51

This work is a result - University of Rhode Island. DISCUSSION ... ~M tilus californianus Conrad from Piano Beach, Coal Ofl Point, ... and important member of the rocky intertidal

Embed Size (px)

Citation preview

This work is a result of research sponsored byNOAA Office of Sea Grant, Department of Commerceunder Grant 82-35227 to the University of SouthernCalifornia. The U.S. Government is authorized to

produce and distribute reprints for governmentalpurposes notwithstanding any copyright notation thatmay appear hereon.

USC-SG-4-74

Table of Contents

~Pa e~ ~ ~ ~ ~ ~ ~ iList of Figures

List of Tables ~ 4 ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ' ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ii

ABSTRACT e- ~ a ~ ~ ~ . ~ ~ ~ ~ ~ ~ e ~ ~

I. INTRODUCTION ~ ~ ~ ~ t ~ 1

II. MATERIALS AND METHODS

Collection of Mussels

of Collecting SitesAcquisition and ChemiCrude Oil

Experimental DesignData Analysis

and PhysiographyA.

~ ~ ~ ~ ~ 4 ~ ~ ~ 2cal Analysis ofB.

~ ~ ~ ~ 4 ~ ~ ~ ~ ~ 4C ~

D.~ 4 ~ 5~ ~ 6

I I I ~ RESULTS ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ e ~ o ~ 6

Crude Oil AnalysisSize Analysis of MusselsSurvival at each locality wseason and concentration of

Results of Observed Spawnin

A.

B.

C.

6 8~ ~

ith regard tocrude oil 9

13D. g s ~ t ~ ~

IV. DISCUSSION ~ ~ ~ ~ ~ i a 1 3

A Size ~ ~ ~ ~ ~ ~ ~ ~ 4 ~ ~ ~ ~ ~ ~ t ~ ~ ~ ~ ~ ~ 13.B. Season and Location ....................... 14

~ ~ ~ ~ ~ 18V. ACKNOWLEDGMENTS

~ ~ y y 2 0List of Figures ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ e ~ ~ ~ ~ ~

I.ist of Tables ~ ~ t ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ 21

22-41Figures and Tables

Literature Cited .................................. ~ 42

LIST OF FIGURES

~Pa e

Chart of California coastlineshowing sampling localities andwaste discharge sites ............ 23

Figure 1

Gas chromatogram of Santa Barbaracrude used in all experiments ..... 24

Figure 2

Gas chromatogram of Coal Oil Pointnatural oil seep crude ........... 24

Figure 3

Gas chromatogram of emulsifiedphase ! 0.45~ dia.! of SantaBarbara crude oil ................ 2S

Figure 4

Gas chromatogram of soluble orpartitioned fractions of SantaBarbara crude from experimental con-centration a! lxlO ppm oil, b!lxl04ppm oil, c! lxl05ppm oil

Figures 5a,b,c

25-26

Experiments from October 6, 1970to December 1, 1970 .........,,... 28-29

Figures 6-9

Experiments from December 30, 1970to March 30, 1971 .........,...... 30-31

Figures 10-13

Experiments from March 8, 1971to April 30, 1971 ......,.......... 32-33

Figures 14-17

Experiments from May 17, 1972.to July 15, 1972 ................. 34-35

Figures 18-21

Experiments from August 10, 1971to October 8, 1971 ............... 36-37

Figures 22-2S

Exper iments from October 20, 1971to December 21, 1971 ...........,. 38-39

Figures 26-29

Figure 30 ersusRelationship of size in length vbody weight in ~M tilne ceiifotni from Dr. Dale Straughan, personcommunication

anus

al

40

LIST OP TABLES

~Pa e

Table 1 ~

Results of size analysis of living anddead animals exposed to crude oil, aswell as size analysis of live anddead control animals

Table 2.

Table 3. Dates of detected spawnings duringexperiments ~ ~ ~ a 4l

Coordinates for sampling local it iesof ~M tilus califocnianua along Cali-fornia coast ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ a ~ ~ ~ 22

ABSTRACT

~M tilus californianus Conrad from Piano Beach, Coal OflPoint, Palos Verdes and Santa Catalina Island, California wereexposed to Santa Barbara crude oil under laboratory conditions.The oil dispersed into two phases both of which contacted themussels directly; an emulsified phase of small globules ! 0.45~ ! and a soluble or partitioned phase 0.45~ !. Sixtwo-month studies spanning twelve months were conducted. Theanimals were maintained in a constant environment chamber at15'C+1'. ~M tilus californianus succumbed faster and in highernumbers in the lxl0 ppm concentration of crude oil than in thelower concentrations of lxl0 ppm and lxl04ppm crude oil. Largerexperimental animals from Pismo Beach, Coal Oil Point and SantaCatalina exhibited significantly higher mortalities than theirsmaller counterparts. The two most' susceptible populationswere from Pismo Beach and Santa Catalina Island. Each popula-tion had different months of highest mortalities, with noencompassing seasonal pattern for all the groups. No corre-lation between periods of highest mortality and spawnings wasdetected.

I. INTRODUCTION

The increase in of f shore oil exploration and productionas well as transportation in the last few years has broughtwith it a corresponding increase in the potential hazard ofoil spillage. In the event of an offshore spill, depending onwind, tide and current conditions, oil will reach the inter-tidal area along the exposed coastlines first.

The mussel ~M tilus californianus Conrad i.s a conspicuousand important member of the rocky intertidal community. Thenumerical dominance of this organism is reflected in the denseclumps, often several centimeters deep, carpeting the middleintertidal zone. The fact that ~M tilus is a suspension feedinggeneralist, capable of efficient space utilization, probablyaccounts for its dominance Paine 1966, Levinton 1972!. Rao�953! estimates that M. californianus pumps from 1-2 liters ofwater per hour through its mantle cavity, depending on thesize of the individual and the water temperature. Food andother suspended or dissolved materials enter the mantle cavityon this feeding and respiratory current. This same watercurrent can bring emulsified, or partitioned fractions of con-taminating oil into the mantle cavity of mussels.

~M tilus cali.fornianus is itself a food source for manyspecies of animals, including man. In addition to t his, andpossibly more important biologically is the extensive micro-community supported within the masses of mussels. The byssusthreads, which attach the mussel to its substrate, catch andtrap tremendous amounts of detritus and sediment which pro-vide both a home and food source for a wide range of marineinvertebrates. The stability of this microcommunity is inti-mately tied to the well being of its sheltering mussel bed.An oil spill or some other major disturbance which destroysthe mussel bed would also eliminate this entire associationof organisms. This in turn could directly affect predatorand other groups which are dependent on these organisms. Chan�973! mentions his concern for the microcommunity harboredin the Duxbury reef mussel beds following the oil spill inSan Francisco Bay.

~M tilus californianus is abundant in most areas through-out its geographic range Alaska to Baja California! . Thusdifferent localities, both those exposed to chronic naturaloil seepage and those not exposed, can be sampled throughouta year without unduly stressing individual populations. Theseecological and practical considerations make ~M ti.lus califor-nianus an ideal choice for experimental study, with the re-sulting data having more than localized significance.

Generalizations about crude oil and its effects onmarine organisms can be a hazardous practice since there arenumerous physical and biological factors which may modifythese effects. The massive mortalities of marine organisms,expected both by Chan �973! in the San Francisco area andworkers studying the aftermath of the 1969 Santa Barbaraspill did not occur. Nicholson and Cimberg �971!, studyingthe Santa Barbara spill, suggest that mortality among inter-tidal animals e.g. Chthamalus fissus, was a result of physicaleffects, i.e. smothering and not chemical toxicity. Chan�973! reports similar results for limpets and barnaclesfollowing the San Francisco spill. Straughan �972! listednine specific variables which can and do alter the effects ofoil during a spill in nature. These included, l! the type ofoil spilled, 2! the dose of oil, 3! the weather conditionsprevailing at the time of the spill, and 4! the biota of theaffected area.

