\ O N Tl SS& (F) fr June 19 n, abun co o 3 T3 -H /j CJ pL4 £ 4-J CO -H CO r- o T3 r^ CX C C^ g CO — i O CJ ^ j-i - — . U CD CO - ^ ,0 CD rd B -H .3 cO cd a c u > CD ^ - — * o o ex TJ CJ -J X 2 CO >• ex o C h 1! T3 •H O I XI cO 4-1 1—1 rH CO 1/1 nJ u C T3 -^\. s b O CD X cd >-j -H N rt O 4-1 >^ X) i-) M-i CO iH o H * 4-1 CO CO 3 i- CO CO C 4-1 CO H C H 3 0J CD E CO > S-i E 0) 01 QJ id 0! pi CO 3 u Ul u tri 3 co i — i *■— tr ^ % ^ (ti tx to 3 (0 3 C CO *6 U c -5 ro >> i — i in 3 c c r^- 2 PC in w u 1- O •H O! cu F— 1 to in en pK ta tfl en in her esi LO > 4_, W H ■D + 147 June: The sulphur bacterium Beggiatoa , which is an indicator of reduced bottoms, is not uncommon in the area, especially in depressions with stagnant water, often at the bottom of the littoral slope. At the oil polluted stations, Beggiatoa covered greater than normal areas even on flat bottoms in June 1978. A higher frequency of Mysidae could be observed at the reference station G and around the Asko Laboratory (adjacent area) than at contaminated stations. A peculiar growth of the tips of Fucus vesiculosus was also observed at station B and in lower frequency at Station C. It looked as if young plants had settled on the old degraded plant tips. On the shore at station B, a beach clean-up was carried out in June, causing a light oil film on the surface water. Samples taken on the deep Mytilus bottoms (8 m depth) visibly contained oil. The Cladophora belt was a bit broader in June, probably due to low water during spring. 7.3.2.2 Calculations from collected data November: A summary of biomass and frequency of dominating animal groups in the Cladophora and Fucus belt in November is made in Table 7.3.1. In some vegetation belts at the most contaminated stations, animal species groups, common at the reference station, were absent (Table 7.3.2). The frequency of vagile (and semisessile) forms was very low at stations B, C, D, E and F, probably as a direct response to oil contamination. The mussel Mytilus edulis dominated the biomass in the entire area. At the reference Station G about 20% of the total biomass in the Fucus belt consisted of other species. The vagile forms con- tributed only a few per cent. In the Cladophora belt about 40% of the total animal biomass consisted of vagile forms. At all other stations Mytilus together with Balanus improvisus dominated the fauna biomass totally, particularly at stations B and F. The vagile forms played a minor part. June and comparisons with November: The dominating plant species from three vegetation belts are listed in Table 7.3.3. Only five spe- cies dominated the biomasses in the specific zones. Cladophora glomerata replaced Ceramium tenuicorne in the shallowest zone (C), and Pilayella 148 e ■ — ^111 oo r~- I OC — I vO I i m > S i o> on i i un i I OC CC I o C z eg c XJ a. c CD .13 3 en 3- on -j- in in i r- i o> ~- — —i i — i CC — — — — CO r- o\] i on on o m ui o + + + c to • 00 4J r- /^N C <^ O CO r-l ^w c H CD o E c O 3 g T3 i-J Uh CD -c b. J= c •J IB >M r- o r- o^ c _ V a k- to CD -a x: c £ 3 u .o > to o z -i o C E •H ok) CO U £ i — i o vi "O CD E o 01 f= i-H > -C >. T3 CO to CO OJ S C_) xle-i — I J> I— I M O U CU "3 CO -C -C H >. o O -C o 4J *-> M Cfl CO >H u a. a a* ezes? 6609Z S8Z + skz 68^7+ TSE2 8ZVI2 ZIL21 ZLZZ + 9S<7+ 996C 0ZZ + 9CCZI 9IIT 9620T C2VC6 9<7CT5 zsze 9*76 + S££c! <7cTVT V^Zc: •79Z9 1 3 <*i p4 E o X5 re E-i I*- — i cm t— ( + 1 +1 OC CM — I C — ' CM . ■0 3 re ID 13 U M 3 (J •H u k* tj cu + + + + re c x: o ft. J3 M n ft -H o C •H M E O w * 3 4-> E O C OJ •H « w z; Q re •H l-l re 01 E r— * w OJ tu o T3 u 3 3 Ed I* — CO re cm Ei O E E E CM CM . _ in o *J F— | —4 J .£ + a ft o 1 V] El ■o 3 re re u u \ 3 b Z •- ■- JZ 0J Cj 4-> •O T3 II II + + 152 u a. o -o CD J2 rH S-J CD e CD > o cfl CO 00 H en 00 C •H ^! e a) u o >, s + "H O cO CO CD CO U cd e -a o c 00 CD (3 CL CD •H cfl C/3 T3 /-> • C fn OO til ^rv ON QJ CO r-H oo 3 CO cj 4-) 3 C Pn 0) o cd cj> c Ph w to CM rO 00 ■H p^ 153 to o oo 1-1 ■H cu CI, CD CU cu > X 1 c 4-1 ■r-t X CU 4-1 ,c •H u s c oc •H r-~ •H > CO O co Z V 3 CO T3 E C o rc M .-^ « qj CU CJ 4-1 43 •H C -a CU CU c u i—l •H u XI CJ -O S i QJ O CO e ;j rt S cd O c -M -H cd c o CO •H U CO ex 6 u 60 •H 5-i CD <3J rH X •H -U 00 S-i CO 3 > 4-1 155 The percentage of abundance and biomass of dominating and vagile species is shown in Tables 7.3.5a and b. Gammarus spp . and Iaera spp . had a very low share of the abundance in November. In many vegetation belts they were totally absent, for example at station B. In June, Iaera was still missing at that station, except in the Ceramium + Mytilus belt, where a single individual was found. In June, gammarids occurred in high abundance at all stations (Table 7.3.3 and 7.3.5a). Most of the gammarids were juveniles. Only 8% 'of the samples contained adult individuals at Station B. Correspond- ing frequencies for stations D and G were 75% and 80%, respectively. 7.3.2.3 Comparison with data from oil analysis of Mytilus edulis The oil content of Mytilus edulis at different times of the year is given in section 11.3.1 and Tables 11.1 and 11.2. The analyses showed a low level of oil before the contamination (10/27/77), and a drastic increase after the spill. The oil content then decreased during the year as shown by successive analyses. The high mortality/absence of all vagile forms mentioned above at the contaminated stations in Novem- ber corresponds well with the high oil content analyzed in Mytilus . The return of vagile forms indicates a healthier environment. This is also indicated by the decreasing oil content analyzed in Mytilus . 7.3.3 Discussion The vegetation, animal diversity and biomass for each vegetation belt usually showed higher values at the reference Station G, especially in November but also in June (Tables 7.3.1, 7.3.2, and 7.3.3). Station G may be considered to be a typical station for the area. No previous investigations in the surrounding area indicate that it would be extreme in any way (Haage, 1975, 1976; Jansson, 1974; Jansson and Kautsky, 1977, Kautsky, 1974; Wallentinus, 1976). The parallel work of Notini also corroborates this (section 7.2). No effect of the oil spill on the benthic macrovegetation was observed. The same observations were made in the Baltic by Notini (1978) and Ravanko (1972). They investigated accidents which happened at the same season, before the spring growth. 156 c o •H 4-1 T) (0 ClJ e W ■H m B o ro ■H i-i -n c c o ■H -> C) in ■n c u ro ffl 0J N) 61! •H ffl 4-J 4-1 O c 0J i—l .C 0) 3 OS a. c^ lo \D ■—I CM lT> ^H + LD ^H snaeuiuieo I I o I CO O r-< CO C\] CT\ IT) r-l EJ9BJ Bsqaopi 157 CM i— i I C^ I I — I I o o — < -T «* I m-i £ ^^ ro o o 4-1 r~- ^-^ W snxjaXw snaeuiiuBO FJ9BI en oo o o o o ^h cni in co ld m eaqqopi 158 The peculiar growth of Fucus observed at station B was probably an effect of the low water which exposed the Fucus tips to the air, causing the protective mucilage layer to dry out. This layer may prevent the Fucus plants from being damaged by contamination. The Fucus plants is rather resistant to light and heavy fuel oil exposure, as shown in outdoor laboratory tests (Ganning and Billing, 1974; Notini, 1978). Results from station B indicate a doubling of the Fucus biomass. This can probably be explained by the heterogeneity of the Fucus belt at that station. A very small change in the sampling site could change the results drastically. It is expected that the Fucus belt would not increase as much as the results indicate. The increase of Fucus biomass (40%) at station D ( Fucus belt) might be explained in the same way. The increase of animal biomass in the Fucus belt at station B in June was caused mainly by increased numbers of Mytilus . These may be the same individuals which were removed and transported from the shal- lower situated Cladophora belt due to narcosis from oil contamination or through mechanical stress during the cleanup operations on the shore. It could also be due to the larger Fucus biomass sampled. Some crustaceans, which have a hydrophobic wax layer on their cuticle, are very sensitive to oil contact (Notini and Hagstrom, 1974). This may explain the drastic decrease of the crustaceans (vagile forms) at the oil-contaminated stations in November. The rapid recruitment of vagile forms from adjacent areas in the Baltic has also been observed by Pelkonen and Tulkki (1972) and Notini (1978). This is also indicated by the high proportion of young in- dividuals at the contaminated station. The lack of chironomid larvae at station B (Table 7.3.3) may be an effect from the oil spill as these organisms have proved to be sensitive to oil contamination (Bengtsson and Berggren, 1972). However, the sampling was too small to allow any definite conclusions. The results of the oil analysis from the reference station G are confusing. The high oil content in Mytilus is not correlated with absence of vagile forms as at the visibly contaminated stations. 159 This study has indicated a rapid recovery of the phytal system. The oil spill occurred at the beginning of a season of low activity in the phytal zone. The low temperature and ice cover from January to March minimize plant and animal metabolism and growth. Many individuals usually die during winter due to senescence. During these 3-4 months of low activity, before the spring growth started, parts of the remaining oil in the littoral were probably washed out and the toxic fractions diluted. Thus the oil spill did not affect the fauna and flora of this system as much as if the spill had occurred in the beginning of the growth season, in April to June. 7.3.4 References Bengtsson, L. and H. Berggren. 1972. The bottom fauna in an oil-contaminated lake, AMBIO 1: 141-144. Dybern, B.I., H. Ackefors and R. Elmgren. (eds.). 1976. Recommendations on Methods for Marine Biological Studies in the Baltic sea. Baltic Marine Biologists Publ. 1:1-98. Ganning B. and U. Billing. 1974. Effects on community metabolism of oil and chemically dispersed oil on Baltic Bladder Wrack ( Fucus vesiculosus ) . In: Beyon, L.R. & E.B. Cowell (eds.). Ecological aspects of the toxicity testing of oils and dispersants. Applied Science Publishers, Ltd., London. 53-61. Haage, P. 1975. Quantitative Investigations of the Baltic Fucus Belt Macrofauna. 2. Quantitative Seasonal Fluctuations, Contrib. Asko Lab. Univ. Stockholm. 9:1-88. Haage, P. 1976. Quantitative Investigations of the Baltic Fucus Belt Macrofauna. 3. Seasonal Variation in Biomass, Reproduction and Population Dynamics of the Dominant Taxa, Contrib. Asko Lab. Univ. Stockholm. 10:1-84. Jansson, A.M. 1974. Community Structure, Modelling and Simulation of the Cladophora Ecosystem in the Baltic Sea. Contrib. Asko Lab. Univ. Stockholm. 57l-130. Jansson, A.M. and N. Kautsky. 1977. Quantitative Survey of Hard Bottom Communities in a Baltic Archipelago. In Keegen, B.F., P. O'Ceidigh and P.J.S. Boaden (eds.), Biology of Benthic Organisms, Pergamon Press, New York. 359-366. 160 Kautsky, N. 1974. Quantitative Investigations of the Red Algal Belt in the Asko Area, Northern Baltic Proper, Contrib. Asko Lab. Univ. Stockholm. 3:1-29. Notini , M. 1978. Longterm Effects of an Oil Spill on Fucus Macrofauna in a Small Baltic Bay, J. Fish. Res. Board Can. 35 (5) : 745-753 . Notini, M. and A. Hagstrom. 1974. Effects of oils on Baltic littoral community, as studied in an outdoor model test system. In: Marine Pollution Monitoring (Petroleum) Symposium and Workshop. Gaithersburg, Maryland, NBS Spec. Publ. 409, 251-254. Pelkonen, K. and P. Tulkki. 1972. The Palva Oil Tanker Disaster in the Finnish S.W. Archipelago. III. The Littoral Fauna of the Oil Polluted Area. Aqua Fenn . 1972:129-139. Ravanko , 0. 1972. The Palva Oil Tanker Disaster in the Finnish S.W. Archipelago. V. The Littoral and Aquatic Flora of the Polluted Area, Aqua Fenn., 1972:142-144. Wallentinus, I. 1976. Environmental Influences on Benthic Macrovegetation in the Trosa-Asko Area, Northern Baltic Proper. I. Hydrographical and Chemical Parameters, and the Macrophytic communities, Contrib. Asko Lab. Univ. Stockholm. 15:1-138. 7.4 In situ respiration of three littoral communities near the Tsesis oil spill (Bjorn Guterstom) 7.4.1 Introduction In order to determine if the oil had affected the community meta- bolism in situ measurements in plastic bags were made on three typical littoral communities of the northern Baltic: 1. the Fucus vesiculosus community 2. the Mytilus edulis community 3. the "shallow soft bottoms" with Ma coma balthica and Hydrobia spp. as the dominating macrofauna components. 161 7.4.2 .Method s Part of the communities were enclosed in plastic bags (0 = 0.5 m, volume 30-70 litres, Fig. 7.4.1, details see Guterstom, 1977) using SCUBA diving. The oxygen consumption was measured with a YSI oxygen electrode on several occasions over a 24 hr. period. Experiments were run at the most contaminated stations and at a similarly exposed, "unpolluted" reference station. Further data from these stations are presented in sections 7.2 and 7.3. Due to failure of the Mytilus experiment at the reference locality, these results were compared with results from laboratory experiments. 7.4.3 Results and discussion The results from the three communities investigated show a typical picture with relatively low respiration at this time of the year (Novem- ber, Table 7.4.1). Only the Fucus communities at two oil polluted localities showed higher respiration than similar unpolluted localities (Table 7.4.1). An increased respiration due to exposure to oil at increasing concentrations was found in outdoor experiments with F. vesiculosus from the northern Baltic (Ganning and Billing, 1974). The shallow soft bottoms showed the same respiration at both localities. The same was found with Mytilus from oil polluted localities compared with mussels from unpolluted areas. At the oil polluted localities oil could be seen in the plastic bags after the incubation periods as it had floated up as small drops under the plexiglass covers of the plastic bags in all experiments. In June 1978 oil was still leaching out of the investigated sediments. Due to the low biological activity during the winter and the few replicates in each experiment, no significant difference in respiration could be found. As shown in Chapter 11 (Tables 11.1-11.4) and discussed in section 7.2.4, the "unpolluted" reference station was later found also to have been contaminated by the oil, even though not by visible slicks. Any effects which might influence the community respiration through changes in animal populations would therefore be expected to be found during the following summer and autumn. Notini (1978) found lower macrofauna populations at oil polluted Fucus communities of the northern Baltic compared to unpolluted communities. 162 50 cm Fig. 7.4.1 Plastic bag 1: buoy 2; plexiglass lid 3; stopper 4; upper ring 5; rubber band 6; plastic 7; lower ring (open) . 163 Table 7.4.1 In situ respiration of three oil polluted littoral communi- ties near the Tsesis oil spill compared with unpolluted communities in the same area. Reference station underlined Locality (see Fig . 