The purpose of the present experiments is to examine theeffects of some of these variables, namely, size, geographiclocation, previous exposure to natural chronic oil pollution,and seasonality on the tolerance of ~M tilus californianus toSanta Barbara crude oil. Kanter et al �971! present a pre-liminary discussion of intraspecific and geographic factorsinfluencing the "toxic" effects of crude oil on M. californi-anus.

There are a large number of uncontrollable variablesoperating in the field. The best way to determine the effectof the variables listed is to simpIify conditions by perform-ing laboratory experiments. This series of experiments willconsider only long-term two months! exposure to low levelsof crude oil, i.e. 10X crude oil in seawater. Other researchcurrently in progress considers factors such as short-termexposure to high levels of oil pollution, i.e. 100X crude oil,the effects on closely related species and natural chronicexposure to oil versus man-made pollution.

IX MATERIALS AND METHODS

A. Collection of Mussels and Physiography of Collecting Sites

~M tiles californianus Conrad were collected from fourlocalities along the California coast. Figure 1 and Table 1show the location and coordinates of each collection site.

The aim of this study is to examine the effects of oil onphysiologically different populations of mussels. Thereforethe populations selected were from sites influenced by vary-ing physical conditions. The sites included one area exposedto chronic oil pollution Goal Oil Point! and three areaswith minimal or no exposure to oil Pismo Beach, Palos Verdesand Santa Catalina Island!. One area was exposed to pollutionfrom a large city Palos Verdes!; while the other three areasbeing relatively remote from any large population centers werenot. In one area Pismo Beach! the mussels are exposed t.othe cold water of the California current and at another local-ity Santa Catalina Island!, the animals are at least periodi-cally exposed to the warm waters of the Davidson current.

Pismo Beach is north of the Santa Barbara Channel andfree from the influence of any major population Center, e.g.Los Angeles. Coal Oil Point is a natural oil seep localitywithin the Santa Barbara Channel. The third locality PalosVerdes is south of the Santa Barbara Channel and within theinfluence of the metropolitan Los Angeles area. The Los Angelescity and county sewage outfall systems and the city's harborentrance empty into the ocean close by Figure 1!. SantaCatalina Island is a non-mainland locality 40 km away from anymajor population center and south of the Santa Barbara Channel.

All experimental museels were collected at low tide fromthe mid tidal level. After removal from the substrate theanimals were immediately transported back to the laboratory.Care was taken not to unduly stress the animals, i.e. minimalaerial exposure time. In the laboratory the mussels were exam-ined and the macro organisms attached to the shell and amongthe byssus threads removed. The mussels were then acclimatedin aquaria with fresh filtered and aerated sea water for 24hours at 15 + 1 C.

The mussels from the Pismo Beach locality Pismo! werecollected from large clumps of animals attached to wooden pierpilings. Coal Oil Point Coal Oil! mussels were collected off

These three animals formed dense mats covering numerous boul-ders found at this site.

Palos Verdes mussels were collected from a large rock out-crop. These animals were in very dense, "carpet-like" matsdominated by mussels.

Santa Catalina Island Catalina! mussels were collectedfrom an exposed rock some dist, ance offshore. This large rockoutcrop Bird Rock! is located outside of Isthmus Cove, Santa

Catalina Island. This population also formed dense matsdominated entirely by ~M tilne.

B. Acquisition and Chemical Analysis of Crude Oil

The crude oil used in these experiments came from off-shore oil fields located on the Rincon trend, Santa BarbaraCalifornia. Oil samples were obtained as needed from a shore-site facility. Gas chromatography was used to analyze theconsistency between samples, and the composition of the crudeoil. A "fingerprinting" technique was employed using a Perkin-Elmer gas chromatograph. The gas chromatographic conditionsincluded dual flame ionization detectors with 10 ft. x 1/8 in.o.d. stainless steel columns packed with lOX GV-101 onChromosorb-W, 80-100 mesh, acid washed, and dimethyldichloro-silane treated. The temperature of the in!ection port was325 C, the manifold and detector temperature was 340'C. Thecolumn was programmed to 50-325 C at 5'C/min. with a Heliumcarrier gas.

When the crude oil was added to each experimentalaquaria, three distinct phases were obtained. The firstphase was represented by the bulk of the crude floating onthe surface of the seawater. This phase is essentially thesame as the whole crude oil. When the aquarium, containingoil and seawater, is aerated and kept in the environmentalchamber the next two phases arise. One was an emulsion oftiny globules p 0.45~ diameter! suspended within the sea-water. The other was the soluble or partitioned crude oilfractions 0.45~ diameter! in the seawater.

Analysis of the latter two phases was carried out asfollows. Three separatory funnels were set up inside an en-vironmental chamber. To each was added filtered seawater andcrude oil to concentrations duplicating those described inmate lais and methods section C Experimental Design! lxl0 ppm,lx10 ppm and lxl05ppm crude oil. Each mixture was vigorouslyaerated for 24 hours. At the end of this period the seawaterphase was separated from the emulsified phase by the use of aNillipore filtration system. The emulsified phase was col-lected on 0.45~ pore size! Millipore filters, the soluble partitioned! phase was retained in the seawater filtrate.Both phases were triply extracted with 50 ml of pentane andthen concentrated in a Rotoevaporator. Both phases were ana-lyzed by gas chromatography using the same "fingerprinting"technique as mentioned above, but using a Hewlitt-Packardmodel 700 gas chromatograph. The chromatographic conditions

were similar to those above using flame ionization detectorswith 10 ft. x 1/8 in. o.d. stainless steel columns packed with10X OV-1 on Chromosorb-W, 80-100 mesh, acid washed and di-methyldichlorosilane treated. The temperature of the injectionport was 300 C, the manifold and detector temperature was390'C. However, the columns were programmed to 120' - 320'Cat 5 C/min. The carrier gas used was 40 ml/min. of Helium.

C. Experimental Design

Eighty acclimated mussels representing a distinct sizerange Table 2! were selected from each locality for eachgroup of experiments. The eighty animals were divided intofour groups of twenty animals and each group placed in aseparate aquarium. For each locality; one aquarium contained4,000 ml filtered seawater only the control!; one aquariumcontained 4 ml oil plus 3,996 ml filtered seawater lxl0 ppmoil!; one aquarium contained 40 ml oil plus 3,960 ml filteredseawater lxl04ppm oil!; and one aquarium contained 400 ml oilplus 3,600 m1 filtered seawater lxl05ppm oil!. All aquariawere kept in a constant environment chamber at a temperatureof 15 + 1'C* with a photoperiod of 14 hours light and 10 hoursdarkness and vigorously and continuously aerated. Seawater andoil were changed at 48 hour intervals. Mortality was recordedat each change in terms of numbers and size class distribution.At the end of each experiment, the shell length of survivorswas recorded. Healthy animals were those that produced newbyssus threads. Smith �968! used this same criteria in histoxicity studies of BF 1002 following the Torrey Canyon spill.An animal was j udged to be dead when the posterior adductormuscle failed to close the gapping shell halves on stimulationby mechanical pressure.

A total of six experiments were completed in t' he periodOctober 6, 1970 to December 21, 1971 with each experiment last-ing approximately two mouths. One experiment conducted Nay 17,1971 to July 14, 1971 was repeated May 17, 1972 to July 15,1972 because of inexplicably high mortality in control groups.

* Environmental chamber broke-down Oct. 20, 1971Apr. 9, 1971, Sept. 8, 1971, Sept. 30, 1971Jun. 17, 1972

D. Data Analysis

The relationship between size and susceptibility to crudeoil was examined by comparing the mean sizes of the survivingand dead experimental animals. This involved multiple com-parisions, thus increasing the probability of a type I error Siegel 1956 p. 160, Steel and Torrie 1960 p. 106!. The lsdi.e. least significant difference Steel and Torrie 1960 p. 106!test was employed to combat this problem. It is an overalltest for the planned comparisons of paired means. Basicallyit is a students-t test using a pooled error variance. Thelevel of significance was set at O.OS.