5.1.1) Date Temp. Respiration ± i Biomass (°C) (mg 2 g dr.wt " h" ) (g dr.wt) Fucus vesiculo- sus (1.5 m depth] D 1-2.11.77 8.0 D 2-3.11.77 8.0 G 2-3.11.77 8.0 B 14-15.11.77 6.8 B 14-15.11.77 6.8 G 14-15.11.77 6.8 0.29 0.38 0.16 0.07 0.17 0.04 79 Fucus 79 ? t 80 it 43 ii 27 it 31 n Mytilus edulis (2 m depth) 1) 1-2.11.77 8.0 1) 2-3.11.77 8.0 B 9.11.77 7.8 Laboratory 17.11.77 7.7 0.10 0.06 0.09 0.10 ± 0.0 237 Mytilus incl. shell 72 176 n = 10 ind. (recalculated from Sec. 10.2) "Shallow so £t mg 2 2 m _i h Ind . m •2 Macrofauna 2 (dr.wt . g m ) bottoms" (1 m depth) D 16-17, .11. .77 6.4 18.4 4669 234 D 16-17, ,11. .77 6.4 24.5 14616 52 G 16-17, .11, ,77 5.5 20.9 7917 106 164 As oil is still present and is leaching out of the bottoms in- vestigated here, the respiration of these littoral communities is pro- bably influenced to some extent. 7.4.4 References Ganning B. and U. Billing. 1974. Effects on community metabolism of oil and chemically dispersed oil on Baltic bladder wrack Fucus vesiculosus. In: Beyon, L.R. , and E.B. Cowell (eds.). Ecological aspects of the toxicity testing of oils and dispersants. Appl. Sci. Publ. Ltd. England. 53-61. Guterstam, B. 1977. An in situ study of the primary production and the metabolism of a Baltic Fucus vesiculosus L. community. In: Keegan, B.F., P. O'Ceidigh and P.J.S. Boaden (eds.). Biology of benthic organisms, Pergamon Press, Oxford, 311-319. Notini, M. , 1978. Long-term effects of an oil spill on Fucus macrofauna in a small Baltic bay. J. Fish. Res. Board Can. 35:745-753. 165 8. Impact of Oil on the Supralittoral Zone CHAPTER 8: IMPACT OF OIL ON THE SUPRALITTORAL ZONE 8.1 Damage to shore vegetation (Anders Lindhe) 8.1.1 Introduction This study was not begun until the middle of June and by then all the affected coastline including all the fine-grain sediment shores suitable for field study had been cleaned up. This was also true of the bay near Lisb'kalv which was supposed to be left undone. Accordingly, traces of damage by oil upon the vegetation were mostly masked by the mechanical effect of the cleanup. Any sites for studying the long-term effect of oil upon vegetation could, for the same reason, not be found. Since the cleanup on land was mostly mechanical, the effect of chemicals could not be studied. A rough estimate of the acute effects of oil on plants could, however, be gained by studying some non-cleaned areas in the Stockholm archipelago affected by oil from the ship Oktavius . The rest of the work was devoted to studies of the cleanup measures from a biological point of view. 8.1.2 Methods The field work consisted of visits to different oil-affected and cleaned-up areas during June, July and August. In early summer, sites in the Stockholm Archipelago (at Varmdblandet and some nearby islands), affected by oil from the Oktavius and Michail Kalinin oil spills, were studied. The Oktavius oil had been on the shore for about as long a time as the Tsesis oil, though it was of a slightly different type. The Michail Kalinin oil, thin diesel oil, had only been on the shore for a couple of weeks when the studies were made. Some fine-grain sand shores on Torb and Svardsb, heavily affected by Tsesis oil, were also visited at this time. These sites were revisited at the end of summer when the entire coastline between Oren on southern Torb and Hbviksholmarna on Svardsb were investigated on foot. The firm, "Sanerings Konsult", is thanked for valuable assistance with transport and information. 169 8.1.3 Results: Acute effects on vegetation In general the plants seemed to be astonishingly little affected by growing on or in close contact with heaps of oil on the shores. This impression is characteristic of all the areas and different types of oil studied. At places with a dense layer of oil, however, the plants showed frequent signs of abnormality or injury. It is, of course, impossible in any one instance to distinguish with certainty between effects caused by oil or, for example, by parasites or local variation in water supply. However, the following observations have been made several times and are, with fair reliability, oil-dependent. Filipendual ulmaria and Lysimachia vulgaris were very dwarfish when growing on soil heavily impregnated with oil. Seedlings of these grow- ing through a thick layer of oil also showed strong malformations. Deformities were also observed in Valriana officinalis and Tussilago farfara . Discoloration, reminiscent of water shortage, was noted in Festuca rubra , Agrostis stolonifera , Atriplix lati folia and Equisetum arvense when the plants were on or close to oil on the shore. Some specimens seemed less sensitive and C i r s i urn arvense was found on several occasions growing on thick heaps of oil without visible effects. 8.1.4 Some comments on the clean-up methods In the Toro-Svardso area there are several different types of shore. Rocks, stone, gravel and sand, in places exposed to wave action, and fine-grain sediment shores with reeds in more sheltered bays. Rock dominated areas: The only vegetation are lichens and micro- scopic algae. No attempts were made to estimate how these were affected by oil. On these kinds of shores the oil, which soils the rocks with a more or less thick covering, is removed by spraying with water at high pressure, scrubbing with oil-dissolving chemicals or steam treatment, which makes the oil easy-flowing. All methods, however, only transfer the oil to the water. It is hardly possible to take care of and destroy all the oil on these shores and, instead take all possible measures to prevent the oil from reaching them. 170 Stones: These shores are less steep than the rocky ones and, for this reason, can hold more oil. The vegetation consists of sparse stands of plants rooted in the finer sediment between stones. The greater part of the oil can be collected by suction pump and spades after which a more careful cleanup takes place as described above for rocks. Stones, too contaminated for cleaning, can also be removed. One of the problems with this kind of shore is that some oil always per- colates to the gravel between the stones and is out of reach of cleaning, in time, this oil is likely to bleed and pollute the water below. Clean-up operations are difficult and very time consuming and so, if possible, the oil should be prevented from reaching these shores also. Gravel and sand: These shores are rare in the area. Stable parts often carry a complete covering of vegetation. If oil affects this, it is harvested and removed. Filthy sand and gravel are also dug up and carried away. A particular example of a sandy beach is Reveln near Oren on southern Toro. The entire point is made up of sand with gravel on the northern shore and fine sediment on the western side. The area is protected on account of the rich bird life in spring and autumn. The ground on the point is very sensitive to wear. Unfortunately, heavy vehicles have been used in connection with the cleanup operations and deep tracks, which will be slow to heal, have resulted. Also, the traffic across the point has not been properly channeled, which might have lessened the damage. It would have been best to arrange transport by boat. However, the cleaning of the beach proper is satisfactory and the re-visit in August showed that the shore vegetation was recovering quite nicely. Fine grain sediment: These shores are found in sheltered bays, and if ungrazed, are dominated by common reed. Being very level they can hold much oil. During the cleanup after the Tsesis accident, the oil was actively steered into such bays where it was collected by suction pump, and the oil saturated reeds were harvested and removed. Any remaining oil was raked together and destroyed. Several places like these were visited in June and re-visited in August. The impression was that the vegetation was very little damaged. The reeds recovering after 171 harvesting seemed completely normal and on re-visiting it was not always easy to distinguish oil affected and non-affected sites. Even the smaller species seemed to have recovered - for instance Myosotis palustuis and (at Varmdo) the tiny annual Montia fontana . Thus clean-up operations on these shores seem to have worked very well. The oil was taken away from the area to be destroyed instead of being washed back into the sea and the cleaned areas seemed to recover fairly soon. 8.1.5 Concluding remarks No very significant oil-induced vegetation damage could be demon- strated in this study. One cannot however eliminate the possibility that some effects will show only after a considerable period has elapsed. What will happen to the oil that has been left in the soil despite the cleanup? How is it broken down and into what substances? Are any of these more toxic than the oil itself? Are some substances so slowly accumulated by roots and rhizomes that injuries will not show until much later? These and other important questions can only be answered by further studies including experimental field work. Cleaning operations must aim towards removal and destruction of the oil. Pollution of rocky and stony shores must for this reason be mini- mized by channeling the oil towards sheltered bays at an early stage when this is at all possible. This will also prove to be economically beneficial since cleaning of rocks and stones is very time consuming. Absorption of the oil using bark splinters and sowing of grass, which have been tried in some places, seem very doubtful methods from a bio- logical point of view since the oil will remain with unknown conse- quences. The results will at best give rise to unnatural systems with species poorly adapted to the gradients of shore salinity. 172 2 Effects on the supralittoral fauna (Maria Foberg) 2.1 Introduction The importance of having knowledge on the effects of an oil spill, not only in the water but also on the adjacent areas above the water- line, is quite obvious since few investigations have been carried out for this purpose, especially in the Baltic Sea. In order to obtain comparable samples the wrack bed was chosen as a sampling area as it was a biotope which occurred at several of the stations. 8.2.2 Materials and methods Three supralittoral stations situated close to stations B, C and D and one close to the reference station G (Fig. 7.2.1) were investigated during the first two weeks of August 1978, over nine months after the spill. On two occasions - August 2 and August 8 - six pit-fall traps containing water and detergent (used for decreasing the surface tension) were placed in the wrack bed at each station for 24 hours. At the same time two quantitative samples were randomly taken at each station with a frame (20 x 20 cm) so that 10-15 mm of the upper part of the ground material was included in the sample - making a total of 12 traps (10 at station G since two were lost) and 4 quantitative samples from each station . The samples were sorted (the quantitative first by hand and then with a Tullgren apparatus, as described in Backlund, 1945), determined to taxonomic group (family), in some cases to species, and counted. In those cases where a large number of spring-tails (Collembola) were found it was not possible to determine every individual. Instead a subsample was examined and the systematic groups recorded. A comparison between number of individuals and number of systematic groups at the different stations was made (Tables 8.2.1, 8.2.2). The number of individuals in the quantitative samples was recalcu- lated to 100 g wrack dry weight. All newly hatched isopods, which were presumably caught while still clinging to their mother, were left out of the calculations, but the numbers are given in Table 8.2.3 (and later in Table 8.2.6). 173 Table 8.2.1 Occurrence of systematic groups and number of individuals in the pit-fall traps at the four stations. Figures after ± are standard error of mean. Av = rage No Number Average No of Total No of Stn of ind/trap of traps syst grp/trap syst grp B 20±6 12 8±2 28 D 30±5 12 9±1 29 C 39±8 12 10±1 30 G 57±9 10 12±2 33 174 Table 8.2.2 Percentual distribution of the different groups of animals found in pit-fall traps at the four stations The column "other" includes Orthoptera, Thysanoptera (Insecta) and Myriapods. Stn B Stn D Stn C Stn G (Norr (Lind- (Tistel- (FifSng) Skotskar) holmen) holmen) % % 7 /o % Insecta : Collembola 2.1 24.2 4.1 3.5 Hemiptera 4.5 0.4 8.5 Hetereptera 0.2 0.4 Coleoptera, adults 5.9 4.2 6.6 3.9 larvae 0.7 3.9 0.7 Hymenoptera : Parasitie 1.3 0.8 0.4 2.3 Formicidae 24.4 18.5 0.9 5.7 Apidae 0.4 Diptera : Brachycera 21.0 6.2 17.4 21.2 Nematocera 3.8 4.2 3.2 1.4 Crustacea : Porcellio scaber 23.1 22.8 6.7 1.9 Orchestia gammarellus 0.2 10.4 Arachnoidea : Araneae 6.7 12.6 54.7 39.2 Arcarina 2.5 Opiliones 0.6 Other 2.1 0.6 1.1 1.2 175 Table 8.2.3 Total number of individuals (x) caught in the traps at the four stations and percent (7 ) frequency of occurrence. Stn B Stn D Stn C St a G er of Traps 12 12 1. ? 12 X 7 /o X % X ' % X % CTA Collembola fam. Entomobryidae 1 8 65 100 9 20 ti Poduridae 3 17 8 17 9 33 ii Isotomidae 13 17 10 8 11 50 ti Sminthuridae 1 8 ysanoptera fam. Thripsidae 1 8 Hemiptera Homoptera fam. Aphididae Cicadellidae 16 42 2 17 42 6 70 40 eteroptera fam. ii Lygaeidae , Scolopostethus sp. 3 17 2 17 Coleoptera fam. Carabidae 9 50 ii , Pterostichus niger 1 8 14 67 10 50 ii , Harpalus sp. 1 8 ii , Metabletus sp . 1 8 ti Hydrophilidae , Cercyon sp . 11 50 3 20 ii Staphylinidae 1 8 3 25 it Silphidae, Thanatophilus sp . 9 25 2 17 »i Ptiliidae 2 17 7 40 ii Scarabaeidae , Geotrupes sp. 1 8 M Elateridae 1 8 11 Nitidulidae 3 25 II Coccinellidae 1 8 1 8 It Curculionidae 1 8 2 10 larvae (mostly Pterostichus niger) 16 42 17 67 4 40 ymenoptera fam. Ichneumonidae 1 8 4 40 Braconidae, Syncrasis fucicola 1 8 2 17 3 30 , Pemphredon lugubris 1 8 Proctotrupoidae , Basalys sp . 1 10 " , Tricopria sp . 2 10 , Codus sp. 1 10 , Trimorus sp. 1 10 Diapriidae, Diapria sp . 2 17 1 10 Pteromalidae , Urolepis maritima 1 8 Formicidae, Myrmiea rubra 3 17 1 8 6 30 , Formica fusca 4 25 2 8 8 50 " , F. rufa 5 25 4 33 1 10 , F. ruginodes 1 8 " " , Lasius niger 45 67 58 92 4 33 17 50 " , Camponotus herculeanum 1 8 Apidae, Bombus sp . 1 8 176 Table 8.2.3 (cont'd) Diptera Brachycera fam. Dolichopodidae Phoridae 7o X % " Pipunculidae " Syrphidae " Sciomyzidae " Ephydridae " , Ephydra macellaria alandica " Agromyzidae " Chloropidae " Tachinidae " Muscidae , Coenosia mollicula 4 17 5 25 45 83 80 80 8 42 3 1 25 8 2 1 17 8 2 1 20 10 8 33 3 17 4 20 1 8 22 75 8 42 18 83 24 60 1 8 1 8 2 8 1 10 1 8 1 10 2 17 2 1 17 8 1 1 8 8 7 20 2 17 1 8 1 8 2 17 1 10 1 8 1 8 4 25 1 8 1 8 11 42 9 5 42 8 3 20 4 33 2 17 1 8 3 30 3 17 2 17 4 17 2 20 1 8 O & 17 3 20 55 100 81 92 31 83 11 70 81 8 " , Dexiopsis lacteipennis " , Lispe tentaculata Anthomyiidae Scatophagidae " , Scatophaga litorea 1 Nematocera fam. Corethridae " Chironomidae , adults " " , larvae " Mycetophilidae " Cecidomyidae unidentified Orthoptera fam. Acrididae CRUSTACEA Isopoda Porcellio scaber, adults " " , newly hatched Amphipoda Orchestia gamma re llus 1 8 59 70 ARACHNOIDEA Aranae fam. Linyphiidae 8 33 26 67 171 100 205 100 Lycosidae 8 42 18 50 84 92 17 60 " Gnaphosidae, Micaria sp . 18 Acarina fam. Ixidae, Ixodes ricinus 6 17 Opiliones 2 17 1 10 MYRIAPODA Chilopoda fam. Lithobidae, Lithobius sp. 1 8 Diplopoda fam. Iulidae 2 17 CASTROPODA Stylommatophora fam. Arionidae 2 20 MAMMALIA Insectivora fam. Soricidae, Sorex sp. 1 10 177 The orders found were: Insecta: Thysanoptera , Collembola, Orthop- tera, Hemiptera, Coleoptera, Hymenoptera and Diptera. Crustacea: Isopoda (Porcellio scaber ) and Amphipoda (Orchestia gamma re 1 lus ) . Arachnoidea: Araneae, Arcarina and Opiliones. Myriapoda : Diplopoda and Chilopoda. Oligochaeta: Plesiopora and Gastropoda: Stylommato- phora . The genera Pachydrilus and Enchytraeus (Oligochaeta) have not been taken into consideration and were left out in all calculations. This was due to the difficulty of obtaining reliable abundance estimates for these highly contagiously distributed animals. The estimated numbers are given in Table 8.2.4 and percent distribution in Table 8.2.5. 8.2.3 Description of the stations At station B and D there was a small belt of totally dry wrack only a couple of centimeters deep and 15 cm wide, which, according to Backlund (1945), could be characterized as a wrack string. The wrack consisted almost exclusively of Fucus vesiculosus with some Cladophora. At these two stations a few oil slicks could be seen on some stones but not in the wrack. Station B, one of the most heavily contaminated stations, was subject to cleanup in the middle of June 1978, during which time parts of the ground were dug up and turned over. At station C, also heavily contaminated, a rather large slick of oil was found in one of the quantitative samples. Here, the wrack accumulation was much larger, about 10-15 cm deep and 50 cm wide, con- sisting of decaying wet wrack. At station G, the reference station, the wrack bed was similar in size to that at C and consisted of decaying wet wrack. At this station there was a bed of reed adjacent to the sampling area, and as a result the wrack contained much material from Phragmites australiensis . 8.2.4 Results 8.2.4.1 Pitfall traps At station G there were more individuals and systematic groups in each trap as well as a higher total number of systematic groups found 178 Table 8.2.4 Occurrence of systematic groups and number of individuals in the quantitative samples at the four stations. Figures after + are standard error of mean. Average No of ind/lOOg Number Average No Total No Estimated No of wrack (dry of of syst grp/ of syst Enchytr/Pachydr per sample 200 250 2000 4000 Stn weight) Samples sample grp B 85+23 4 7+1 11 D 57+31 4 6+3 13 C 175+93 4 8+2 13 G 162+25 4 10+2 13 179 Table 8.2.5 Percent distribution of the different groups of animals found in the quantitative samples at the four stations. Stn B Stn D Stn C Stn G (Norr (Lind- (Tistel- (FifSng) Skb'tskar ) holmen) holmen) % 7 /o 7 /o % Insecta : Collembola Hemiptera Coleoptera, adults , larvae Hymenoptera : parasitic Formicidae Diptera : Brachycera, adults , larvae Nematocera Crustacea : Porcellio s caber Orchestia gammarellus Arachnoidea : Araneae Acarina 2. 2 1. .5 5. 1 0. ,7 0. . i 2, ,2 0. ,7 40. .9 2. .2 33. .6 12.8 2.4 0.8 0.8 7.2 3.2 45.6 2.4 23.2 30. .9 0. ,4 2. ,4 0, .4 0, .2 5. .7 0, .4 14. .1 10 .8 34 .8 42.9 0.2 6.7 0.5 0.2 0.2 1.8 2.0 1.8 6.3 37.2 Myriapeda 10.2 1.6 180 compared to the other stations. For more detailed data see Table 8.2.1 and Fig. 8.2.1. The percentual distribution of the different groups of animals (Table 8.2.2, Fig. 8.2.3) shows some similarity between station B and D and between station C and G, respectively. At station B and D, Formicidae (ants) and the species Porcellio scaber (woodlouse) dominated -- at B together with Diptera (flies and gnats) and at D with Collembola (spring- tails), while at station C and G, Araneae (spiders) were strikingly dominant, followed by Diptera. The gamma rid Orch estia gammarellus was frequent only at station G where it comprised 10.5% (59 individuals) of the total number of individuals found, while at station C the corres- ponding figure was 0.2% (1 individual). No gammarid could be found at the other stations. At station G there also seemed to be more parasitic hymenopteras and cicadas (Homoptera) than at the other stations. The little wood- louse Porcellio scaber appeared in small numbers at station C and G compared to stations B and D. In one of the traps at station G a shrew (Sorex sp.) was found in company with some fleas. These latter were not included in the calcu- lations . 8.2.4.2 Quantitative samples The average number of individuals/lOOg wrack dry weight at stations G and C was two to three times higher than at stations B and D (Table 8.2.6) The total number of systematic groups found was 11 at station B and 13 at each of the other stations (Table 8.2.4, Fig. 8.2.2). The percentual distribution of the systematic groups (Table 8.2.5 and Fig. 8.2.4) in this case also showed a similarity between stations B and D and stations C and G, respectively. Porcellio scaber was clearly dominant at station B and D, followed by Acarina (ticks), while Araneae seemed to be less frequent at these stations. Myriapoda (millepedes) made up 10.2% of the total number of individuals found at station B while at C and G this group could not be found either in the quanti- tative samples or in the traps. At station C and G Acarina and 181 mean No of ind / trap 60- 50- 40- 30- 20- 10- total No of families 1 -40 -30 -20 -10 mean No of ind/IOOg wrack 250- 200- 150- 100- 50- total No of families -16 -12 -8 -4 B B Fig. 0.2.1 Average number of individuals (Istandard error of mean) in the pit-fall traps (o) and total number of systematic groups (a) at the four stations . Fig. 8.2.2 Average number of individuals/ lOOg wrack, dry weight (o) (istandard error of mean) and total number of systematic groups (A) in the quantitative samples at the four stations. 182 Table 8.2.6 Total number of individuals (x) occurring in the four quanti' tative samples at the four stations and percent occurrence frequency (%) . St n B Stn D St n C St n G X I X % X 7 10 X % INSECTA Collembola fam. Entomobryidae 3 25 9 75 (i Poduridae 4 50 5 50 399 75 " Isotomidae 3 25 153 25' Hemiptera Homoptera fam. Aphididae 2 25 1 25 2 50 Heteropotera unid entif ied 1 25 Coleoptera fam . Hydrophilidae 8 100 7 75 it Staphylinidae 2 50 2 25 4 50 5 75 1! Ptiliidae 3 Nitidulidae 2 larvae 1 25 50 25 1 25 2 25 50 75 Hymenoptera fam. Ichneumonidae 1 25 n Braconidae 1 25 ii Proctotrupoidae 1 25 it Formicidae, Lasius niger 1 25 2 50 Diptera Brachycera " Agromyzidae 1 25 1 25 larvae 3 50 9 75 29 100 17 100 Nematocera " it Chironomidae Cecidomyidae 1 25 3 1 50 25 1 25 CRUSTACEA Isopoda Porcellio scaber, adults 56 100 57 75 72 100 19 100 it ", newly hatched 79 50 44 50 4 25 10 25 Amphipoda Orchestia gammarellus, adults 17 75 K , newly hatched 15 25 ARACHNOIDEA Aranae fam. Linyphiidae 3 75 3 25 52 100 59 100 1! Lycosidae 3 50 Acarina fam. Ixidae, Ixodes ricinus 46 100 29 75 178 100 346 100 MYRIAPODA Diplopoda fam. Iulidae 14 75 2 25 OLIGOCHAETA Plesiopora Pachydrilus sp. and Enchytraeus sp. (estimated numbers) 200 75 250 100 2000 100 4000 100 183 Insecta : I Collembola ^d Hemiptera llli Goleoptera Hymenoptera Crustacea : K3 Porcellio 1_±J Orchestia Arachnoidea : >=] Diptera Araneae(Ar), Acarina □ Other 2.3 Percentual distribution of the different groups of animals in the pit-fall traps at the four stations. The area of the circles is correlated to average number of individuals per trap. B Fig. 8.2.4 Percentual distribution of the different groups of animals in the quantitative samples at the four stations. The area of the circles is correlated to average number of individual s/lOOg wrack, dry weight. 184 Collembola were the dominant groups. Station G showed a low number of Porcellio scaber , only 2.0% while the gammarid Orchestia gammarellus appeared at this station exclusively. 8.2.5 Discussion The result of the pit-fall trap study shows that the total number of systematic groups was somewhat higher at the reference station and that the average number of individuals caught was 1.5 to 3 times higher than at the other stations. The quantitative samples however, show that the total number of systematic groups was similar for all stations except station B, which had a slightly lower figure. The average number of individuals per 100 g wrack was clearly higher at station G compared to stations B and D, but a little lower than at station C. This could be due to the wrack at G being too wet for some animals, which would explain the low number of Porcellio at this station (Backlund, 1945). The fact that the gammarid Orchestia was frequent only at G and absent at the contaminated stations might be a result of the oil spill, but it could also be a consequence of the humidity of the wrack at G (Backlund, 1945). This does not explain the almost complete absence of this species at C, where the wrack was as wet as at G. The differences in humidity of the wrack are clearly shown by the abundance of the group Plesiopora (Oligochaeta) which demands humidity for thriving (Backlund 1945). This 2 2 group is much more frequent at G and C (about 27,000/m and 13,000/m 2 2 resp., estimated values) than at B and D (about 1 , 100/m and 1,300/m resp.). The largest part of this group was Pachydrilus sp . and the rest Enchytraeus sp . Pachydri lus is a typical animal of the seashore and abundant in wrack beds, being the main food source for many animals living there, e.g., spiders of the genus Erigone sp . and larvae of the coleopteran Cercyon sp. (Backlund, 1945). Most of the flies (Brachycera) in the quantitative samples were larvae because the adults are able to fly and thus escaped capture. The great number of flies in the traps could be due to their being attracted by the smell of the detergent, as most of them are not true inhabitants of the seashore (Chinery, 1976). Species which truly belong to this 185 biotope, for instance Lispe tentaculata (family Muscidae), Dexiopsis lacteipennis (family Muscidae, according to Hugo Andersson, Lund, pers. comm., not known in this county before), Scatophaga litorea (family Scatophagidae) , and Ephydra macellaria alandica (family Ephydridae) contribute only 10% of the total number of flies while Dolichopodidae and Agromyzidae made up the major parts (49% and 26%, resp.)- Some of the flies are parasites on other insects, e.g., individuals of Phoridae (Chinery, 1976) and Tachinidae (Lindroth, 1967). One individual of the family Pipunculidae was found at Station G. Members of this group are known as parasites of cicadas (Homoptera) (Lindroth, 1967) and were found only at this station. Some of the parasitic hymenopteras are fly parasites, e.g., Tricopria sp. and Trimorus sp . These were also found at Station G where the greatest number of flies occurred. The only hymenoptera species which truly belongs to the shore is Urolepis maritima which appeared at Station G. Most of the Ichneumonidae would seem to have come from the woods (Cederholm, pers. comm.). Perhaps these were also attracted by the detergent. According to Backlund (1945) nearly every wrack bed contains parasitic hymenopteras living on dipterous larvae. The high number of ants (Formicidae) at Stations B and D could be explained by ant hills in the wood beyond the shores. A greater part of the spiders consisted of the species Oedothorax retusus , (family Linyphiidae) , small web-spinning spiders which are found especially in littoral vegetation and on the surface layer of wrack (Backlund, 1945). The genus Erigone (family Linyphiidae) was also found. One third of all the spiders was made up of Lycosidae, hunting spiders which probably live to a greater extent on flies (Backlund, 1945). The genus Arctosa , Trochosa and Pardosa were also found. All of the ticks found belonged to the species Ixodes ricinus , which live in decaying organic material and lay their eggs in the ground vegetation (Ursing, 1971). At Station G there was more vegetation around the wrack bed, which could have influenced the number of aphides (Aphididae) which occurred here in larger numbers than at any other station. One species (Hyalopterus pruni) has for instance Phragmites as a summer host (Lindroth, 1967). It is possible that the individuals found belonged to this species, but no further examination was made. 186 The dominant group of Coleoptera (beetles) were carabides of the species Pterostichus niger followed by Hydrophilidae of the genus Cercyon, common in decaying wrack on shores (Chinery, 1976). These occurred only at Stations C and G, where the humidity was high, since they feed on even wetter wrack than Orchestia (Backlund, 1945). The family Ptilidae was also frequently found, some species of which are very common in wrack beds (Backlund, 1945). Staphylinides , of which many species live on shores and feed on Diptera larvae, also occurred in the samples as well as Nitidulidae, which lives on decaying material (Chinery, 1976). In all calculations the newly hatched individuals of Porcellio and Orchestia were omitted. At station B there were 81 newly hatched Porcellio (among 19 adults) still clinging to the mother adult in one trap, and in one quantitative sample there were 75 newly hatched individuals of the same species among 6 adults at the same station. If these had been taken into account, the average number of individuals in each trap would have increased to 27 (instead of 20), just a little less than at the other stations and the average number of individuals per 100 g wrack would then become not 85 but 158. This would make the figure comparable to that at station G. In fact, for the small amount of material examined the differences in number of individuals and systematic groups between the stations are not so large that they could not be explained by the natural variation caused by the environment. The wrack bed is a rela- tively variable environment, strongly affected by storms. Nevertheless it was of a similar nature throughout the whole area investigated during the sampling period, so that a comparison could be made between the stations . The humidity is a factor of great importance to the animals laving in the wrack itself (^Backlund, 1945). Water makes the wrack much softer and more accessible and therefore easier to eat. Orch estia for example cannot eat hard and dry wrack (Backlund, 1945). Another important factor influencing the wrack fauna are the interactions over the "bound- aries" to adjacent biotopes (Backlund, 1945). At station G for example the bed of reed comprises another biotope containing other types of animals supplying immigrants to the wrack fauna. Thus the surrounding biotopes influence wrack beds both quantitatively and qualitatively. 187 The low number of individuals at station B could be a result of the oil spill but it is more likely a result of the ground being turned over during the clean-up procedure, a measure which destroyed the wrack more or less completely. Since there were no storms during this period, when Fucus might have blown ashore, the remaining wrack bed became rather diminished, and this in turn influenced the animal life. 8.2.6 Conclusions Unfortunately no similar investigation had been made in this area either before or directly after the accident, which makes it difficult to draw firm conclusions. The present investigation indicates, however, that the oil spill from Tsesis has not had any great and long lasting negative effects on the fauna of the wrack belts of the shore. It seemed to recover rather quickly, due both to the short generation time and the vagility of the animals. Another important factor was the time of the accident. It took place in the late autumn when many animals had already left the upper, most affected part of the ground in preparation for hibernation. Had it happened in the early spring it is possible that the results would have been different and shown longer lasting effect . 