Control survivor and mortality size data were treated ina similar mannet' to determine the normal trend of mortalityin each population.

III . RESULTS

A. Crude Oil Analysis

The crude oil used throughout the entire period of ex-periments was found to be a typical full range low paraffiniccrude. The chromatogram shown in Figure 2 displays a widecarbon number distribution from C2 to C3> with large peakstypical of n-paraffins. Due to solvent masking effects, com-pounds below CIO could not be determined accurately duringanalysis of the various phases, and therefore are not discussedin detail. All samples used in the experiments were found tobe identical with only slight variations in the concentrationsof the front end weathering products.

The crude oil was also found to be extremely similar tothat spilled from platform "A" during the 1969 Santa Barbaraoil spill. This similarity was ascertained by the super-imposition of gas chromatograms of the two crude oils ThomasJ. Meyers personal communication!. Hence the results of thisstudy can be related to field observations following the SantaBarbara spill.

Coal Oil Point crude from the natural oil seep was alsoanalyzed Figure 3!. This oil was found to differ distinctlyfrom the crude used in this study. There was a definite absenceof the n-paraffins and isoprenoids. Instead, a broad, unre-solved envelope of compounds composed of aromatics, napthenes,

and higher end asphaltic components were present.

During the initial period of weathering either in natureor under these laboratory conditions, the physical processesof evaporation, dissolution or mechanical dispersion are theprimary mechanisms for change in the character of the crude.Under the experimental conditions of this study, the crudeforms three distinct phases. The first comprises the bulkof the crude, and floats on the surface forming a surface film.This oil has the composition of the full range crude, differ-ing only by slight evaporative volatile losses.

The next two phases result from physical dispersion andthe combined effects of partitioning and solubility of crudeeil in seawater. These two phases are perhaps the most impor-tant to this work because they act to bring various crudefractions into direct contact with the submerged mussels.

The emulsified phase ! 0.45~ dia., Figure 4! is composedof small globules kept in suspension by the action of theaeration system. Soylan and Tripp �971! found a similarphase appearing as they increased the turbulence of theirstirring. I have personally observed this same sort of emul-sion forming under natural conditions of wind and wave action.The emulsified phase displayed a composition very similar tothe full range crude Figures 2 and 4!. Volatiles below Cll areabsent, this loss is most likely due to the vigorous aerationof the aquaria, and/or solvent masking effects. A similar lossbelow the C>2 range was found by Smith and Maclntyre �971! intheir artificial oi1 slick studies. In nature the crude oilspill is not a closed system, therefore, equilibration betweengas and water phases never actually occurs. Instead, normalwave action and turbulence caused by wind and waves will beresponsible for considerable loss of volatiles. In this respect,the experiments simulated natural loss of volatiles and lowmolecular weight compounds. Since volatility and solubility aredirectly related NcAuliffe, 1966!, the fractions that remaincontain the less soluble higher boiling point compounds. Thegas chromatogram of the emulsified phase Figure 4! shows thepresence of the higher molecular weight n-paraffins, branchedparaffins and small amounts of aromatics and napthenes. Inaddition, the end of the gas chromatogram where resolution hasdecreased shows the presence of some residuum asphalts and as-phaltenes. The emulsified phase, because of its suspension,may be pumped into the mantle cavity of the mussel during res-piration and feeding.

The third phase Figures 5a,b,c! is composed of crudefractions that are either dissolved or partitioned into the sea-water. These fractions, like the emulsified phase can enter in-to the mussels' cavity. They represent the fraction collected

during filtration �.45~ diameter. This phase forms a dis-tinct chromatographic envelope which includes the more solublearomatics, i.e. mononuclear benzene-like derivatives, highmolecular weight alkylbenzenes, and other polynuclear aro-matics. However its chromatogram differed from the crude oilin lacking n-paraffin and isoprenoid peaks. This phase aswith the emulsified phase, displays a lack of volatiles, thisloss again associated with the aeration.

The concentration of this last " soluble" ! phase wascalculated for the various experimental aquaria. This wasaccomplished by simple disc integration of the area underneaththe gas chromatogram curves. Although the original experi-mental concentrations were lxl0 ppm, lx104ppm, and lxlO ppmcrude oil by volume, the actual concentrations were much lessbeing lxlO >ppm, 5xl0 lppm, and 11x10 lppm respectively'

B. Size Analysis

The results of the size analysis can be found in Table2. No mortalities were recorded from the Palos Verdes controlgroups, and very few total mortalities were observed in theexperimental groups from this locality. Analysis showed nosignificant size difference between the groups of survivorsor mortalities. In addition the animals from this localityappeared in the field to be smaller than those from otherlocalities. This observation was born out with mean sizes ofexperimental animals being considerably smaller than experi-mental animals from other localities Table 2!.

Norta3.ity was significantly higher among larger animalsthan among smaller animals in experimental groups from PismoBeach, Coal Oil Point, and Santa Catalina Table 2!. Thesedifferences in size were significant at P< 0.001, PC 0.02,and P< 0.01 respectively.

The Pismo Beach and Coal Oil Point control results Table 2! indicate that there fs no significant differencebetween the sizes of the living or dead animals. However, asignificant PC 0.01! difference exists between the sizes ofthe 1iving and dead in the Catalina control groups. Thisdifference is negative, indicating a normal trend of deathin the larger animals from this population.

t:. Survival at each locality with regard to season andconcentration of crude ail

�! Figures 6-9 contain the analyzed data for theexperiments conducted fram October 6, 1970 throughDecember 1, 1970.

A single mortality was recorded from the Pismo Beachcontxol grou~ Figure 6! while no mortalities were recordedfor the lxlG ppm and lx104ppm groups. During the first tendays of the experiment four deaths were recorded from thelx10 ppm group.

A single mortality was recorded from the Coal Oil Pointcontrol grou~ Figure 7! while no mortalities were recordedfor the lxlO ppm and lxl04ppm groups. During the first six-teen days of the experiment six deaths were recorded from thelx105ppm group.

No mortalities were recorded fram the Palos Verdes control

or lx10 ppm groups Figure 8! during the experimental periodsThe lxlO~ppm and lx105ppm gx'oups suffered low mortalities of 3and 2 animals respectively. These occurred during the firsteight days of the experiment.

Inexplicably high mortalities were recorded for the SantaCatalina Is. control animals Figure 9!. Mortalities beganon the eighteenth day and by the twentieth day deaths exceeded50X. After twenty-four days only 4 animals remained aliveand no further deaths were recorded. Na mortalities were re-corded from the lxl03ppm and lx104ppm groups of animals. Highmortalities were recorded for the lxl0>ppm group beginning onthe sixteenth day and dropping to less than 50X survivorshipby day twenty. Mortalities continued until the thirty-fourthday when only 2 an.imals remained alive.

�! Figures 1G-13 contain the analyzed data for theexperiments conducted fram December 30, 1970 throughMarch 3, 1971.

The Pisma Beach control group suffered 2 mortalities, bothof these occurring during the first twenty-one days Figure 10!.No mortalities and 1 mortality were recorded from the lxl03ppmand lx104ppm graups of animals respectively. Two mortalitieswere recorded at the start of the experiments from the lx10>ppmgroups

No mortalities were recorded for the Coal Oil Point con-trol animals Figure ll!. Deaths occurring near the end ofthe experiments resulted in law mortalities of 2 and 1animals from the lx103ppm and lx10 ppm groups respectively.Three deaths occurring at various times were recorded for thelx105ppm group.

No mortalities were recorded from the control or ex-perimental groups of animals collected at Palos Verdes Figure 12!.

The Santa Catalina Is. experiments were conducted fromJanuary ll, 1971 through March 3, 1971- This delay resultedfrom collection problems. No mortalities were recorded fromthe control or experimental groups collected at this locality Figure 13!.