8.2.7 Acknowledgements The following persons deserve thanks for great help with exam- ination of the animals: Dr. Carl-Cedric Coulianos (Brachycera and Coleoptera), Dr. Torbjorn Kronestedt (Araneae) , Dr. Lennart Cederholm (Hymenoptera) and Dr. Hugo Andersson (Brachycera). 8.2.8 References Backlund, H.O. 1945. Wrack fauna of Sweden and Finland, Ecology and Chorology, Lund. Acad, diss., Opuscula entomologica Suppl . 5: 1-236. Chinery, M. 1976. Nordeuropas lnsekter, En Bestamningsbok. Handledning for bestamning av samtliga insektf amiljer i nordeuropa. Albert Bonniers forlag AB, Stockholm. 1-371. (In Swedish). 188 Lindroth, CM. 1967. Entomologi , Biologi 7, Almqvist and Wiksell, Stockholm. 1-236. (In Swedish). Lindroth, C.H. 1942. V§ra skalbaggar och hur man kanner igen dem, Del I-III, Albert Bonniers Forlag AB, Stockholm. 1-222. (In Swedish) . Ursing, B. 1971. Ryggradslosa djur, P. A. Norstedt och Sb'ner Forlag, Stockholm. 1-369. (In Swedish). 189 CHAPTER 9: IMPACT OF OIL ON LOCAL FISH FAUNA (Sture Nellbring, Sture Hansson, Gunnar Aneer and Lars Westin) 9.1 Introduction From earlier echo sounding surveys (Aneer et al., 1978) it is known that pelagic fish (herring, Clupea harengus membras , L. and sprat, Sprattus sprattus L.) are very abundant in the spill area during late autumn and winter. It is also known that herring spawn in the archi- pelago in the spring. In view of the not inconsiderable local com- mercial fishery, the following questions were of interest after the spill: Were the pelagic fish still present in the area and were the fish contaminated by the oil? Did the oil affect the spawning grounds and the hatching results (c.f. Linden, 1976)? 9.2 Material and methods To investigate the occurrence of pelagic fish in the polluted area, echo sounding surveys were carried out on four occasions (11 November 1977, 15 December 1977, 11 January 1978, and 12 April 1978). The survey routes are illustrated in Fig. 9.1. Herring were caught with gill-nets or by trawling. Organs (gills, stomachs) and whole fish from the reference and affected areas were deep-frozen for oil analysis. In order to investigate spawning grounds, 100 randomly chosen non-polluted stations and 20 polluted stations were visited in June 1978 (about seven months after the spill) by SCUBA divers to see if any spawning had taken place. At each station, the divers investigated the bottom from the shoreline down to a depth of about 10 m. At stations in the affected area where spawning had taken place and at four non-polluted reference stations, egg samples were taken for hatching in the laboratory. At each station two samples were taken by cutting and removing sections of algae with attached roe. In the labor- atory the two samples were mixed and ten sub-samples of about 50-100 193 TORO 2 km Fig. 9.1 Echosounding survey routes ^P = Tsesis — " = Survey routes (scale 1:50,000) 194 eggs each were placed in hatching chambers with a volume of about 100 ml (Fig. 9.2). These were kept in the hatching chamber system for 7-12 days. Two or three days after the appearance of hatched larvae in a chamber, the contents were removed and preserved in 4% formaldehyde for later examination. The temperature in the chamber system was between 10 and 14 C, corresponding to the in situ temperature for that time of year. The oxygen content of the water in the hatching chambers was almost saturated. 9.3 Results The echo sounding surveys did not indicate any decrease or disappear- ance of pelagic fish within the oil spill area. Furthermore, chemical analysis of herring organs and whole fish showed no indication of oil contamination (Boehm, pers. comm.). The frequency of spawning grounds was lower in the contaminated area than in the reference area (20%, n =20 and 45%, n =100). (Fig. 9.3 2 2 and 9.4). A x test was carried out (x - 3.3, d.f. = 1) and the result was found to be significant at the 7% level. There was no significant difference in the number of malformed larvae. Only two larvae with enlarged areas anterior to the yolk sac were found in two samples from the polluted area (picture, Linden, 1976). The hatching results are presented in Table 9.1. These are divided into two columns, representing two different counting methods. The first column (observed hatching) gives the percentage of hatched larvae. The second column (theoretical hatching) gives the percentage of hatched larvae plus the percentage of eggs with live nearly hatched larvae. The hatching of eggs from one reference station, Jutskar, was unsuccessful. This was probably due to a severe fungal infection, the causes of which are unknown. As can be seen in Table 9.1, the average hatching success was lower in the oil polluted areas. The hatching success of eggs from Tistelholmen, the most affected area, was extremely poor, about 20% (theoretical hatching) . 195 c/> a) 42 ■u T3 (U CO CO e P > P 0.006, observed hatching, z = -2.96, 0.0032 > P > 0.0026). It is worth noting, however, that one of the polluted stations, Tistelholmen, gave highly variable results, with significant differences between the three sampling occasions (rank-sum test for several samples (Dixon and Massey 1969:345), H observed = 97.9, d.f. = 2, H theoretical 94.6, d.f. = 2, p << 0.01 in both cases). 9.4 Discussion The analysis of herring did not indicate any contamination by oil (Boehm, pers. coram. )• However it would be of great interest to trace the path of the oil through the food chain by carrying out oil analyses on flounder ( Pleuronectus f lesus ) . These fish feed mainly upon Ma coma balthica and the blue mussel ( Mytilus edulis ) both of which have been shown to contain considerable amounts of oil. The low frequency of spawning by herring in the affected area may indicate effects of oil pollution. It may, however, also be due to the differences that undoubtedly exist between the polluted and reference areas as regards exposure and sediment type distribution on the fairly shallow bottoms, where herring spawn was found. On comparison, it is clear that the hatching of herring eggs was less successful in samples from the oil affected area. Significant differences also between samples from within the polluted area may, however, indicate that factors other than oil pollution may have in- fluenced hatching rate. One such factor, repeatedly observed in the hatching experiments, was fungal infection of the roe. From previous studies it is known that the presence of benthic crustaceans decreases the amount of fungal growth on fish roe (Oseid, 1977). In the spring after the oil spill, hardly any adult Gammarids were present in the polluted area. It is therefore possible that their absence led to increased fungal attack on fish roe. 200 9.5 References Aneer, G. , A. Lindquist and L. Westin. 1978. Winter concentrations of Baltic herring ( Clupea harengus var . membra s L.) Contrib. Asko Lab., Univ. Stockholm. No. 21. 1-16. Dixon and Massey. 1969. Introduction to statistical analysis. McGraw- Hill. Kogakusha, Tokyo. 1-638. Linden, 0. 1976. The influence of crude oil and mixtures of crude oil/ dispersants on the ontogenic development of the Baltic herring Clupea harengus membra s . Ambio 5: 136-140. Oseid, D.M. 1977. Control of fungus growth on fish eggs by Asellus militaris and Gammarus pseudolimnaeus . Trans. Am. Fish. Soc. 106: 192-195. 201 10. Laboratory Studies Carried Out in Connection with the Spill CHAPTER 10: LABORATORY STUDIES CARRIED OUT IN CONNECTION WITH THE SPILL 10.1 Introduction In connection with the Tsesis oil spill some laboratory studies were carried out on animals brought to the Asko laboratory from the affected area. The reason for these investigations was to provide information on possible sublethal effects of the oil on organisms in the area . Three types of investigations were carried out: 10.2 Respiratory measurements on the mussel Mytilus edulis . 10.3 Measurements of byssus formation by Mytilus edulis . 10.4 Burrowing behavior in the clam Ma coma balthica . 10.2 Respiration measurements on the mussel Mytilus edulis (Sture Hansson) The rate of energy turnover in organisms is known to be affected by pollutants. One indication of oil-induced effects on the energy turnover would be changes in the respiration rate. However, measurements of the respiration rate alone cannot be used to determine how energy turnover is affected as a number of other components in the budget might also be affected, e.g., growth rate or food ingestion. 10.2.1 Materials and methods Mussels were collected by SCUBA diving from the upper 3 m at oil polluted littoral stations and from the same depth range at an unpolluted reference station by means of a dredge from the shore. The mussels and 2 liters of the surrounding water were collected in ethylene containers which were sealed underwater to avoid contamination by the surface oil slicks . In the laboratory the mussels were allowed to acclimatize for 2-4 hours before the start of the experiment. The temperature during the experiments did not differ by more than two degrees from that in situ , and field temperature and salinity varied negligibly between stations. 205 Animals from unpolluted localities were incubated in clean water, while those from polluted localities were incubated in water collected together with the mussels. The measurements were made on specimens with a mean shellfree dry-weight of 31 mg (S.D. = 8 mg , shell length = 17-22 mm). Single specimens were incubated for 50 to 150 minutes in syringes with a volume of 10 ml and a diameter of about 15 mm. In order to study the acute effects of high oil concentrations, fresh mussels were incubated in a laboratory prepared oil-in-water mixture. This mixture was prepared from 10 ml of Tsesis oil and 900 ml sea water which was mixed using a magnetic stirrer for 18 hours at 20 C. The water phase served as an incubation medium and was used both undi- luted and diluted (1:10) and was added fresh at the start of the incu- bation. Mussels incubated in uncontaminated Baltic sea water were used as a reference. The oxygen concentration was measured in control syringes with both polluted and unpolluted water which had been incubated as long as the syringes with animals present. These values were then subtracted from measurements with animals present, before the oxygen consumption of the mussels was calculated. Concentrations were determined with a Radiometer oxygen electrode (D 616) connected to a Radiometer PHM 716 with a PO -module, PHA 930. 10.2.2 Results Data from the respiration measurements are presented in Tables 10.2.1 and 10.2.2 and illustrated in Fig. 10.2.1. The respiration rates of M. edulis from different localities, and those from the experiments with different incubation water, were examined with Bartlett's test (Bailey 1976:189) and with analysis of variance (Dixon and Massey, 1969:156-162) in order to determine whether variances or means were significantly different. When variances in the measure- ments from all stations (polluted and unpolluted) were compared no 2 differences were found (x = 8.17, df = 6). However, in testing the means by analysis of variance, a p-value between 0.005 and 0.001 (F/£. co\ ~ 4-098) W as obtained rejecting the hypothesis that all the samples came from populations with the same respiration rates. 206 X X a 03 00 cu 2 CM O 00 "2. P a 4-1 •p P CO O >> cu a a p P cu cu OJ a o 03 u u E o p 5-( a CO s-l •H •H 03 cu cu cu U P P a T3 4_> ft CU CO O o ■H (t ft U X ■H 4-1 u m CO o 03 P a Ch ^1 r-4 03 o a cw 5-1 u o .a o p :o ■ H •P P o J£4 Oh s^s P ■H 5-4 fX4 CO CO 03 3 CU < CU LO P X PS o> •H ID E cu ft a P X CO 3 oo > 4-J •H a 4-1 B cu fl 00 •H O a T3 P o CO X) TD -o •H p co 3 CU a P Xi -0 CO 4-J cu 03 00 CU 03 P •H P CO 5-1 cu CO CU 3 •H P X 3 T— 1 CO CO p CU a 1— I 03 a u >> 03 o X r— 1 p a 5-1 CU ft 1 P rP •H cfl cu u Q a 1— 1 Xi a cu •P 00 CU o 4-1 O •H P ■rH cu OJ 03 4-J 5-1 a B 3 03 o i CO P cu u >^P P X! m P r— 1 a H -a 3 cu ft CO CO u s u •H cu OJ a • cu o ^-( OJ p o H H 3 u u • T3 4-1 cu CN CU 1 XI a « i-H H cu r^ • i— 1 00 cu a OJ >. 4-> X X 00 03 <+H CO CO o ■H P o CO rp 03 a IP CO > r— I r-- lo vO to r-~ r-« r-- o vo r-- r-» \£> -J" lo CN CJ1 CTi 0> CT\ CT\ CT - ! CTN ~4" -Cl" -CT O O O O I— I ,— I r-\ r — r~- r — r~- i — r~- i — r~- r-~ i~-~ r — r — r~- r— ■ TD T3 TD -a ~o T3 CU cu CU CU CU CU 4-1 p 4-1 4-1 4-J 4J a 3 3 3 3 3 "O H i-H rH rH rH rH cu r-4 1— 1 rH rH rH .— < 4-J O o o o o o a P4 ft CP a CP P4 r— 1 rH >, >> >> >■> >. >> O rP rH rH rH rH 1— 1 &4 ■H •H •H 4-1 ■P •H > > > xi > > P 03 03 03 00 03 CO o CU CU CU ■H CU CU 2; 35 X W P? W W a 4-1 CO cu w Q rH o CJ PQ CJ a a a a a a a cu o o o o o o S-l •H •H •H ■p •p •H cu 4-) 4-1 p 4-1 4-1 P lh 03 03 03 03 03 CO CU 4-J 4-J 4-1 4-J 4-> P PBJ CO CO CO CO CO CO 5-1 CU P CO -a cu o Oh O rH rH 03 •H CJ P 5-1 03 -D a £0 a 03 cu T3 CU 4-> 03 X 3 CJ a >p O CU cu CO CO cu p * 03 CO p a o a «h O P ■p 03 P -H 03 > 5-4 CU •H p PP co X> QJ CJ T3 P •H Oh LO CO o> CU C£ CM o 00 s • *- — ' > cu 4-J -3 X 00 CO •H cu 3 3 >> 03 u CU Q S3 CO CN ■p ft rH O E cp 3 rH CU T3 Eh CU CU i—4 X 3 03 cu p CO Q X CO H S-l CU P CO a o CO X 3 CJ a o 1 — 1 r- r~ vD LO co lO CM O0 oo lo lo o o -Cl- co CO O0 co oo 00 CM CM lO co CM OO OO r-- r~ TD S-l CU cu 4-1 4-J 3 CO i— 1 3 O p CU -o ft 4-J CU CO 4-J a 3 3 CO T— 1 CU f — s -o S-l r-\ r—t ^ CU cu o CJ p p ft ON 3 03 CU <—>, 3 >. 5-1 II r-\ n^ 3 o a 4-1 P H Oh CO X X CU CU 00 •H p i— 1 rH •H E 03 •H CJ rP 3 O 207 500-- 100- Respi ration ( /jg 2 /g-h ) Temp. 9 °C o T3 CD (0 +-> +J 00 CU o c QJ CU <+- CU B Temp. 8 U C -a CD 4-> rO _Q 13 U 5- CU n3 l (T3 4-> i— C •• O Ol (_> •i- S- O CU +-> • • n3 1 — - s 3 TD Z3 CU "O c C 4-> CU -a cu +-> to c 03 +-> O C_) Fig. 10.2.1 Respiration rates (means with 95% confidence intervals) of Mytilus edulis from clean and oil-polluted localities(A) and of animals incubated in clean and artificially oil-contaminated water (B) 208 From Fig. 10.2.1 it can be seen that animals from three of the stations (D, E, F) showed lower respiration rates than animals from other stations. This result can, however, be explained by the fact that M. edulis from stations D, E and F were incubated in water with lower oxygen contents than those from the reference station (Table 10.2.1). It is known that low oxygen concentrations depress the respiration of Myt ilus edulis (Bayne, Thompson and Widdows, 1976). If the remaining stations (B, C, G) and the reference station (all had about the same oxygen content in the incubation water. Table 10.2.1 ) were compared no differences could be found between variances (x = 1.93, df = 3) or between means (^/o.oq\ ~ 0.85). The results from the experiment with Myt i lus incubated in water with different oil-concentrations (Table 10.2.2) also failed to show statistically significant differences between variances (x = 1.44, df = 2) or means (F. . = 1.26). 