�! Figure 14-17 contain the analyzed data for theexperiments conducted from March 8, 1971 throughApril 30, 1971.

A single mortality was recorded from the Pismo Beachcontrol group Figure 14! while no mortalities occurred in thelxlO ppm grou~ of animals. Four mortalities were recordedfrom the lxlO ppm group of animals, most of these occurringduring the last two days of experiments. Dead animals wererecorded from the LxlO~ppm group after two days of experiments.After six days, mortalities exceeded 50X and after twa weeks,there were only 4 surviving animals. Two more deaths wererecorded from this group in the last few days of experiments.

Very few deaths were recorded for either the control orexperimental animals from Coal Oil Point for this experimentalperi~d Figure 15!. One mortality occurred in the control andIx10 ppm groups while none occurred in the lxlO or Lx105ppmgroups of animals.

No mortalities were recorded from the control or LxL04ppmgroups of animals collected at Palos Verdes Figure 16!. Twomortalities, occurring at the end of the experimental periodwere recorded fram the lx10 ppm group. The LxL05ppm groupsuffered 4 mortalities during the last four days of experiments.

No mortalities were recorded from the control or lxl04ppmgroups of animals collected at Santa Catalina Is. Figure 17!.Only 1 mortality was recorded from the lx103ppm group. Highmortalities were recorded from the lxlO ppm group of animalsbeginning at day two. Deaths exceeded the 50X level by daytwenty and leveled off on the thirtieth day with 2 survivinganimals.

� 10-

�! Figures 18-21 contain the results of the repeat ex-periments conducted from May 17, 1972 through July15, 1972.

No mortalities were recorded from the control, lxlO ppm,3

or lxl0 ~pm groups of animals from Pismo Beach Figure 18! .The lx10 ppm group suffered 5 mortalities occurring betweenday two and twenty-two.

No mortalities were recorded from the control, lx103ppmor lxl0 ppm groups of animals from Coal Oil Point Figure 19!.A single mortality vas recorded from the lx105ppm group.

No mortalities were recorded from the control, lxlO ppmor lxl0 ppm groups of animals collected at Palos Verdes Figure20!. The lxl05ppm group suffered 4 mortalities between fourtiethday and the end of the experimental period.

No mortalities were recorded from the control group col-lected at Santa Catalina Is. Figure 21!. No mortalities wererecorded from the lxl03ppm group until day twenty-two. Asharp increase in mortalities was noted at this time with deathsexceeding 50X by day twenty-eight and by day thirty there wereno survivors. The lxl04ppm group suffered heavy mortalitiesfrom the beginning of the experiments. Mortalities exceededthe 50X level by day eight and only 1 animal remained alive atday thirty. Mortalities began at day six in the lxl05ppm group.Deaths exceeded the 50X level by day twenty-two, and deathscontinued with no surviving animals past the thirtieth day.

�! Figures 22-25 contain the results of the experimentsconducted from August 10, 1971 through October 8,1971.

Two mortalities were recorded at the beginning of the ex-periments for the Pismo Beach control animals Figure 22!. Nomortalities were recorded from the lxlO ppm group until daythirty-six. Beginning at day thirty-six high mortalities wererecorded and by day forty-two deaths exceeded the 50X level.On day forty-eight mortalities ceased vith 3 surviving animals.The lxl0 ppm group followed a similar pattern with the onsetof mortality at day thirty-six. High mortalities were then re-corded. with deaths exceeding the 50X level by day fifty andleveling off by day fifty-two with 9 surviving animals. Thelxl05ppm group suffered the highest mortalities in the shortesttime period. Mortalities began on day twenty-two with deathsreaching the 50X level by day twenty-six. There were no sur-viving individuals from this group by day forty-four.

-11-

During the first ten days 2 deaths occurred in the CoalOil Point contxol animals Figure 23!. Low mortalities of 2and 0 were recorded from the lxl03ppm and lx104ppm groups ofanimals respectively. The lx10 ppm suffered high mortalitiesbeginning at day twenty-eight. Deaths reached the 50K levelby day fifty and by the end of the experimental period only 8animals remained alive.

No mortalities were recorded from the Palos Verdes con-trol group Figure 24!. The lxl0 ppm and lxl04ppm groupssuffered 1 and 2 mortalities respectively. Only 3 deaths wererecorded from the Ix105ppm group of animals these occurringat various times throughout the experiment.

No mortalities were recorded from the Santa Catalina Is.control group Figure 25!. The lxlO ppm and lxl0 ppm groupssuffered law mortalities of 2 and 1 respectively. High mor-talities were recorded for the lx10 ppm group beginning on daytwenty. Survivorship dropped to 15 by day twenty-two whereit levels off with no further mortalities.

�! Figures 26-29 contain the results of experimentsconducted from October 20, 1971 through December21, 1971.

Two mortalities vere recorded from the Pismo Beach con-trol group Figure 26!, while only 1 mortaltiy was observedfrom the lx10 ppm group of animals. Relatively high suvivor-ship �8 out of 20! vas noted for the lxl04ppm group untilday twenty-six. Four additional deaths between this time andthe end of the experiments resulted in 14 survivors. No mor-talities were recorded from the lxlO ppm group until daytwenty-eight. From this time through day thirty-four, mor-talities increased resulting in 13 surviving animals.

No mortalities were recorded for either the control orlx10 ppm groups collected from Coal Oil Point Figure 27!.A single mortality vas recorded from the lxlO ppm group and2 deaths were observed in the lxl04ppm group of animals.

No mortalities vere recorded from the control or lxlO ppmgroups collected at Palos Verdes Figure 28!. The lxlG ppmgroup suffered 3 mortalities which occurred between the four-teenth and twenty-sixth day. Two deaths were recorded fromthe Ix10 ppm group of animals.

Moderate mortalities � out of 20! were recorded from theSanta Catalina control animals Figure 29!. These deathsoccurred between day ten and day twenty. The lxlO ppm group

D. Results of Observed Spawnings

Spawning occurred during several of the experiments. Onlyobvious spawns, involving several animals, were discerniblewithin the oil and water mixtures in experimental aquaria. Therecorded occurrences Table 3! can be assumed to represent aminimal number of spawns. It is quite likely that individualspawns would pass undetected as the gametes are diluted andmasked within the aquaria. There was no correlation betweendates of spawning and periods of high mortality.

IV. DISCUSSION

A. Size

Mortality was related to the size of the mussel in threeof the four geographic populations studied. Mortality washighest among the larger animals from the Pismo Beach, Coal OilPoint and Santa Catalina experimental populations. There wasno relationship between animal size and mortality for animalscollected at Palos Verdes. If more deaths had occurred amongthe Palos Verdes animals perhaps a susceptible size range wouldemerge.

These results may be attributed strictly to physiologicaldifferences between these geographically different populations.Experiments comparing physiological data for animals from eachlocality might help in interpreting these findings. Rao's�953! observations supply the basis for a possible explanationfor these results. He reports that for N. californianus main-tained at minimal temperatures, pumping rates increase with wetweight until a certain peak point and then decline with in-creasing weight. In the present experiments, size data length!

� 13-

suffered high mortalities beginning afterexceeded the 5OX level by day sixteen, anthere were only 5 survivors by day twentytalities after this. Four mortalities ocppm group between day eight and sixteen.recorded from the lx10 ppm group. Deathsand by the end of the experimental periodanimals remained alive.

day twelve. Deathsd continued so that

and no further mor-

curred in the lxlOHigh mortalities werebegan on day eighteenonly 5OX of the

is not directly camparable to Rao's wet weight. Figure 30 D. Straughan, personal communication! shows an exponentialincrease in weight with a corresponding length increase forM. californianus from Ellwood Beach, California Figure 1!.Assuming this type of relationship for M. Californianus fromother localities, the following explanation is advanced forthe present findings. We can assume that the smaller animals those that lived! were pumping less water thxough their mantlecavity than the larger animals those that died!. This isalso true for the Palas Verdes animals which were not anly thesmallest animals overall but also the group suffering the low-est number of mortalities.

Pumping rate and water transport are related to feedingand respiration in M. californianus as well as other mollusks Jdrgenson, 1952, 1966!. This would mean that ail componentscarried in the water would pass through the mantle cavity ata rate proportional to the pumping rate. The gas chromato-graphic results indicate that both the soluble partitioned!fractions, and those fractions emulsified in the seawater couldbe, and probably were, pumped into the mantle cavity. Smi,th�968! observed oil globules in the mantle cavity of seeminglyhealthy M. edulis during the Torey Canyon study. The emulsi-fied fractions could be responsible for the deaths in experi-mental animals buc more likely they would be voided as faecesor pseudofaeces. Lee et al �972! reported discharge ofcontaminating petroleum hydrocarbons by M. edulis. The otherpossible fate for either of these phases is temporary or per-manent retention in the tissues. Blumer et al �970! found

Pumping rate appears to be one important factor in ex-plaining the contrasting results of size and mortality fromcrude oil. Radioactive tracer experiments similar to thoseconducted by Lee et al l972! are a promising approach tosolving the dilemma of which crude fractions and in what quanti-ties are responsible for observed deaths.

B. Season and Location

The toxicity studies were conducted over a twelve month

period. A seasonal pattern was anticipated which would pro-vide a basis for predicting periods of highest susceptibilityto crude oil effects. No general pattern was detected thatencompassed all the groups tested. Instead each geographicpopulation reacted in its own unique way.

-14-

Coal Oil Point and Palos Verdes animals were the two

most tolerant groups tested. The consistently high survivalin all concentrations during each experiment was proof ofthis. Only once did deaths ever exceed 50K and this was inCoal Oil Point lxl0 ppm oil during the experiments of August10, 1971 to October 8, 1971. In the remaining experimentsmortalities never exceeded 30X for either locality. The CoalOil Point findings were consistent with the findings of Kanteret al �971! for the lower concentrations, and with slightdifferences in the lx10 ppm oil group- Chan et al �973!testing the tolerance of M. californianus, from Duxbury reef,to Bunker C oil found similar high survival.

To explain the high resistance to the effects of crudeoil in the Palos Verdes and Coal Oil Point animals, it isnecessary to look at these individual geographic populations.Coal Oil Point is subj ected daily to crude oil coming fromthe natural seeps found at this locality. Allen et al �969!estimates from 11 to 160 barrels per day escape from the seeps.The portion of this reaching the intertidal animals varieswith prevailing tides, currents and wind conditions. Theanimals settling and living here are subjected to this environ-mental stress. The result can be selection in the watercolumn of larvae! and/or on the substrate for more fit indi-viduals. The individuals surviving may at least tolerate oracclimate to the persistent presence of oil. Whatever thespecific reason, the fact remains that these animals are lesssensitive to the toxic effects of crude oil than eitherCatalina or Pismo Beach groups. The high survival in Coal OilPoint animals adds further support to the low mortality find-ings of Nicholson and Cimberg �971! in field studies of thissame area following the Santa Barbara spill.

The explanation for the highly tolerant population ofanimals from Palos Verdes is open to even more speculationthan that for Coal Oil Point animals. Emery �960! cites thepresence of several natural oil seeps In the vicinity of PalosVerdes. These seeps however, appear to be far enough offshoreand infrequent enough in activity to have little if any sig-nificant effect on the intertidal animals. The proximity ofthe Los Angeles County and City sewage outfalls makes theselikely sources of environmental stress. As with the naturalseep oil of Coal Oil Point, the effluents of these outfallscould provide selective pressures for the establishment ofresistent populations. Galloway �972! reports exceptionallyhigh concentrations of capper and other trace metals in thearea surrounding the county and city outfalls. Undoubtedlyfurther investigation will yield data on waste petroleumproducts also in this effluent. Again the exact factorsoperating to establish this tolerant population are not clear,and one can only speculate on the likely possibilities.

-15-

The animals from Santa Catalina and Pismo Beach suffered

the greatest number of mortalities from the crude oil. Themortality figures varied not only with the month but also withthe concentration of crude oil. During the months of highestmortalities deaths in general occurred sooner and were greaterin number in the lx105ppm oil than any other concentration.

Santa Catalina animals show exceptionally high mortali-ties 7 50K! in all concentrations during the summer months ofMay and June. In addition the animals in lxl05ppm oil suffer-ed high mortalities during March, October, November andDecember.

Pismo Beach animals suffered high mortalities in allconcentrations during the fall months of September and October.In addition there was high mortality in the lxl05ppm oil groupduring the spring month of March.

The exact reasons for the high susceptability of thesetwo populations is not known. However, these animals are fromtwo localities with relatively "pollution free" environments.There are no natural oil seeps or major sewage effluentsadjacent to either locality, which would provide the selectivepressures presumably acting to establish the hardy Coal OilPoint and Palos Verdes populations.

During the months of highest mortalities for the variouspopulations and concentrations of crude oil two general patternsof mortality emerged. In some instances, e.g. Figure 21 SantaCatalina Is. lxlO ppm oil, mortalities began almost i~mediatelyat the outset of the experiment and continued until there werefew or no surviving individuals. The entire populationappeared to be susceptible to the crude oil. In most instancesthis pattern of mortality was associated with the groups ex-posed to the higher concentrations i.e. lx10 ppm, of crude oil.5

This pattern would suggest the "die-off" was a result of acutetoxic effects of certain crude oil fractions.

The second pattern of mortality is displayed by groupssubjected to all the concentrations, and particularly thelx10 ppm group, e-g- Figures 22 and 23. The trend shows high5

survival for approximately three to four weeks of exposure tothe crude oil. This was followed by a period of rapid and con-tinuous mortalities until few or no animals remained alive.One can only speculate on the explanation behind this pattern.A behavioral or physiological modification on the part of themussel to avoid or minimize contact with the oil is likelyiSupport for this idea is supplied by observations made duringthe 48 hour changing of seawater and crude oil. When thefresh seawater was added the mussels would separate their shellhalves and appear to be pumping water and respiring normally.

-16-

When the crude oil was added to the experimental aquaria themussels were observed to immediately close their shells. Howlong they can remain closed or the time period for adductorfatigue is not known, nor is the effectiveness of this avoid-ance behavior. This can only be ascertained with furtherstudy. Whatever behavioral or physiological change does occur,it is nonetheless apparent that this group of animals can sur-vive for three to four weeks time when subjected to chronicexposure to Santa Barbara crude oil.

Two patterns of survival arise as a result of moderatemortality among experimental animals. The first pattern ap-pears after initial mortality of a few mussels at the beginningof the experiment Figures 6,8,18!. Following these initialdeaths, there is a leveling off with survivors remaining aliveuntil the end of the experimental period. The second pattern,not quite as common, is mortality of a few mussels afterapproximately three weeks, followed by no further deaths Figure 21,25,28!. These patterns of survival appear to be aresult of acclimation by the animals. That is, a few initialdeaths occur during critical periods, and the remaining organ-isms acclimate to tolerate the presence of oil and thereforesurvive.

Successful spawning was noted at various times throughoutthe experiments for all localities. No direct correlation wasfound to exist between mortality and periods of observed spawn-ing. It has been reported by many authors Whedon 1936! thatspawning occurs year round with peak periods that vary from onelocation to another. The spawnings that occurred in these ex-periments appear to only incidently coincide with those reportedin the literature. Spawning obviously is related to the physio-logical state of the animal. If the organism is in a ripe statewhen collected, handling or thermal shock may initiate spawning.

No explanation is offered at the present time to accountfor the freak high mortalities observed in the control groupfrom Santa Catalina October 6, 1970 through December 1, 1970,However experiments are now in progress which are e~amining in-consistencies such as this. The possibility of heterogeneousmussel populations containing "healthy" and "unhealthy" indi-viduals or "old age" classes is being seriously considered.