10.2.3 Co nclusions and discussion According to this study, oil contamination of the water did not significantly change the respiration of Myt i lus edulis . However, due to the large variability in respiration, (mean values of approximately 650 pg /g/h and standard deviations about 150 pg o /g/h) , changes would have had to be large, in order to be detected. To detect, with a probability of 90%, a difference in respiration rate of about 20%, the sample size ought to be some 30 specimens (Dixon and Massey 1969, Table A-12a). 10.2.4 References Bailey, N.T.J. 1976. Statistical Methods in Biology. 10th ed. Hodder and Stoughton, London. 1-198. Bayne, B.L., R.J. Thompson and J. Widdows. 1976. Physiology I. In: Bayne, B.L. (ed.). Marine mussels. Their ecology and physiology. Cambridge University Press. Cambridge. 121-206. Dixon, W.J. and F.J. Massey. 1969. Introduction to Statistical Analysis. 3rd ed., McGraw-Hill, Kogakusha, Tokyo, 1-638. 209 10.3 Measurements of byssus formation by the mussel Mytilus edulis (Olle Linden and Maria Foberg) Methods: Oil-affected mussels were collected in the littoral zone off Toro about one week after the grounding of the Tsesis . At this time large quantities of oil were floating on the surface. The mussels were collected by a diver at 4-5 m depth and brought to the laboratory in plastic bags. Care was taken to avoid contamination by the surface oil film. Reference mussels were collected close to the Asko laboratory in a similar but unpolluted biotope. In the laboratory the mussels were allowed to acclimatize for a couple of hours. After this period mussels of two length-classes (ju- veniles: up to 10 mm, adults: over 10 mm) were placed on petri-dishes which were immersed in unpolluted sea water. Twenty to 25 animals of each length-class from each locality were exposed to a continuous flow of sea water in ~10 litre plastic jars. The number of mussels byssally attached to the dish after 3, 4, 5 and 6 hours was noted. Results and discussion : Fig. 10.3.1 shows the percentage of mussels byssally attached during the course of the experiment. Indi- viduals from the impacted area showed decreased tendency to attach to the substrate. This tendency is more pronounced among adult individuals compared to juveniles. Reduction or absence of byssus thread production in mussels ( Mytilus edulis ) or related species under the influence of oil, has previously been observed under laboratory conditions (Smith 1968, Swedmark et al. 1973, Eisler 1973, Linden 1977). It is obvious that affected byssal activity under natural conditions must be considered a serious stress syndrome. Mussels incapable of byssal thread production will be unable to remain in their natural habitat as threads are used as mooring lines. They will consequently be washed away and their chances of reattachment at a suitable place are probably rather small. The experiments, although performed under laboratory conditions, indicate that the mussels in the impacted area were subjected to such effects. 210 Q LU I- < O o en xi cfl QJ U X. ■u en 3 •s. en >. rO E U o IW o 4-1 ^ o a o ~ c cu en 4-J 4-1 rH ,—> 3 en T3 •H cd H 3 II ~ a) -a ctj Cfl a T3 H c •H cfl ~J >. en 5 en 3 3 •r- 1 II ^H O pq ■I - ) i — i • ro • o 1 — 1 . 60 ■H Pk sav3yHi snssAa 9Niwd0d sivwinv 211 • References Eisler, R. 1973. Latent effects of Iranian crude oil and a chemical oil dispersant on red sea molluscs. Israel Journ. Zool . 22: 97-105. Linden, 0. 1977. Sublethal effects of oil on molluscs species from the Baltic Sea. Wat. Air Soil Pollut. 8: 550-558. Smith, E. (ed.) 1968. Torrey Canyon, pollution and marine life. Cambridge Univ. Press. 1-196. Swedmark, M., A. Cranmo, and S. Kollberg. 1973. Effects of oil dis- persants and oil emulsions on marine animals. Water Res., 7(11): 1649-1672. 212 10.4 Burrowing behavior in the clam Macoma balthica (Olle Linden) Methods: Oil -affected clams were collected at 15-20 m depth off Toro about one week after the grounding of the ship. At this time large quantities of oil were floating on the surface. The clams were collected by a diver and brought to the laboratory in closed plastic jars. Refer- ence animals were collected off the Asko laboratory in an unpolluted area of similiar depth. After acclimatization in the laboratory for 2-3 hours, about 20 clams of each of three length-classes (3-4 mm, 4-9 mm, 10-15 nun) from each locality were spread out over the surface of a 5 cm thick, uncon- taminated sediment layer in several glass jars. The glass jars were immersed in basins containing unpolluted sea water. The number of clams that had buried themselves completely was counted at several time intervals . Results and discussions. The rates of burrowing of the bivalve are shown in Figure 10.4.1. On the whole the control animals showed the highest burrowing rate, and all individuals had buried completely within 60 min. The animals from the polluted area showed a clearly decreased burrowing rate. Affected burrowing behavior among clams under the influence of oil pollution has previously been reported by Shaw et al. (1976), Linden (1977) and Taylor and Karinen (1977). As this experiment was performed under laboratory conditions, the ecological consequences of the observed effects are still unproven. However, under natural conditions behavioral disturbances such as im- paired burrowing will most probably, in the long run, be unfavorable for this organism. Clams unable to burrow normally may have their food finding affected as they are deposit feeders. Furthermore, the impor- tance of burrowing behavior in the predator-prey relationship is obvious. The results reported here have demonstrated that the clam Macoma balthica during the Tsesis oil spill may have been subjected to such sublethal effects. 213 cc Z3 o I— I > < X LU PQ CD O en CQ CD O X CO CO _l < % 100 ANIMALS FROM A CONTROL AREA 50- 7. 100 • - 3-4 MM p = 8 " . = 10-15 " ANIMALS FROM A CONTAMINATED LOCALITY 60 105 mN Fig. 10.4.1 Macoma balthica burrowing behaviour 214 References Linden, 0. 1977. Sublethal effects of oil on molluscs species from the Baltic Sea. Wat. Air Soil Pollut. 8:550-558. Shaw, D.G., A.J. Paul, L.M. Cheek and H.M. Feder. 1976. Ma coma balthica an indicator of oil pollution. Mar. Pollut. Bull. 7: 29-31. Taylor, T.L. and Karinen, J.F. 1977. Response of the clam Ma coma balthica exposed to Prudhoe Bay crude oil as unmixed oil, water- soluble fraction, and oil-contaminated sediment in the laboratory. In: Wolfe, D.A. (ed.): Fate and effects of petroleum hydrocarbons in marine organisms and ecosystems. Pergamon Press, Oxford. 229-237. !15 11. The Analytical Chemistry of Mytilu s eduHs, Macoma balthica Sediment Trap ana Surface Sediment Samples CHAPTER 11: THE ANA1YTICAL CHEMISTRY OF MYTILUS EPULIS, MACOMA BALTHICA SEDIMENT TRAP AND SURFACE SEDIMENT SAMPLES" (Paul D. Boehm, Judith Barak, David Fiest and Adria Elskus) 11.1 Introduction It has become increasingly apparent in recent years that the tem- poral impacts of spilled oil in the. marine environment become more prolonged when the fate of petroleum hydrocarbons includes transport into the sediment. Recent studies by Teal et al. (1978), Keizer et al. (1978), and Mayo et al. (1978) indicate that aromatic and aliphatic hydrocarbons from spilled petroleum persist in the sedimentary environ- ment for substantial periods of time (years). While degradative pro- cesses, both chemical and microbial, act to alter the composition of the oil in sediment, the toxicants and carcinogens may persist in the sed- iment depending on factors such as sediment grain size, wave energy (Owens, 1978), and oxidation state of the sediment (Anderson et al. , 1978). There have been few field studies directly pertaining to the sedimentation of oil owing to natural processes. At least three pro- cesses can lead to the transport of petroleum from a positively buoyant state in the water column to negatively buoyant state reaching the sediment. First, some oils by virtue of an initial density close to that of water (e.g., Bunker C) can weather at sea, lose volatile or soluble components, and sink (Conomos, 1975). Such behavior was apparently observed after a Bunker C spill in cold water off the coast of Greenland (Mattson and Grose, 1978). Secondly, oil can adsorb to living particulate matter or detrital particles and sink due to sedi- mentation. This process is dependent on the availability of sediment particles as well as the nature of the particulate matter (National Academy of Sciences, 1975; Meyers and Quinn, 1973; Poirier and Thiel, 1941). Another route of transport to the benthos is by ingestion of oil by zooplankters followed by fecal pellet transport (Conover, 1971). A further indirect process is deposition after landfall by seaward trans- port of beach and intertidal sediment with associated petroleum. " Work carried out by Energy Resources Co., Inc. under Contract No. MO-A01-78-4178 to OCSEAP. 219 Shaw et al. (1976) have suggested that Ma coma balthiea represents an ideal organism to monitor exposure of the benthos to pollutant input due to its deposit feeding behavior, while others have suggested that filter feeders (e.g., mussels, Mytilus species) are good indicators of oil in the aqueous environment due to their manner of processing large volumes of water. Due to their widespread distribution in coastal waters, mussels ( Mytilus sp.) have been closely studied in laboratory and field experi- ments for their responses to hydrocarbon pollutant exposures. Studies on the hydrocarbon chemical content of species of mussels have been performed relating to (1) background or chronic input levels in Mytilus galloprovincialis (Fossato and Siviero, 1974), M. edulis (Ehrhardt and Heinemann, 1975; Rudling, 1976), M. californianus and M. edulis (DiSalvo et al., 1975); (2) laboratory hydrocarbon uptake and depuration studies (Lee et al., 1972; Fossato and Canzonier, 1976; Kanter, 1974; Clark and Finley, 1975); (3) field transplantation (uptake and depuration) studies (DiSalvo et al., 1975; Fossato, 1975; (4) oil spills in the field (Grahl- Nielsen et al., 1978; Clark et al., 1978, among others). Mytilus sp. is a suspension feeder which processes large volumes of water to obtain food and in doing so is exposed to hydrocarbon and other pollutant compounds dispersed in the water or adsorbed to particulates. Mytilus itself is a food source for many animals, but in the Baltic it is not utilized by man. Furthermore its sedentary nature within the littoral community makes Mytilus populations both excellent markers of pollutant exposure within this zone as well as indicators of temporal recovery of the littoral zone. This chapter concentrates on the extent of Tsesis oil exposure of the Mytilus edulis population and the long-term (1 year) tissue hydro- carbon burden of populations around the region affected by the Tsesis oil spills. For details on the spill, see section 1.2. Of particular interest were the changes in both the quantitative and qualitative nature of the aliphatic and aromatic hydrocarbon assemblages in the tissues. Macoma balthiea , a prominent resident of the soft-bottom community, was used as an indicator of pollutant input to the benthos. 220 A year-long study of these animals together with supportive measurements from surface sediment and sediment trap samples will be the means to determine the chemical fate of oil from this spill and propose a des- criptive model for the behavior of the oil after the spill. 11.2 Methods 11.2.1 Sampling Sampling was carried out by the Swedish scientists from the Asko Laboratory, University of Stockholm, and the Swedish Water and Air Pollution Research Institute (IVL) as part of their ecological impact study. Sampling of Mytilus were obtained periodically from eight of the stations in the study area indicated in Figure 11.1 (B, C, D, E, F, G, I, J). Baseline chemical information was obtained by sampling at stations C, D, and G prior to the oil's landfall. A reference station J adjacent to the biological laboratory on Asko Island was sampled in October of 1978 after it became apparent that the original reference site G was indeed impacted by the oil. Ma coma samples were obtained in grab or dredge samples taken at nine soft-bottom stations shown in Fig. 11.1 (C, D, 2, 5, 6, 7, 8, 15, 20) . Station 15 was chosen as the reference station not believed to be impacted by the Tsesis spill. At each station at least 5 grams wet weight of Mytilus and Macoma tissue were obtained. Individual specimens measured from 1 to 3 cm and each sample for analysis consisted of 10 to 60 individuals. Samples were frozen after collection and transported using dry ice for preser- vation. Sediment traps (see section 4.2.3.5 for details) were deployed at three stations in the area (II, IV, V). Sediment samples were collected as described in section 6.2.1. The collected material was stored frozen prior to analysis. 11.2.2 Sample Analysis Once in the laboratory the shells of the frozen samples were rinsed with distilled solvent. The specimens were then shucked, tissues com- bined and weighed and added to 50 ml Teflon capped centrifuge tubes. 221 Figure 11.1 Map of Study Area and _S t a 1 1 q n_ Locations KEY Roman numerals (I, II, III , IV, V, VI) = pelagic sampling stations Arabic numerals (2 , 5, 8, 15, 20, 21 ) = benthic sampling stations Capital letters (B - J, exc . H) = littoral sampling stations or benthic stations in the proximity of a littoral station. 