There was no general annual cycle of susceptability toSanta Barbara crude oil in M. californianus from the four local-ities sampled. Perhaps a large scale study of a single localityspanning several years would yield a distinct pattern.

Straughan's �972! list of factors which can influencechanges in the environment following an oil spill should perhapsbe modified. As a result of this study the "Biota" classification

-]. 7�

should be expanded. When considering the biota one must notonly consider the species, but also intraspecies variations,such as those brought about by size, age, geographical andseasonal population differences.

V. ACKNOWLEDGMENTS

I would like to express my gratitude to the many peoplewho assisted me both in field and laboratory work. Amongthese people I am particularly grateful to Tens Martella,Ronald Harris, William Jessee, and Marion Johnson. I am in-debted to Dr. Dale Straughan for her continual encouragementand valuable criticisms during the development of my ideas.I wish to thank Dr. Thomas J. Meyers for his assistance withgas chromatographic analysis and subsequent valuable discuss-ions. I would also like to thank Robert Smith for supplyinghis computer program and assistance in statistical analysis.I wish to thank Nicholas Condap for his assistance in computerprogramming of the survivorship figures. I would also like tothank Tom Gaines and Don Craggs of Union Oil for supplyingcrude oil during the entire experimental period. My thanksalso to the Marineland of the Pacific for supplying filteredseawater.

Finally I would like to thank Dr. Ian Straughan, Dr.Richard Pieper, Dr. Kristian Fauchald and Dr. David Morafkawho read and made valuable criticisms of a draft of thispaper. The research was supported by Sea Grant P2-35227.

-19-

LIST OF FIGURES

Chart of California coastline showingsampling localities and waste dischargesites.

Figure l.

Figure 2. Gas chromatogram of Santa Barbara crudeused in all experiments.

Figure 3. Gas chromatogram of Coal Oil Point naturaloil seep crude.

Figure 4. Gas chromatogram of emulsified phase ! 0. 45~dia. ! of Santa Barbara crude oil.

Figures 5a,b,c Gas chromatogram of soluble or partitionedfractions of Santa Barbara crude from ex-

perimental concentration a! lx10 ppm oil, b! lxlO ppm oil, c! 1xl05ppm oil.

Figures 6-9 Experiments from October 6, 1970 toDecember 1, 1970.

F igur es 10-13

Experiments f rom March 8, 1971 to April 30,1971.

Figures 14-17

Experiments from May 17, 1972 to July 15,1972.

Figures 18-21

Figures 22-25 Experiments from August 10, 1971 to October8, 1971.

Figures 26-29 Experiments from October 20, 1971 toDecember 21, 1971.

Relationship of size in length versus bodyweight in ~M tilne californianne lfrom Dr.Dale Straughan, personal communication! .

Figure 30

� 20-

Experiments from December 30, 1970 to March30, 1971.

LIST OF TABLES

Coordinates for sampling localities of~N tilua oalifornianua along Californiacoast.

Results of size analysis of living anddead mussels exposed to crude oil, as wellas size analysis of live and dead controlanimals.

Dates of detected spawnings during experi-ments.

� 21-

Table 1. Sampling localities.

CoordinatesLocality

-22-

Pismo Beach

Coal Oil Point

Palos Verdes

Santa Catalina Island

Long. 121'18'20" Lat. 35'3'30"

Long. 119 52'45" Lat. 34 24'30"

Long. 118'21'40" Lat. 33'44'50"

Long. 118 30'00" Lat. 33 28'00"

0 0w N

N rt0>

0OO

I CJE4

Ww

8 0

0 0 8rt fb

0 0 0>W'

Ol

-23-

Figure 2: Gas chromatogram of Santa Barbara crude used inall experiments.

1elyeratare

Figure 3: Gas chromatogram of Coal Oil Point natural oilseep crude.

Figure 4: Gas chromatogram of emulsified phase ~ 0.45 dia.of Santa Barbara crude oil.

IayeritIra

Figure 5a: Gas chromatogram of soluble or partitionedfractions of Santa Barbara crude from experimentalconcentration, lxl0 ppm oil.3

� 25�

Tesyer ~ fere

Figure 5b: Gas chromatogram of soluble or partitioned fractionsof Santa Barbara crude from experimental concentra-tion, 1x104ppm oil.

Te I por at ere 'C

Figure 5c: Gas chromatogram of soluble or partitioned fractionsof Santa Barbara crude from experimental concentra-tion. 1xl0~ppm oil.

-26-

O O O VPv

Efl

O OV F4 Ch

C4

8

O I8 8 8

I I

0 0b0

CljIII

5 I-"III SH WgCh 0

4

0 C4OI

4 S Ac5 HCj M W+Ch 0

S SO

S8 8

O

I I

N

8 8

ICON

4 04Jtfj

0

W 0

N04 04J

00C3 C3

-27�

4J

CIa0 4C4

Ch CII0

4 AQl Cjj0 % W

0

di IW CII ~W N

SCh

<0III W

0

iJFLr

O cO0O

gl Clj

Cl

CCI0 CO

4J Q tjjC4 O

O OCh Ch I V

R I

00 00 I CC!II

1 I I8 8 I

I

8 8 II5 8 I 8Ch W I CO

I

8 8 I8 8 I

N NCO

S 8 88 8 8

I I I I~ 0N N N N

~ p!4J Ql ll

CllO 44l W Il CCICO W ~ 4

O0 9!g W 0 t5

aj0 Id t5V W 'U

KST TD FISIOCS' PISSD ~ IIIID/14 TO IITITID

44

L. -OD -%- -4- � 5- -5- -I- -5- -N- � S- � CS - }L -5- � I � � I- � I

! 'I ~ ~ II 'll

TIME IN DAYS

Figure 6: Survivorship curves for Pismo Beach animalsOct. 6, 1970 to Dec. 1, 1970.

ICT TD PI~St OOS DTL POIIDT IDI ~ IIS TO 'ICJIITI

I ~

TIME. IN DAYS

Figure 7: Survivorship curves for Coal Oil Point animals,Oct. 6, 1970 ta Dec. 1, 1970.

-28-

'I III'I I!

lt11ltIIIIE

0

KUS

ElLCJ

K 2

llIC'I I! !

ty 'I92IlltI 'IIZ14

Z.K

IC3LPI I

OXITOX.IXI ~ IPS OILIXII PPS OILIXII PISI OIL

P 4-t11 ID lt ID II Dl DI I ~ IO It St IC 14 ~ ~ 'll 'll IC I ~ ~ I ST ~ 'I Ct SI 40

~ COOSLIXI ~ PPtl DILIXII PPII Otl,IX'l4 IPS Ott

t t I ~ 14 II I ~ 'IC I ~ ID tt II IC I ~ 14 It I ~ II I ~ TS ~ t TT I t ~ 44 It I ~ 44 SS 14

110 ID 01~Or 00LJD TCDJJO 101 TITD TO Itll/It~ ~- - 05--

I ~

� ltIT~ 10

0 I ~ ~ 10 11 I'I 10 10 00 tt ~ 'I t92 tt ll tt I'I I ~ Jt 10 lt 00 'I ~ 'll ~ 4 ~ I 50 51 55 I ~TIME IV DAY5

Figure 8: Survivorship curves for Palos Verdes animalsOct. 7, 1970 to Dec. 1, 1970-

XXT ID I JDDOX DDTID COTE.QXT 15. Idld/It TO lt. 'Iltd

' k~+-+

~ III It 11 I ~ I ~ 54 00 10 00 tl 10 31 30 ll 11 Td It 00 11 ~ 0 ~ tt 10 III 50 10

T IME IN I!FIYS

Figure 9: Survivorship curves for Santa Catalina Islandanimals, Oct. 6, 1970 to Dec. 1, 1970.

� 29-

JTNZ

! lt! 'l l

IJ

I

LJJLJJJJJKZ

ll10Z

11!! IT

LTJIJll11tZI IIK

CLz

a

JJJEZ

- JL- - 4- -4- -k- � 4- � JL- - 4- - 4 - - k- � D- - JL- -D- � 4 � � D � � D � � D � -4- -O- -k

'I

4 -m -a- -D- � y%- -0- -8- -0- - D- � W- -I- -I- -51

CXXITIDX1510 0011 DIL1 0 10 000 OILIt lt 0011 Dll

DDm!OLlldld 000 OILI XIII Pill PIL1110 0010 OJL

att Tc e I elena slain cclxn Itrlerte lo Irlrl Icoslls.IX'll 1st OILIX'Ie 1st oiLIe'ie tsx CXL

11Z: II

ia

IX 11I ~il

~ ieZ I

ISCIXLLICXT ~

K Z ~ ~ ia it tt ii ia I ~ It et te te Ia te ie aa 11 te t ~ ~ ~ te t ~ ~ I ee It ae ea IeTIME IN DAYS

Figure 10: Survivorship curves for Piano Reach animalsDec. 30, 1970 to Mar, 3, 1971,

~ TO alar CCTX OIL eOIXT lareeltl Ia lrlrTI

ixle 1st oILIXI I Iet OILixi ~ ltx on.

'I ~

IT~ is

I ~! !it

Crl~ ii

IJS ItIIll

Z

0 I?IIS

CXSL 'I

Z ~ I I ts tt tt ti ie te tt tt t ~ te 11 Ie I ~ la te ta 92I v ~ 92a t ~ It It It as Ii taTIME IN DAYS

Figure ll: Survivorship curves for Coal Oil Point animalsDec. 30, 1970 to Mar. 3, 1911.

-30-

~ IT 10 41401451 41401 450055,1 II 14 10 I104104

I 400 DII'I 000 DI,1111 000 Dl

~+-+~~+-+-4~4-+~4-

11 4

14

!

1Ji?I

1

51 I 'll I I i! 14 .I ll 11 . ' ll 11 ll 11 I ~ 14 1 ~ 44 Il 41 14 44 4 ~T [MF 'X '.!I=El 0

Figure 12: Survivorship curves for Paloe Verdee animalsDec, 30, 1970 to Mar. 3, 1971.

OCT ID tl~ .' %Weal CHILI' 14, II I I/Tl TD Ill/IlCDANLI 4 I 4 1011 0 ELI I lt Ittl OIL1414 IIW 41510

I ~

ll1515

5? 14

IO IV3CLEI3 5EZ

14 10 10 tl tl I ~ I ~ 30 It I ~ 14 30 ~ 0 41 'I ~ 10 ~ ~ 04 10 5'I ~ 'I I ~ 501 I ~ I I 11 I ~

TIVE IN UHVII

Figure 13: Survivorship curves for Santa Catalina Islandanimals, Jan. 1, 1971 to Mar. 3, 1971.

� 31-

~ CT Tlt TROIS'S ~ IOTHI OCRCIT Itlt I IO till/ t 1I OIITROCPtt PPT OII Rl'I I'Pl' OltI lilt PPT O1l

I ~IS

IlZ I ~

11

IJISItI ~

Z

I

11!

E!

K Z T I I ~ ill IT I ~ IS I I I ~ Tl I'I SI I ~ 11 It t tl sl 'IO 11 I ~ SO I ~ III I ~ I I 'll 10t! "1E ! N DAvs

Figure 14: Survivorship curves f or Pismo Beach animalsMar. 8, 1971 to Apr. 30, 1971.

Ri'T TO TICCNCC COOL Ol POIRT I/ ~ Ill TO t IO",COTTTRR-.Cll Ptlt It',,S.O Pttt Ol1st II Pttl Ol.

I ~IT

Z IOIO

CL'I ~V! IO

TSK 11

OZ

I

n EAR'i'ID

E Z' I lit ts tl I Il ts st It tc 11 IS 11 I I 1 ~ 92I t tt tt tl IT tt I tt 11r <NE IN DH<~

Figure 15: Survivorship curves for Coal Oil Point animalsMar. 8, 1971 to Apr. 30, 1971.

� 32-

~ Xt TO Flail: ~ TVOVO I/O/Tl TO ~ /IO/Tl

TIME IN INLAYS

Figure 16: Survivorship curves for Palos Verdes animalsMar. 8, 1971 to Apr. 30, 1971.

SIT TO /IOISFSI 'VIIII IBTIXIIOI IS I/ ~ //I IO ~ /II//I

ItI 'IZ

II! I92Pg 'I l

llIIIIE ZSI

O

Ly

IV-. O V.. -V- V- V-..V -V V

t I I I 1I TI Is Is II tl tt I92 ts Tl III It 11 IS I'I II tt II II ~ I IS tf sl IS II II1 Ibid IN DRIES

Figure 17: Survivorship curves for Santa Catalina animalsMar. 8, 1971 to Apr. 30, 1971.

-33-

ITI ~Z

IOISIt

/! lt11IO

Z

O

ISLIJ OE.Z

/I

S

� V

'V � � VI t

IXVOOXIllf OFO OIL1XIO Otll OILIXIX OIO Oll.

I Oll I SOLIXI I IOO OILIIll ~ IVtl 0'llIXI ~ IVII OIL

t r to t ~4 p OOI 054 o trrrirl ro rrtlrrl IDIO L.1510II»0

POII 0 I I.PPII OIL

OIL

TIME IN DAYS

Figure 18: Survivorship curves f or Pismo Beach animalsMay 17, 1972 to July 15, 1972.

r t 10 pl Loll �01 ott potpr lrtrrrl to Irrtrll

I ~

tI I I I 11 11 r'I Il r ~

I ME Itt DA' '-

Figure 19: Survivorship curves for CoaL Oil Point animalsMay 17, 1972 to July 15, 1972.

-34-

IOPI! I

t14

14FZ Z ~

o �IZIEZ }

tr11Z

Il y I I

11 �rlII10E Z0

0 I

4J 5 IZ

~~ tI I ~ 0 14 10 ll 10 10 00 00 01 to t ~ 10 00 lt lt I ~ ~ P tl 92I 920 92 ~ 54 'll 01 00 I ~ II

orlrOOI110 PPP OIL5 I I PPtl 0 I II 5 I II PPPI OIL

t � ' ' ' ' ' t" ' ' I ' M- ~11 tl I ~ tt Zl tl ''I 11 I Il 11 Ir 4 tt rl 0 fl' " rl Ir

011 TD TI044115' PPLOT XIODXX I'tl/11 TO I/11/llXOPTOOL%14 PPlt OILI

t'110 %01% OILI 114 I%tv DIL14

'O- � IE- � lO- -54% � -O

IZ! tt

I!

V t~ ~.. V P4 '~l4 I 1 tt It tt lt 11 xl Iv 1% 92'T' 4 T~ 4 4 . PT ~ 4 tl ~ T~

11 I/ ". I I I tl %1 1% ll 10 VT tl 'I IIX ll 10T 5M[ ' g' I tLf'4 '.

Figure 20: Survivorship curves for Palos Verdes animalsMay 17, 1972 to July 15P 1972.