222 58049.7'N — I 17043.8'E 223 The digestion, extraction, and fractionation schemes were similar to those developed by Warner (1976) except that the digestion was performed using a 0.5 N KOH/distilled water/distilled methanol system heated in a boiling water bath for 4 hours to achieve complete digestion and hence release of hydrocarbons from the cellular matrix. Internal standards were added prior to digestion and carried through the entire procedure (f, = androstane; f„ = hexaethyl benzene). Standards were also added to sediment and sediment trap samples (see below). The digestate was extracted three times with distilled hexane in the centrifuge tube. The extracts were combined, concentrated to 0.5 ml, weighed on a Cahn electro- balance, and fractionated on an alumina over silica gel column (Boehm, 1978). Two fractions corresponding to the aliphatic or f, (hexane eluate) and the aromatic/olef inic or f„ (methylene chloride eluate) hydrocarbons were obtained for gas chromatographic analysis. Sediment samples were extracted using the method of Boehm and Quinn (1978) and fractionated as stated above. Sediment trap samples ("^1 gram) were extracted in closed centrifuge tubes with a methanol-hexane solvent mix in a boiling water bath for 4 hours. The solvent was obtained, concentrated, and fractionated as stated. Chromatographic fractions were concentrated to 50 |jl and a 1 |Jl subsample was injected into a Hewlett Packard Model 5840A gas chromato- graph equipped with a flame ionization detector. Samples were chromato- graphed in the splitless injection mode on a 15-m (0.25-mm i.d.) SE-30 glass capillary column (J and W Scientific; ^50,000 theoretical plates). The column oven was temperature programmed from 60 C to 275 C at 3 C per minute. The injection port and detector temperatures were 250 C and 300 C, respectively. Peak areas were obtained using a digital integration option which resets the baseline at every valley. Thus peak areas above the unresolved complex mixture (UCM) or hump were digitized. Peak area and retention time information for the sample peaks and the internal standard was transmitted using an HP 18861 digital interface to a PDP 10 computer which computed retention index and quantitative data according to programs developed at ERCO. 224 The unresolved complex mixture (UCM) was quantified using planimetry according to published procedures of Boehm and Quinn (1978) and Farrington and Quinn (1973), and relating its area to that of the internal standard through an integration unit to planimeter unit conversion factor. Combined glass capillary gas chromatography/mass spectrometry was performed on a Hewlett-Packard 5985 GC/MS/computer system for peak identifications and quantifications, using selected ion monitoring. GC/MS was used primarily as a tool for investigating aromatic (f~) hydrocarbon fraction contents of a selected set of tissue and sediment trap samples. Quantitative GC/MS was performed on selected aromatic hydrocarbon samples, GC/MS response factors were computed by examining the instru- mental response of a given amount of aromatic standard relative to that for the internal standard (hexaethyl benzene). Response factors for components for which no authentic standards were available (e.g. C~ phenanthrene , C„ and C~ fluorenes, C., and C~ dibenzothiophenes) were computed by extrapolation from the factors for parent and monomethylated compounds. Total ion currents for parent (M ) peaks were obtained for each component of interest, relative response factors applied, and converted to concentration units by comparison to the internal standard amount. 11.3 Results 11.3.1 Mytilus edulis All of the stations in the littoral zone were sampled from the time of the spill event through early May 1978. In addition, two of the stations C and D, along the eastern shoreline, and one near the spill off Fifong Island (G) were sampled prior to landfall of the oil (base- line samples) and through October of 1978. Results of these analyses are presented in Tables 11.1 through 11.4. 11.3.1.1 Aliphatic Hydrocarbons The data in Table 11.1 document a very rapid uptake of Tsesis oil by the littoral bivalve Mytilus. Aliphatic hydrocarbon concentrations 225 TABLE 11.1 ALIPHATIC HYDROCARBON DATA ON MYTILUS EDULIS SAMPLES Fj ALIPHATICS ( yq/g ) ] DATE DRY WEIGHT PRIS PHY PRIS/ PHY ALK/ SAMPLI RESOLVED UCM TOTAL ISO (13-18) (C) 10-27-77 1.54 1.9 71.0 72.9 nd 0.08 - - 11-14-77 0.64 7,538.1 25,708.3 33,246.4 286.4 429.6 0.67 2.67 12-14-77 0.32 1,138.8 13,087.8 14,226.6 102.5 133.4 0.77 0.06 5-2-78 1.55 92.1 1,785.4 1,877.5 3.7 5.3 0.70 0.09 6-20-78 0.62 25.1 799.8 824.9 1.7 1.3 1.34 0.45 8-23-78 2.10 16.0 958.0 974.0 0.4 - - - 10-30-78 2.24 3.3 958.0 161.3 0.2 0.3 0.67 0.30 (D) 10-27-77 1.39 2.5 20.3 22.8 nd 0.2 - - 11-30-77 2.88 1,388.8 6,258.1 7,646.9 49.3 83.9 0.59 1.79 12-14-77 1.17 47.0 922.1 969.1 6.0 6.7 0.90 0.03 5-2-78 1.75 74.0 923.0 997.1 1.2 1.7 0.70 0.26 8-23-78 1.68 6.8 513.0 519.8 - - - - 10-30-78 1.79 2.5 65.5 67.0 - - - - (G) 10-27-77 1.89 2.7 79.7 82.4 0.2 0.05 3.5 - 11-09-77 1.17 706.7 2,002.0 2,708.7 14.2 43.9 0.32 1.6 12-14-77 0.96 86.5 688.8 775.3 5.6 7.0 0.80 0.3 5-02-78 0.39 53.9 1,399.5 1,453.4 4.2 6.3 0.66 0.06 8-23-78 1.71 66.0 687.0 753.0 6.6 10.1 0.65 0.16 10-30-70 2.12 1.9 50.5 52.4 0.1 - - 1.3 (B) 11-09-77 3.16 1,126.6 2,976.1 4,102.8 29.7 59.0 0.50 1.67 12-14-77 1.14 267.2 2,187.2 2,454.4 19.5 29.1 0.67 0.06 5-2-78 1.40 61.9 1,558.6 1,620.5 3.9 6.4 0.61 0.10 (E) 11-09-77 2.36 123.7 653.3 777.0 11.1 15.3 0.73 0.66 12-14-77 1.31 177.9 1,911.4 2,089.3 23.1 29.2 0.79 0.06 5-2-78 1.06 45.5 1,166.3 1,211.8 5.2 7.3 0.70 0.07 (F) 11-9-77 0.90 574.9 1,328.0 1,902.9 21.0 50.5 0.42 0.53 12-14-77 0.89 92.9 1,012.7 1,105.6 9.7 13.5 0.72 0.03 5-2-78 0.86 42.7 1,150.4 1,192.1 3.1 4.9 0.64 0.09 (I) 11-9-77 2.29 338.3 1,6 51.1 1,989.4 21.3 31.9 0.67 0.64 (J) 11-2-78 (Control) 2.81 1.3 12.2 13.5 — — — — TSESIS oil — — — — — — 0.54 7.0 UCM=unresolved complex mixture; PRIS=pristane; PHY=phytane; ALK/IS0= n-alkane-to-isoprenoid ratio over the range n ~ C io io Tsee also Table 11.5). — 1 J-lo 226 TABLE 11.2 AROMATIC HYDROCARBON GROSS PARAMETER CONCENTRATIONS IN MYTILUS EDULIS DATE F 2 (AROMATICS) SAMPLE RESOLVED UCM TOTAL (C) 10-27-77 1.7 14.8 16.5 11-14-77 456.6 16,917.9 17,374.5 12-14-77 297.5 21,275.0 21 ,572.5 5-02-78 18.9 1,394.0 1,412.9 6-20-78 5.6 461.7 467.3 8-23-78 2.8 106.0 108.8 10-30-78 14.0 426.8 440.0 (D) 10-27-77 2.5 46.0 48.5 11-30-77 141.4 3,689.1 3,830.5 12-14-77 11.7 852.9 864.5 5-02-78 7.3 478.0 485.3 8-23-78 2.8 56.7 59.5 / /-i \ 10-30-78 5.8 110.4 116.2 (G) 10-27-77 8.5 210.0 218.5 11-09-77 99.6 6,236.7 6,336.3 12-14-77 2.4 470.3 472.7 5-02-78 12.3 1,366.3 1,378.6 8-23-78 7.8 86.1 93.9 10-30-78 5.0 48.5 53.5 (B) 11-09-77 174.7 4,636.6 4,811.3 12-14-77 47.6 2,057.4 2,105.0 5-02-78 23.4 894.1 917.5 (E) 11-09-77 1.9 858.0 859.0 12-14-77 20.2 2,297.1 2,317.3 5-02-78 44.8 1,418.8 1,463.6 (F) 11-9-77 68.0 3,015.0 3,083.0 12-14-77 7.9 1,620.0 1,627.9 5-02-78 12.2 1,053.1 1,065.3 (I) 11-09-77 9.8 2,400.1 2,409.0 (J) 11-02-78 3.3 12.3 15.6 (Control) 227 TABLE 11. 3 CONCENTRATIONS OF AROMATIC HYDROCARBON COMPOUNDS IN MYTILUS EDULIS AS DETERMINED BY GC/MS NAPHTHA- PHENAN- DIBENZO- LENES THRENES THIOPHENES STATION DATE (yg/g) (yg/g) (yg/g) (F) 11-09-77 32.3 24.5 _a 12-14-77 nd b 3.2 - 5-2-78 nd 1.6 - (D) 11-30-77 23.2 31.7 - 12-14-77 0.1 7.9 - 5-2-78 0.03 0.6 0.5 8-23-78 0.15 0.08 - 10-30-78 0.005 0.17 - (B) 11-09-77 79.2 61.3 34.3 12-14-77 3.4 14.0 14.1 5-2-78 0.3 1.7 1.0 (G) 10-27-77 (Baseline) nd 1.2 0.71 12-14-77 0.82 5.4 1.8 10-30-78 0.23 0.13 - (I) 11-09-77 0.7 16.91 - a Not searched . ^None detected . 228 01 0) c c 01 0) I— 1 rfl CO CX ! 1 i r C 1 a VO in o *r O 1 ^G O •« i c ! ■U -H 0) 1 l (-. 1 o —1 O o o o as 1 X X C i ° 1 j a 4J oj ffl O £ i H CO ^* ^ f C N Ph ca ' vo O »* *< m o m m I c o i Q • • m • • « • ■ • • • « « « « « 1 ^H QJ "H i ~* •a CD «• i a e p» m o « * C o c r« o 41 « « * • CD ■" H C 0) >> 0) C ^3 c. i"> o in <«> •» VO o <-4 in «»> CO o r^ VO CTI o •» r» l»> rH OJ -U «•» CO S-i CU (J o m in vo © o O o o o CD e» o 00 IN o e» 4-1 4-1 -H mm r» •• ui o ci Dh CO 4J V a. o ** *m «-# o o <-« *» a w+ r» CN * 00 ^^ o crj C CO' Ml in U o r» - o o a o •^ o r» M o en —4 © o •r C ai « -C OJ J ^h a c a;o; — « c vo in CO o o "0 c •a c CO in C» t- ■» » 13 N> m >s 0) ,C ^H 42 4-1 >, (X 63[U 1 u u (N O o o o o C* «-4 o OB a rv 0) J3 O ICQ' o CU ■"< ""• 6 4-1 -H w x'si •H JJ J 3 S5N X c a. •o — •o o o .. "3 o o •JO rt »J IN T3 ~s -a C 4-1 •H O J <'CJ u c • c • ■ e c • • • • • • C • • " S-i N QJ 4J C i-t JO J in o a o o m o a IN o o tr« J! < C OJ >• hI^I s a Z in o f-7 o « r» r» r» 0) •« ^3 2 o^'cc' M «•■> ^ « ^3 •e T3 ■e • • • "O • "3 •3 • t— 1 0) "H "vT ej c o» C c C o C a o ^ o c o C C o d C -a ** «4 ^ m rC 0) * 2 u SS a z 4J J-l rH i— 1 M !Q W9 « CO CT> x: x: >. i— 1 >tl2 z •o r» T3 •o O o •a »H o f~ W1 fX o •3 °3 ■o CL 4-1 ^C w cap in (J c i»> C c o o c o o ~* o o — ^ C C c CO C 4J 63 2 ° «^ cO OJ H OJ -H T3 J o cn'cu a CO cs's Z •o ■3 "O ^3 CO o •0 ^3 o VO O o «N ■a CO •a •a T3 >,£ -d m < cs Cd'O e C c C c C • • c c c C £ Oh 4-1 £H o o 63' • S'J« w •3 ■O a o o o T3 ■o o o o ON o <-i a m a o o T3 "3 ■a met hyl ene, etec a: <'<' z c c • t c c C C c II 4-> -C T3 a ajjal o o o o o O o o o 0) P-, >* < a S S 01 X u 2h > ■J < m OS in o i-l vo CO in m —4 1/1 —I m o o m a CO o« IN vO CM o uxi ^ 4-1 a w PI . o a> co ^ •-* r* •» CO mmt en o VO o •V « « H-t w a 5" H » » » » •» ^ » 3 QJ N II EH o 2 2 i *" i <-> » r* n ** "" a •m o>« a r~ «» o « m \o r» *» m a r-i a . hC c ,rf H 01 ■» a >n r~ tn -* -* r» ■*r l^l m o« -* IN m o ' U cO 1— 1 3 *■*• X a 1 v« at ^ «-• c* •• « o CO o VO <=> ^ C >, M s z 1 r» •• « r> -* ** r« "> QJ rC 3 OJ -C 4J cy: * d a oj S a oj g ^ « oa IHH O 2 :> m ♦ p» n oo CO \n ^ o r» so ^r o e» IN CO ! CO >-. II O I Q ■j o CK : 43 43 C ca ~ » IN VO — 1 4-1 4-1 H C J 43 0) pq -H 1 1 1 ! a r» i — r» CO r* r» CO p> f~ CO COT3 ; 11 oj r» I r» 00 r» p» r» p» 1 r» 1 r» r» i CU • I o I CO r- 1 1 1 1 C^ i CO c^ l a r*- ! H pL, r, O i H ' r» co *e r» i o r» » a o •v p> o » r~ I < • in ^m t <*"» CI n ^M n «-l 1 1 <— i 1 e< 1 0*^-1 S-i t a ' i l 1 r« < o i-H M 1 IN 1 Q OJ 1 «■- Csl CO z >, CM 4-1 l o I OJ - * £ 1 /""N 1 s s~\ s-~^ •— v ^-s ^ HP : < O pq 1— 1 •> PQ I CO 1 w ^-- v.^ v^ P-i O II O O -X 229 in living tissue exceed 30 mg/g at the most heavily impacted station C. Animals at the other shoreline stations D and E contain aliphatic hydro- carbon levels of 7.6 mg/g and 4.1 mg/g respectively during the November 1977 sampling. These values are 300 to 500 times the background hydro- carbon levels in tissues from these stations. Although depuration of aliphatic hydrocarbons appears to commence very soon after the initial impact in animals surviving the oil's landfall, it is not until 1 year after the event that tissue hydrocarbon levels appear to reapproach the background levels. However, examination of the detailed hydrocarbon composition of the tissues 1 year after the spill still reveals petro- leum hydrocarbon inputs at some stations. This will be discussed. High resolution gas chromatograms of the tissue hydrocarbon com- positions reveal a rapidly changing suite of hydrocarbons in the mus- sels. Initially, tissues from the most heavily impacted stations appear to have taken up hydrocarbons very similar in composition to the spilled oil (Fig. 11.2a and 11.2b). However, all samples obtained in December, 2 months after the spill, reveal GC patterns indicating substantial alteration of the cargo oil pattern, with n-alkanes being preferentially degraded compared with corresponding branched and isoprenoid compounds (Fig. 11.2c and 11. 2d). At several stations, most notably I, even the November tissue samples already exhibit notable preferential n-alkane degradation throughout the entire boiling range of the oil (approxi- mately n-C, through n-C ) . The GC of Fig. 11.2c illustrates such a case, where the isoprenoid hydrocarbons, having retention indices of 1370, 1460 (farnesane), 1560, 1650, 1710 (pristane), and 1812 (phytane) , become chromatographically prominent due to their greater resistance to microbial degradative processes (Kator, 1973). The rapid alteration in the tissue hydrocarbon composition from an essentially unaltered oil to an n-alkane-depleted, isoprenoid-enriched assemblage is the single most important aspect of the changing hydro- carbon chemistry of Mytilus tissue. This change is expressed in Table 11.1 as the ALK/IS0 ratio, which is simply the ratio of n-alkane to isoprenoid hydrocarbons in the n-C, to n-C, boiling range. In the ij lo spilled cargo oil the n-alkanes predominate (ALK/ISO = 7.0). The most 230 Figure 11.2 Representative Glass Capillary Gas Chroma to grams of Mytilus edulis Aliphatic Hydrocarbons A - Baseline (pre-spill) B - TSESIS cargo C - November (Station B) D - November (Station I) E - December (Station B) F - May 1978 (Station B) G - October 1978 (Station C) CH = cholestane (internal standard); n- c = n-alkane having x carbon atoms; FA = farnesane; PR = pristane; PH = phytane; UCM = unresolved complex mixture; 1370, 1460, 1650 = retention indices of these peaks 231 LZO-H 913- IN SE3-N «3-N EZ3-N CI3-N II3-N 0J3-N — S13-N -^ 813- N -5 il 3-N "_p- 913-N ^7 313- "'-^ -4 232 heavily oiled samples exhibit an ALK/ISO ratio from 1.7 at station B to 2.7 at C, both in early November. Thus, even at the earliest sampling period the petroleum hydrocarbons already exhibit an altered pattern. Hydrocarbon GC profiles from particulate material in the water column in the region obtained from sediment trap deployments (see sec- tion on sediment traps) indicate that the oil sampled in the water during the first week in November also has been dramatically altered (see Fig. 11.6), presumably due to microbial degradation. Even the earliest trap samples obtained exhibit a degraded pattern (ALK/ISO = 1.0) and within 10 days all of the remaining oil dispersed in the water column has been severely altered (ALK/ISO = 0.1). As previously mentioned, the gross hydrocarbon parameters indicate that Mytilus remaining in the region have eliminated most of the initial hydrocarbon burden 1 year after the spill and pre-spill levels are being approached (Table 11.1). However, GC profiles (Fig. 11. 2G) of samples taken from station C in October of 1978 still exhibit significant levels of petroleum hydrocarbons albeit a degraded assemblage consisting mainly of UCM material . 11.3.1.2 Aromatic Hydrocarbons After receiving an initial hydrocarbon tissue burden at levels of 5 mg/g to 20 mg/g at the most heavily oiled stations, levels in Mytilus decrease but remain at levels 100 times the background levels of the gross parameters (Table 11.2) 7 months after the spill. One year after the spill levels appear to approach background at all stations except the most heavily impacted station, C. However, a close examination of the gas chromatographic profiles of representative aromatic hydrocarbon fractions throughout the year reveals that residual petroleum material is still present in the tissues one year after the spill (Fig. 11.3). Therefore the gross hydrocarbon parameters (Table 11.2) can be mis- leading without a consideration of both the GC run and the detailed aromatic hydrocarbon composition. The latter was determined by sub- jecting a set of samples for combined gas chromatographic/mass spec- trometric analysis and quantifying individual aromatic compounds which are otherwise obscured in the GC run due to their low levels relative to the entire suite of aromatic compounds. 233 Figure 11 . 3 Repre s e n tative Glass Capillar y Gas Chromatograms o f Mytilus edulis Aromat i c Hydrocarbons A - Baseline (pre-spill, Station C) B - TSESIS cargo C - November 1977 (Station B) D - December 1977 (Station C) E - May 1978 (Station B) F - October 1978 (Station C) G - October 1978 (Station G) TMB = substituted trimethyl benzene HEB = hexaethyl benzene (internal standard) 234 J J N'3 (NJ 3N31VHXH«fVN 235 As can be seen in Tables 11.3 and 11.4, levels of individual hydro- carbon compounds initially as high a 40 pg/g decrease rapidly to the 3 (Jg/g to 10 (Jg/g level in December and to the 0.1 (Jg/g level in October of 1978. It is interesting to note that one of the baseline or pre- spill samples (station G) contains significant quantities of substituted phenanthrenes as well as the organo-sulphur compounds, the dibenzothio- phenes. These are probably the residual aromatics from previous pollu- tion events. The other baseline sample from station D also exhibits small quantities of substituted aromatics (see Fig. 11.3a) as well. After one year (October 30, 1978) it is apparent that pre-spill levels are being approached, although the detailed aromatic composition does not mirror the pre-spill composition; i.e., one year after the spill mussels still contain substituted naphthalenes whereas the pre-spill samples contained none. As was the case for the aliphatic uptake patterns, the aromatic uptake profiles reveal temporal changes. On an absolute basis all aromatic components appear to have decreased substantially during the first month of post-spill depuration (Tables 11.2 and 11.4). On a comparative basis consider Fig. 11.4. Samples obtained after the first two weeks, post-spill, exhibit a slightly altered aromatic composition vis-a-vis the spilled oil. Soon thereafter the biphenyl and fluorene compounds decrease to non-detectable levels (May 1978) in the tissues and remain so until biphenyl reappears in tissues and, with the naph- thalenes, assumes a prominent comparative importance one year after the spill. It should be noted that Fig. 11.4 presents the data on a com- parative basis with the absolute sum of the total components in Fig. 11.4 decreasing throughout the year from 165.4 (Jg/g to 0-7 Mg/g- During the year there appears to be a greater comparative loss of the naphtha- lenes, fluorene, and biphenyl, the lower boiling compounds, in addition to the parent organo-sulphur compound dibenzothiophene. The substituted dibenzothiophenes and substituted phenanthrenes are less readily depurated on a comparative basis. The aromatic hydrocarbon gas chromatograms also reveal a series of substituted trimethylbenzene (TMB) compounds that are relatively long- 236 Figure 11.4 Comparative Plot of Aromatic Hydrocarbon Composition of TSESIS Oil and Mytilus edulis Normalized to Trimethyl Phenanthrene N, C]N, C2N, C3N = naphthalene (N), methyl N, dimethyl N, trimethyl N BP = biphenyl F = fluorene DBT, CiDBT, C2DBT = dibenzothiophene (DBT), methyl DBT, dimethyl DBT P, C]P, C2P, C3P = phenanthrene (P), methyl P, dimethyl P, trimethyl P 237 — CO o> a. Ol ON rr 1 ob it a.~- a-. ~ £ m r> ID £* » 2 (T» (O J£ l*» CO en en j£ > r» r- — . 00 O a) re en CD Z u> QS- k. _ «■■» **». ** «0 o m > o £So CO CO «• 2 «••••> 3 3 3 3 CO CO -i LU iu *» M *•**"*» f/t fn > WV > >■ > H 1- 2 t- 22 2 f «Q <&ir * »6 * 23i lived in the tissues. The structures of these compounds were examined by GC/MS. A selection search for m/e=133 (Fig. 11.5) shows that the series is present in the Tsesis oil although in relatively small amounts. However, the series emerges rapidly, presumably due to its retention by the organisms or due to difficulty in biotransformation, and persists until 1 year after the spill when analyses fail to reveal significant quantities of TMB compounds. 11.3.2 Sediment Traps The material collected in the sediment traps at Stations II, IV, and V contains large amounts of weathered Tsesis cargo oil. Concentra- tions of Tsesis oil in the traps are presented in Table 11.5 and illu- strate that the rate of deposition of oil through the water column is similar at the upwind station (IV), at the wreck site (II), and downwind in the direction of the movement of the visible slick (V). Thus, it appears that oil is dispersed in the water column, adsorbs to detrital material, is quickly weathered, and is redistributed by subsurface water movement through the study region. The petroleum hydrocarbon deposition rate was high during the second week after the spill (November 1 to November 9) and presumably was as high or higher during the last week in October. Thereafter the amount of petroleum hydrocarbons in the sedimen- ted material decreased markedly and 1 to 2 months afterward the spill reached very low levels. At these low levels, petrogenic hydrocarbon inputs are non-detectable. Very significant in terms of interpreting benthic as well as inter- tidal bivalve uptake patterns, are the detailed gas chromatograms of the hydrocarbon material found in the traps. As illustrated in Fig. 11.6 and 11.7, the spilled Tsesis oil that is sedimented has been significant- ly altered as early as 1 week after the spill. In the aliphatic fraction (Fig. 11.6), the n-alkane degradation has altered in oil's composition, presumably by microbial agents, to the point where the ALK/ISO ratio is 0.3 to 1.0, down from an initial value of 7.0. In addition, the increased importance of the UCM in these early deposition samples is evident from Fig. 11.6. 239 Figure 11.5 GC/MS Searches for Substituted Trimethyl Benzenes (TMB) A - TMB (m/e 133) searches on respresentat ive Myt ilus and Ma coma samples B - Display of m/e 133 search and total ion chromatograin 240 CH 3 I RH 2 C— CH- CH 3 1 R-H 2 R-CH3 3 R-CH3-CH2 4 R-CH3CH2-CH 2 m/»-l33 B iii.iiu, i.i m.ijil Ll.l 1 1 1_U 1 1 1,1.1 1 1 111 1 11 1 1 1 1 1 1 I 1 1 1 I 1 1 M II 1 M I 1 1 1 1 I I i s ta is aa as aa as 4a 46 sa ss aa « 241 TABLE 11.5 SEDIMENT TRAP HYDROCARBON DATA DATES ALIPHATIC HYDROCARDONS (pg/g) AROMATIC HYDROCARBONS (ug/g) STATION RESOLVED UNRESOLVED TOTAL CPI a AKL/ISO b RES. UNRESOLVED TOTAL V NOV 1 - NOV 9 333 3,276 3,609 0.9 1.0 32 3,624 3,656 NOV 9 - NOV 17 32 907 939 1.1 0.1 5 1,129 1,134 NOV 17 - DEC 4 20 396 416 1.4 <0.05 3 424 427 II NOV 2 - NOV 9 129 1,331 1,460 1.0 0.5 53 3,266 3,319 NOV 9 - NOV 17 94 1,350 1,444 1.2 0.2 18 1,354 1,372 NOV 17 - DEC 21 5 106 111 2.7 <0.05 2 76 78 IV NOV 2 - NOV 9 102 1,783 1,185 1.0 0.3 24 2,730 2,754 NOV 9 - NOV 17 6 20 26 1.5 <0.05 3 34 37 NOV 17 - DEC 21 40 30 70 2.5 <0.05 6 28 34 a CPI ■ Carbon Pre •ference 2( Index - _ n-c n-C 2? + 26 * 2n n-C 29 ) " C 28 + ~" 6 7o bft r v normal alk anea n " C 14 + n-C 15 f n " C 16 + n " C 17 ♦ n- C 18 isoprenolda 1470 + 1560 -f 1650 + 1710 + 1812 whore 1470 - farneaane, 1710 - prietane, and 1812 ■ phytane. 242 Figure 11.6 Representative Glass Capillary Gas Chroma t ograms of Sediment Trap Aliphatic Hydrocarbons A - TSESIS oil (November 2 - November 9) B - Station IV (November 2 - November 9) C - Station V (November 1 - November 9) D - Station V (November 9 - November 17) 243 3 -A 244 Figure 11.7 Repr e sentative G lass Capillary Gas Chroma tog rams of _S ediment Tr ap Aro matic Hydrocarb ons A - TSESIS oil B - Station II (November 2 - November 9) C - Station V (November 9 - November 17) 245 246 Just as the ALK/ISO ratio describes the chromatographic composition of the n-C,„ to n-C, range, the carbon preference index (CPI) describes i J lo the nature of the n-C„. to n-C„ n range. Alkanes within this range can indicate a petrogenic input (CPI = 1.0) or, as the CPI increases above 1, it documents an increased input of terrigenous n-alkanes having their source in vascular plant waxes (Farrington and Meyers, 1975). With increasing time, the hydrocarbon material sedimented in the region more closely reflects these normal terrigenous inputs (Table 11.5), thus illustrating that direct petroleum deposition appears to occur only during the early post-spill period, perhaps on the order of to 3 weeks. Thereafter, it is possible that deposition continues outside of the three station transects due to the slick movement. The changes in the character of the aromatic hydrocarbon composi- tion of the trap material are illustrated in Fig. 11.7 and 11.8. The GC profiles of the aromatic composition (Fig. 11.7) indicate that the amount of the resolved material decreases relative to the whole oil and, as was the case with the aliphatics, the UCM achieves a chromatographic significance. This was observed in the November Ma coma and December Mytilus GC profiles as well. However, unlike the aliphatic hydrocarbon changes which appear to be microbially mediated, changes in the aro- matics appear to depend on solubility considerations, the lower boiling compounds being preferentially lost. The total amount of resolved material decreases rapidly (Tables 11.5 and 11.6). However, on a comparative basis, consider Fig. 11.8 which, like the corresponding consideration of the Mytilus and Macoma data, indicates preferential loss of the soluble naphthalenes, biphenyl, and fluorene compounds, and soon thereafter the dibenzothiophene parent compound (see also Table 11.6). Microbial activity would result instead in loss of or conversion of the methyl side chains in the various homo- logous series. The composition of the aromatic hydrocarbons in the traps is severely altered by the third week after the spill and consists mainly of the phenanthrene series and substituted dibenzothiophenes . 247 Figure 11.8 Comparative Plot of Aromatic Hydrocarbon Composition of Sediment Traps and TSESIS Oil Normalized to Trimethyl Phenanthrene N, C]N, C2N, C3N = naphthalene (N), methyl N, dimethyl N, trimethyl N BP = biphenyl F = fluorene DBT, CiDBT, C2DBT = dibenzothiophene (DBT), methyl DBT, dimethyl DBT P, C]_P, C2P, C3P = phenanthrene (P), methyl P, dimethyl P, trimethyl P 248 I I I I I I o o o > > c u a. cm O a. o -a. •O U m •O •CD Q o a ~ U CO to «2 CO CO > ~ — v co CO a — a — a CO 9) CO 0) CO ^^ "" ^Z £Z ?Z S2 _ £«n £«ni Sds c •in Z • CO u Z ■ CM O z o o o 249 >> o o fM en a> >--w a. m o i-4 m e^g «M| • • • • Ciu E u in o r- «M . Q.JJ wj «** a o cm « CM • • • • JJ — 1 «3 \ o in o r» c> 41 J= C 0u o» e jj » < 3. " 0i W» E- b N Q. K D3 CM ^ <— t m JJ c ^ OT a VO r-\ >J3 «N i a o CM O fM -. «!-£ £- 3 E fj -u 2 o 4) at **— cu (N us «N »^ <-4 E X E- m r» <■- 1 * >— E o ca • • • ^: j= "C s u a o fM O jj jj Q J a e * rn < en C - 01 1 2 b. i 1 o 1 01 u Cu.! > > 41 JJ C ! M JZ J= C Q t r*i • 1 • • C Oi-C i-i (J IT « fM 01 J3 JJ rn |J<4 (1 « -o e fM o \JB X cu' Z f* m "0" JJ s % *-^ 1 cm • 1 • • J= 01 (J O «» i-H a-f-" c ° can oi C O u 1 fM •— t O fMjS 21 z •9 m m »H O JJ ~1 • 1 • • >. c 5 o o fM o J= «■ a' Ql oi e-i a. OjO ~* CO rn u — •a a u X J > D> JB 1 < J \ fM m »n ^" jj •» % ■E- o tr n m fM jz oi a. | O w s. a c fM u' E-> U — «j o» u m! Eh BS o » ISO. < p- rH »-l 2 Os «-i tn en • 2 UU C) r-i O > > > > C U » - CK O O o O 3 • a < m z z z z h B) ^ J u C Z 1J- — : 1 1 1 1 tfl 3 CM 01 < z u u ~ c Q "- 1 en CM fM \ >H 0) > > > > CJ u z c c cu o O O o ^ M R 0> C < > > c NN £ ci H a — js w a t c 250 11.3.3 Surface Sediments Initially 15 sediment samples covering the entire region were analyzed by glass capillary gas chromatography and several subjected to GC/MS analysis for detailed aromatic hydrocarbon determinations. Al- though several different suites of hydrocarbons were observed in the extracts (see Fig. 11.9 and 11.10), no petrogenic inputs were observed that related to the spilled Tsesis oil. The UCM was a prominent feature in most of the chromatograms . The UCM material is generally ascribed to anthropogenic inputs from several possible sources, including urban particulates (Hauser and Pattison, 1972) and weathered petroleum from storm runoff, municipal sewage, and chronic oil spills (Van Vleet and Quiun, 1978; Boehm and Quinn, 1978; Farrington et al., 1978) and is observed in most silt/clay surface sediment in the region (Rudling, 1976). The dominant feature other than the UCM, in the GC traces, was the terrigenous n-alkanes n.-C, ?I - , n-C 07 , n-C~ Q , n-C.,, , and their dom- inance over their even carbon number neighbors, yielding CPI values from 2.5 to 5.4 (Table 11.7) .• The absence of significant quantities of petrogenic hydrocarbons in the surface sediment (0 cm to 2 cm) was puzzling. It was hypothesized that the gravity corer used in the initial sampling might, on impact, blow away the fine floe layer, just on the sediment water interface, where newly deposited material resides. Also the depth of the sediment sampled (2 cm) might cause background levels of hydrocarbons (50 (Jg/g to 1,000 [Jg/g , Table 11.7) to obscure smaller quantities of petrogenic hydrocarbons (see also section 6.4.1). A careful resampling was performed at Station 20 with a wide-bore corer which presumably disturbed the surface layer less. In addition, two sections, the cm to 0.5 cm and 0.5 cm to 1.0 cm layers, were obtained for analysis. Again, the analytical results indicate no petro- leum in the sediment, with the possible exception of one 0.5 to 1.0 cm section, which indicates an input of lower boiling n-alkanes to the terrigenous assemblage (Fig. 11.9a). If this is Tsesis -related hydro- carbon material, then it has been significantly altered. 251 Figure 11.9 Representative Glass Capillary Gas Chromatograms of Surface Sediment Aliphatic Hydrocarbon s A - Station 20 (0.5-1.0 cm) - 1, 2 9 November 19 7 8 B - Station C, 20 December 1977 C - Station 20 (0.5-1.0 cm) - 2, 29 November 1978 252 o i i (A a C U a. I Z U o I to z 2 I J,! z SW4 v* V •*> i ^.u^ WjV UCM Vjty VAJ V-^U/m I JJ •WwVwA-m 4 AJ 1 w~ O 0« a i z uA/ yy^J^ J i*n ;JW UCM ,1 I B -V, >*7 o I z 1 1 IV' *i n ri O I z B3 O I z ■ (Jl 0,/lwu «/•»«*** W UCM 253 Figure 11.10 Representative Glass Capillary Gas Chromatograms of Surface Sediment Aromatic Hydrocarbons Station 20, 8 March 1978 !54 JU JJJ ,CX\rQTl«m i-\ r- IOOOOOOOOCN CTi o p» m CN P* ro r-\ in o in r^ \D O m 1— 1 o r-t O o O o o --i i-H o O ro joooooooooooooooooooooooooo < u M Eh < noo'«Tr^ocT\mfMro-^r~v£>r^inr--r^r~vrir'r , -*«Dr > *coo^rv£iin co H OOOOOOOOOOOOOOOOOOOOOOOOOr-l Ct! cu I Lnrororncs)fornrnv£)vx3roinlflOHC1^N(NHlflO (H i— I rH rH rH CN rHr-lCO ,_) rO ^j. p") co u\H>r s cn(Nnwr iininoroinootnP~cocNcocNr*-co'3'ro H| a !rn^DrooinrrrHO^rinr»r— ifN^c^incriCNOCNrncoo^i— to CN CN CN CN 00 Cd < En CO O M Eh < Eh 63 U zi o u Z 1 o CQ (X ** Cj! O cd a <-> Ml Eh < X Cu M < CO o ■H 4J CO jc a < Eh O Eh D CO Cd Eh > X — « a m — w Cd Eh < Q Cd CU < CO p*inininrocNincoTra^cNp~r^cr*inr^invoror«-ocop--^j , p>co i— ir^ocN-^r- itcnp»ocx»covocncnp»cnO'*i , i— ii— ivOrHP-p-vo N , « , >o5Hffin(Nriinmc\^'?intNcnNoinfN>oDcoc0^ WN^f^H(Nmin>>(NH mroi— iiHnr-CNoincx)r-r^-'!Tmcoini-HCNCNO"<3"CNiHO>x)i-H o^o^mcnnrHr^rHOO'TTrcNCOocNcrN cno^ncNr^vo'^rvDOoco'^cNr-jr^mn p* ■— i'^'VDrrcNr > »ninmmmcNCNinvx5«a , in en o »>D ro p» o i— t i— lorocrto,— ir— i in m CN CN in \£) ^ rH f— ! cNCNrr^ommcocrii— imon-Tr^cNrr^rcNr^vDrH CN CD <£> i-H O ^-imrHcnmr^roininoco-TCNOonr^o rHrncNOOGo,_ip»Ln roinrninrncN o p* p* P- P~ CO I I p- O O I co cn p» I I I r-\ CN ^O r-i r-i CO r» p» p- p- I I o o m cn I I <-{ CN vO P- P» I o r~* r~» p- p~ co r*p»p»r~»p»p-coco l I P» | i | p» r» O O I o O CN I I rocNCOrocNCNCovo I I I I I I I I co jq a CO CO CO CO p* p- p- I I p~ p» I p~ p» I I I I CN I t— ICNVJO,_|cNCNCO' _ o 5 *~- z 8 ° ° & ■ P 2 2 g is c UJ = ^ o^ 268 About 1,100 metric tons of oil were spilled in the region and about 700 tons were recovered during cleanup operations. The oil was driven by the prevailing winds to the northeast and, as a largely unweathered oil slick, impacted the coastline at stations D, B, C, and F. Here the littoral zone was severely impacted from a community standpoint (section 7). Concentrations of petroleum hydrocarbons in tissues were as high as 20,000 to 50,000 pg/g and at this level mortalities seem to have occur- red (section 7.2). As the slick passed through the region, significant quantities of petrogenic hydrocarbons were mixed into the water column and concentra- tions in the water as high as 50 (Jg/£ (greater than 100 times the back- ground levels) were observed (Boehm and Fiest, 1978). In addition, oil dispersed in the water column was sedimented to the bottom presumably through (1) adsorption to detrital material and sinking and/or (2) ingestion of oil droplets by zooplankters followed by fecal pellet transport to the benthos. Visual scrutiny of zooplankters obtained during the early stages of the spill demonstrated that oil had been in- gested by zooplankters in the water column (see section 4.3.4). Oil remaining at the water's surface apparently underwent only slow degradation due to chemical and microbial weathering until landfall occurred. However, petroleum material dispersed in the water column underwent rapid bacterial degradation of the aliphatic hydrocarbons with n-alkanes being rapidly depleted relative to the isoprenoid compounds and rapid removal of the lighter aromatic fraction due to dissolution. Measurements of the bacterial populations following the spill indicate an increase in the bacterial population in the water column possibly due in part to the availability of oil as a carbon source (see section 4.3.3). Those stations receiving secondary impacts of the spilled oil (i.e., those receiving a secondary landfall of oil), most notably sta- tion G (at first this was designated as a control station due to no obvious landfall of the spilled oil), received a degraded oil as seen in the Mytilus tissues due to the longer residence of this petroleum mater- ial in the water. 269 Uptake patterns of the hydrocarbon material indicate that rapid depuration of fresh oil characterized the Mytilus samples during the early months of the spill and throughout the following year degraded Tsesis petroleum was present in Mytilus tissues. One year after the spill, much of the petroleum was gone from mussel tissue except at station C from which samples continued to exhibit aliphatic and aromatic petroleum hydrocarbons in their tissues. This was presumably due to the greater initial exposure of the Mytilus population to oil at this station, Macoma , on the other hand, received a sizeable petroleum impact during the early stages of the spill probably due to direct sedimentation of the oil. After apparent depuration occurred during the winter, a secon- dary impact was observed, especially at station 20. The transport path of this secondary oil might include (1) landfall, (2) sinking at the shoreline with age, and (3) transport and redistribution throughout the 30-meter-depth basin of which station 20 is at the bottom. It is possi- ble that as the water temperature increased and pumping rates of both Mytilus and Macoma increased, the increased activity aided in the depura- tion of the former and recontamination of the latter. The station 20 location appears to be at the focus of the benthic impact of the spill which is observed at all benthic stations (station 2 west of Fifong Island included) except for station 15, the control station south of the island of Asko. Macoma balthica appears to be an excellent indicator of pollutant input to the benthos. As was previous- ly suggested by Shaw et al. (1976), Macoma apparently receives material identical in composition to that captured in the sediment traps. It is puzzling why, even with careful sampling of surface sediment, the direct confirmation of the presence of oil in sediment is ambiguous at best. Bieri and Stamoudis (1977) were also unable to directly confirm the presence of fuel oil in sediment in their experimental oil spill in spite of its obvious presence in benthic organisms. The hydrocarbon material present in the fine floe at the sediment/water interface is difficult to sample even with careful grab or core sampling. Thus, the sedimented hydrocarbons from the Tsesis spill may reside at this diffi- cultly sampled, highly mobile pseudo-surface from which Macoma obtains 270 its food. This fact may also account for the drastic elimination of the important sensitive benthic crustacean, Pontoporeia spp., from station 20, and its failure to reoccupy the station as of August 1978 (see section 6.3.2). The apparent contrasting behavior of Mytilus and Macoma vis-a-vis ingested oil may reflect more the duration of exposure and source trans- port route of the petroleum than any intrinsic differences in the two bivalve species. Depuration of acutely acquired hydrocarbons by Mytilus is apparently accomplished through flushing of water through the animal's gills. Other studies have shown that depuration of acutely acquired petroleum is fairly rapid though perhaps not complete (Fossato and Conzonier, 1976; Anderson, 1975; Kanter, 1974; Stegeman and Teal, 1973; Lee et al., 1972; among others). However, Boehm and Quinn (1978) and DiSalvo et al. (1975) have shown that chronically accumulated hydro- carbons are slow to be eliminated from bivalve tissues, thus suggesting that the duration of exposure is critical to the post-spill chemical recovery of a particular bivalve community. The transport and reintro- duction to and long resistance time of petroleum in the benthic environ- ment in the regions of the Tsesis spill may result in the much slower recovery of Macoma and the entire soft-bottom community from the effects of this spill, and in general points to the environmental complications caused by transport of petroleum to the benthos. 271 11.5 References Anderson, J.W. 1975. Laboratory studies on the effects of oil on marine organisms: an overview. Publication of the American Petroleum Institute 4249:1-70. Anderson, J.W., R.G. Riley, and R.M. Bean. 1978. Recruitment of benthic animals as a function of petroleum hydrocarbon concentrations in sediment. J. Fish Res. Bd. Canada 35:776-790. Bieri, R.H. and V.C. Stamoudis. 1977. The fate of petroleum hydro- carbons from a No. 2 fuel oil spill in a seminatural estuarine en- vironment, in D.A. Wolfe (ed.), Fate and effects of petroleum hydrocarbons in marine organisms and ecosystems. Pergamon Press, Inc. , New York, 332-334. Boehm, P.D. 1978. Hydrocarbon chemistry of Georges Bank and Nantucket Shoals, Inc. Final Report North Atlantic Benchmark Program, Bureau of Land Management, New York. Boehm, P.D. and J.G. Quinn. 1978a. Benthic hydrocarbons of Rhode Island Sound. Est. and Coast. Mar. Sci. 6:471-494. Boehm, P.D. and J.G. Quinn. 1978b. The persistence of chemically accumulated hydrocarbons in the hard shell clam, Mercenaria mercenaria Mar. Biol. 44:227-233. Boehm, P.D. and D.L. Fiest. 1978. Analyses of water samples from the Tsesis oil spill and laboratory experiments on the use of the Niskin bacteriological sterile bag samples. National Oceanic and Atmospheric Administration Report Contract 03-A01-8-4178 , Boulder, Colorado. Clark, R.C., Jr. and J.S. Finley. 1974. Uptake and loss of petroleum hydrocarbons by the mussel, Mytilus edulis , in laboratory experi- ments. Fish. Bull. 73:508-515^ Clark, R.C., B. Patten, and E.E. De Nike. 1978. Observations of a cold water intertidal community after 5 years of a low level, per- sistent oil spill from the General M.C. Meigs, J. Fish. Res. Bd . Canada 35:754-765. Conomos, T.J. 1975. Movement of spilled oil as predicted by estuarine non-tidal drift. Limnol. Oceanog. 20:159-173. Conover, R.J. 1971. Some relations between zooplankton and Bunker C oil in Chedabucto Bay following the wreck of the tanker Arrow . J. Fish. Res. Board Can. 28: 1327-1330. 272 DiSalvo, L.H. , H.E. Guard, and L. Hunter. 1975. Tissue hydrocarbon burden of mussels as potential monitor of environmental hydrocarbon insult. Env. Sci. and Tech. 9:247-251. Erhardt, M. and J. Heinemann. 1975. Hydrocarbons in blue mussels from the Kiel Bight. Env. Poll. 9:263-281. Farrington, J.W. and P. A. Meyers. 1975. Hydrocarbons in the marine environment. In G. Eglinton (ed.), Environmental Chemistry, Vol. 1, The Chemical Society Special Report No. 35, London. Farrington, J.W. and J.G. Quinn. 1973. Petroleum hydrocarbons in Narragansett Bay I. Survey of hydrocarbons in sediments and clams ( Mercenaria mercenaria ) . Est. and Coast. Mar. Sci. 1:71-79. Farrington, J.W., N.M. Frew, P.M. Oschwend, and B.W. Tripp. 1978. Hydrocarbons in cores of Northwestern Atlantic coastal and continental margin sediments. Est. and Coast Mar. Sci. 5:793-808. Fossato, V.U. 1975. Elimination of hvdrocarbons by mussels. Mar. Poll. Bull. 6:7-10. Fossato, V.U. and E. Siviero. 1974. Oil pollution monitoring in the Gulf of Venice using the mussel Mytilus galloprovincialis . Mar. Biol. 25:1-6. Fossato, V.U. and W.J. Canzonier. 1976. Hydrocarbon uptake and loss by the mussel Mytilus edulis . Mar. Biol. 36:243-250. Grahl-Nielsen, 0., J.T. Staveland, and S. Wilhelmsen. 1978. Aromatic hydrocarbons in benthic organisms from coastal areas polluted by Iranian crude oil. J. Fish. Res. Bd. Canada 35:615-623. Hauser, T.R. and J.N. Pattison. 1972. Analysis of aliphatic fraction of air particulate matter. Env. Sci. and Tech. 6:549-555. Kanter, R. 1974. Susceptibility to crude oil with respect to size, season, and geographic location in Mytilus californianus (Bivalvia). University of Southern California Sea Grant Report, USC-SG-4-74: 1-43 . Kator, H. 1973. Utilization of crude oil hydrocarbons by mixed cultures of marine bacteria. In D.G. Ahearn and S.P. Meyers (eds.), The microbial degradation of oil pollutants. Publication No. LSU-SG- 73-01, Louisiana State University, Baton Rouge. Keizer, P.D., T.P. Ahern, J. Dale, and J.H. Vandermeulen. 1978. Residues of Bunker C oil in Chedabucto Bay, Nova Scotia, 6 years after the Arrow spill. J. Fish. Res. Bd. Canada 35:528-535. 273 Lee, R.F., R. Sauerheber, and A. A. Benson. 1972. Petroleum hydrocarbons: uptake and discharge by the marine mussel Mytilus edulis . Science 177:344-346. Mattson, J. and P. Grose. 1978. Impact assessment of the USNS Potomac oil spill. Interim final report, NOAA. Mayo, D.W., D.S. Page, J. Cooley, E. Sorenson, F. Bradley, E.S. Gilfillan, and S.A. Hanson. 1978. Weathering characteristics of petroleum hydrocarbons deposited in fineclay marine sediments, Searsport, Maine, J. Fish. Res. Bd. Canada 35:552-567. Meyers, P. A. and J.G. Quinn. 1973. Association of hydrocarbons and mineral particles in saline solutions. Nature 244:23-24. National Academy of Sciences. 1975. Petroleum hydrocarbons in the marine environment. NAS , Washington, D.C., 1-107. Owens, E.H. 1978. Mechanical dispersal of oil stranded in the littoral zone. J. Fish. Res. Bd. Canada 35:563-572. Poirier, O.A. , and G.A. Thiel. 1941. Deposition of free oil by sediments settling in seawater. Bull. Amer. Assoc. Petr. Geol. 25:2170-2180. Rudling, L. 1976. Oil pollution in the Baltic Sea. A chemical analytical search for monitoring methods. Statens NaturvSrdsverk (Sweden) PM 783:1-80. Shaw, D.G., A.J. Paul, L.M. Clark, and H.M. Feder. 1976. Macoma balthica : an indicator of oil pollution. Mar. Poll. Bull. 7:29-31. Stegeman, J.J. and J.M. Teal. 1973. Accumulation, release, and retention of petroleum hydrocarbons by the oyster Crassostrea virginica . Mar. Biol. 22:37-44. Teal, J.M. , K. Burns, and J. Farrington. 1978. Analyses of aromatic hydrocarbons in intertidal sediments resulting from two spills of No. 2 fuel oil in Buzzards Bay, Massachusetts. J. Fish. Res. Bd. Canada 35:510-520. Van Vleet, E.S. and J.G. Quinn. 1978. Contribution of chronic petroleum inputs to Narragansett Bay and Rhode Island Sound sediments. J. Fish. Res. Bd. Canada 35:536-543. Warner, J.S. 1976. Determination of aliphatic and aromatic hydrocarbons in marine organisms. Anal. Chem. 48:578-583. !74 Appendix 1 Investigations Appendix 1: Investigations PELAGIC INVESTIGATION Station Date (1977) Depth S Pp l ■ amples Ph Z B S I (ref) Oct 25 50 X X X Nov 22 X X X Nov 22-Dec 14 X II Oct Oct Oct Oct Nov Nov Nov 28 29 30 31 02 02-09 09-17 23-25 X X X X X X X X Nov 17-Dec 21 X Samples for Degree of initial oil analyses oil impact 2 ++ x x x III Oct 31 x ++ IV: + V: +++ IV & V Nov 01 IV: X Nov 02 30-32 X Nov 05 V : 25 X X X X Nov 07 X X X Nov 09 X X X X Nov 11 X X X X Nov 14 X X X X Nov 17 X X X Nov 24 X Nov 02-09 X X Nov 09-17 X X Nov 17-Dec 21 X X VI (ref.) Oct 26 36 X X X X Nov 09 X X X X Nov 23 X X X X Nov 09-23 X Nov 23-Dec 14 X Echosounding for fish in Svardsf jarden was also undertaken on Nov 11 and Dec 15 1977, and Jan 11 and Apr 12, 1978. During the period 13-29/6, 1978, herring eggs were collected in the impacted area for laboratory experiments on hatching success and compared with eggs collected between 13/6-6/7 from an unpolluted area west of Asko Laboratory. 1 2 Pp = Primary production = none Ph = Phytoplankton + ■ light Z = Zooplankton ++ = moderate B = Bacteria +++ = heavy S = Sedimentation PHYTAL INVESTIGATION Station Date Depth (m) Samples en 3 • • O T3 P- 3 3 E to •u cd en CO Samples for Degree of initial oil analyses oil impact iJ CD 6 C cu a •H a) en Oct 27 Nov 09 Nov 15 Dec 14 May 02 Jun 20 Aug 28 Oct 30 '77 '78 ii ii ii 1-2 ii profile 1-2 ii ii ii x x X X X X X D Nov 03 Nov 09 Nov 15 Nov 14-15 Nov 17 Dec 14 May 02 Jun 13 Jul 04 Aug 28 Oct 30 Oct 27 Nov 09 Nov 14 Nov 15 Dec 14 May 02 Jun 20 Aug 23 Aug 28 Oct 30 '77 78 77 78 Oct Nov Nov Nov Nov Nov Nov Nov Nov Nov Nov Dec Dec 27 01 01-02 01-02 02 02-03 02-03 09 16-17 24 30 14 20 77 profile 1-2 ii 2 1-2 profile ii 1-2 ii 1-2 ii profile 1-2 x ++-+ 1-2 1-2 profile 1.5 2 1-2 1.5 2 1-2 1 5-6 ii 1-2 5-6 x x X X x X X X X X X X X X X X X X X X X X X X X X X X X X X 278 PHYTAL INVESTIGATION (Cont.) Station Date Depth Samp les Samples for (m) CO CJ 3 C •rA O s T3 3 u CO cd 4-1 e Oh CO 4-1 C n3 3 a* oil CO •H rH 3 -a 0) analyses 4-1 c 0) B •H -o CO D (cont) May Jun Jun Aug Aug 02 13 20 23 28 '78 n M ii M 1-2 profile 1-2 it it X X X X X X Oct 30 M n X X E Nov 09 '77 1-2 X X Nov 10 ii profile X Dec 14 M 1-2 X X May 02 '78 ii X X Aug 28 ii ii X F Nov Nov 09 10 '77 ii 2 profile X X X Dec 14 M 2 X X May 02 '78 ii X X Aug 28 ii ii X Oct 30 ii n X G Oct 27 '77 1-2 X X Nov 02 ii profile X Nov 02- 03 1! 1.5 X Nov 09 11 1-2 X X Nov 14- 15 rt 1.5 X Nov 16- 17 ii 1 X Nov 24 n 11 X Dec 14 ii 1-2 X X May 02 *78 ii X X Jun 14 ii profile X Jun 20 ii 1-2 X Jul 05 1! profile X Aug 23 II 1-2 X Aug 28 II ii X Oct 30 II ii X X Degree of initial oil impact +++ Nov 09 '77 2 Nov 02 '78 1-2 X X mac. in Fucus = macrofauna in Fucus metab stud, quant samp. M. edulis metabolism studies quantitative sampling Mytilus eduli s 279 BENTHAL INVESTIGATION Station Date Depth Samples Samples for Degree of initial (m) cd oil analyses oil impact cfl oil . ana o o ^ •U H • H o u cu C CD QJ cfl Cti J-J CO Q) e B B 6 B 4Jj o •H CD CO ■a ra CJ 4J CO ra o o *5 CO ■a CD Si CO 287 CO ^ !h ^ 03 q |~H CD rCi +J Ms ^ q "i CO ■H 3i CD v SH q tn o< fO S sh tr> U Hh 4J 3i •q |-H i^ M ns to 3 "H CO -q ■H +J i> q >q •H CO •H CD 3 fc> tn 'H q to ■H q •U CD to 4J ■H X t3 qj >q ■U tr, 3 4-1 M q 3 +J ■H s. ^ M Mh ■H ■H O ^ CD O -q to o q o •3 CD *H O -q CO q O ■H Ss CD -U 03 S ■u q CD -q O co CO tQ T3 CD CO 3 £ is ■u CO 288 &1 R a; a) 4J t3 Ss Sh rO CD 291 :0 to ■H O -c: M O i> o o 292 f ^ f jft/ I :t H ■«■!■ ■ ' « **&■ m m 'Wr ';v v '>?:' '' „ ( M ^H 1 | : ' ■ ft W JLm S I r ..IM LirV' i /f 6 to -h o +J is 03 a, QJ S M O ^ QJ 4J CO •H 03 QJ CO 3 o is to I "* ^ O to 4J o to •H +J QJ r2 R O O QJ +J "Q -Q -CJ 'a 0) 0) 4J Si QJ i> O O -iH O QJ TS Ss Cd Uh Uh is O CM Ol QJ E QJ i> O 5: R O QJ 'a to t) *H to c; QJ O tO 293 T3 R M to H CJ (U Sn -V to is to Mh O &, "H +J ■R 4J Sh O R O a, 0) to o M U 'a c A! to :0 to O O tJ -u O t) R HJ M O 4h O Oi 3 0) to o o 294 in s s O CO ■> &< : o St! o +J :0 to 'a CO 4h O O 295 ifff co *: 296 TSBBaBsSj-.. AioooViiVosJ 1 --- am