DET TD PTIDTDOv ~4 0DTIOItOI 15. 4/11/14 TD 4/40/44IXXIT PXE1110 PPII DTLItPTP PPIt DTL1510 PPD OIL

't

& � D

DIIItt

IIII'I'I I

II tI

I tIb ET- -ET- k. a,54,

ET- � Elv't

51- - El- � ET- � El0

I I I I 10 II 1% 10 44 00 10 4% 'I ~ I ~ I ~

I ~ll10'I 5

0C3

:JI

II-40- - E4- � Et- qD- � EL. -54- -ET- -6- -ET- � ET- -Et- � E4- -El

~ I 10I'I I ~ 10 %0 14 %% %5 I ~ 50 5T 0' ~ ~ 0TIFF IN DAYS

-35-

Figure 21: Survivorship curves for Santa Catalina Islandanimals, May 17, 1972 to July 15, 1972,

CET TO CICNOIC ~ IOINI 04NCN Iritrll IO 1014111COOT IIOL1010 PAI OILI Nit PPII OILIXIII PPN OIL

I ~I ~

I 'tl11IltZ

It'I92IXU!lt

ICjZ I

50 50 Pl ~ COTIME !N DAY I

Figure 22: Survivorship curves for Pismo Beach animalsAug. l0, 1971 to Oct. 8, 1971.

XEI ~ PICOXE5' CIAL OIL POINT ICIO/II Tl 14141TICONTOOLIXI ~ PPN OIL101 ~ PAT OILI ~ 14 PIOI II IL

4014

IO

L

T ['tP ! N JPYg

Figure 23: Survivorship curves for Coal Oil Point animalsAug. 10, 1971 to Oct. 8, 1971.

-36-

O IY.V!TCIIZt

Z

IIIlZ

IlI ~

vrut

I ~Z

EC- -0I1II1b

m- "55- � Ct

I II ~ 0 5 Il It 10 14 It ft tl tl 04 I ~ 40 I! II 55 Ic 10 II 11 10 cl 50

I ~ 10 14 IT 1 I I 14 11 11 IC 14 I '10 1 PC 10 51

PEF Tll FIIXQEE PIEQE XEN300 ~ /10/Tl 10 10/I/Tl

00-EL � EL- � EL- � + -g � 0 4- -4- - 4- 4- 4- 4-00--/4 -4--~-4--6 -5--5--I--5--L-L

4 4 � 4- 4 � -4 � 4 � -4 � � 4 � -4

I ~I ~ItVT

tt I ~ Tt XT tt t ~ I ~ 00 II 092 JX I ~ ~ I It I ~ 00 00 00 00 00 ~ I 00'I ~ I 10 I I I'I

TIME IN DRYL'

Figure 24: Survivorship curves for Palos Verdes animalsAug. 10, 1971 to Oct. 8, 1971.

PTT 10 FIOJXX'0 XIXITR LTITRL IWI IE I/ lt/FI TO 10/ ~ /11

10

1 I 'I

T I" ' 'i I/! I

Figure 2S: Survivorship curves for Santa Catalina IslandAug. 10, 1971 to Oct. 8, 1971.

V! I It? II

IIKZ I

0oV0LLTIZIEz

11Z

I

IJ

I/:VTTY

ImZ

IXXP/RTLI X IX Ptlt O'ILIt ld PPR QTI'I X I ~ PPII OIL

I/X/IRQLIXIII PPQ lit101 ~ PPR OILIX/0 PPX 0ll

Q 4- -Q--E--Q -4--Q--Qi -Q--Q--4--4--4- � ~ � -4- -4- -V - Q- Q- Q- -E -W--4- -4-- ~ - 4- -400--4- -4- .4- -d -4-',-4 4- -4- -4- -4- -4. -4- -4 -4- � o � d--4- -4- -4- -4- � 4- -4--4- -4

I

ET- Tj- � ET. - El- � 4- - LT- - El � ET � � EI- � 0- -EE

EP- -0- -ET- -ET- - El .ET- � EI

+ ~ /~T'f~' I ''T l '01 .".. I 'I I / II lt ~ tl "t . I FI 10

'0 1.11 Hea Ial ~ HHHCH 101�'I Ta1aHIHa,51 II PPTJ1010 PPI 11lele PP J!

I ~IeI ~1110

! I, ta 'J !J I: G- -!J II

0

IJ.

T [Ht,. ' HJ Af!

Figure 26: Survivorship curves for Pismo Beach animalsOct. 20, 1971 to Dec. 21, 1971.

a0J-

151 ~

~H~ ~I ~tJ 1 1 I ' ll '1 1 't 1 Tt I Ie 10 ! 1. 11 11

['Al ,

Figure 27: Survivorship curves for Coal Oil Point animalsOct, 20, 1971 to Dec. 21, 1971.

-38-

,! 11

15

11

Ia.n

i!'

! IJ I3 - EJ IT !- I! 6 - II

'T!, J I! I! i!

-+-8- -8- -0--�--8--4--1--0--Pi--4. -0- -A--a--~--H--*--&--R--�0 -'

4O tataI'4 '1 4 pte el ~

at ee 41Iae m at,

14

OI- � Lt- -Ia- � tf � � ae- � It- -41- � O- � Oo o 4 -4--4--O a � O.-l oIt

z

1

VIG.

14IT

Itt

4X.

z f T T I~ ~ 4 Ii I Ia I ~ 44I I I ~ 11 I 1 14 I ~ tt I ~ 1 I 1 ~ 4 11 1 14 ~ 14 et I 'I92

< 10E. I N DR f'8

Figure 28: Survivorship curves for Palos Verdea animalsOct. 20, 1971 to Dec. 21, 1971.

ale Il tIOIOaa fOltA ftlaa teel If. I ~ 144111 Tl lett'1111alalaaLI el ~ tee OIL141 ~ tel OILIa'I ~ ffl Otf

11I ~

14If

K1 ~

C:Z 4I

V!LdF.

z ~ I 14 I ~ I ~ 'lI I ~ tl lf I ~ tl tt ee at IV 44 14 te et 41 'I ~ ~ I al ~ f ll fa fl IeTIVE [N DRYS

Figure 29: Survivorship curves for Santa Catalina Islandanimals, Oct. 20, 1971 to Dec. 21, 1971.

-39-

~ mE

I

C%Ikey

ChCS

SIIELL LENQTII s s!

Figure 30: Relationship of size in length versus body weightin ~M tilua ca1ifornianua from Dr. Dale Straughan,personal communication!.

-40-

Table 3. Dates of detected spavnings.

Location Dates

-41-

P israo Beach

Coal Oil Point

Palos Verdes

March 1, 1971; March 3, 1971

February 12, 1971; March 3, 1971

October 9, 1970; March 10, 1971;October 20, 1971

LITERATURE CITED

Allen, A.A., Schlueter, R.S. 1969. Estimates of Surface Pol-lution, Resulting from Submarine Oil Seeps at PlatformA and Coal Oil Point. Technical Memorandum 1230. GeneralResearch Corp., Santa Barbara, Calif.

Blumer, M., Souza, G. and Sass, J. 1970. Hydrocarbon Pollutionof Edible Shellfish by an Oil Spill. Ref. No. 70-1.Woods Hole Oceanographic Institution. UnpublishedManuscript.

Boylan, D.B. and Tripp, B.W. 1971. Determination of Hydro-carbons in Seawater Extracts of Crude Oil and Crude OilFractions. Nature. 230:44-47.

Chan, G.L. 1973. A Study of the Effects of the San FranciscoOil Spill on Marine Organisms. In Proceedings of JointConference on Prevention and Control of Oil Spills.API, EPA, USCG. Washington D.C.: 741-782.

Emery, K.O. 1960. The Sea Off Southern California. Pub. JohnWiley 6 Sons Inc. New York. 320-321.

Gallowayf J.N. 1972. Global Concepts of Environmental Contami-nation. Mar. Pol. Bull. 3�!:78-79.

Jgrgensen, C.B. 1952. On the Relation Between Water Transportand Food Requirements in Some Marine Filter Feeding In-vertebrates. Biol. Bull. 103:356-363.

1966. "Biology of Suspension Feeding", Pergamon Press.

Kanter, R.G., Straughan, D, Jessee, W.N. 1971. Effects ofExposure to Oil on ~M tilus cslifornienus from DifferentLocalities. In Proceedings of Joint Conference on Pre-vention and Control of Oil Spills. API, EPA, USCG.Washington D.C.:485-488.

Lee, R.F., Sauerheber, R., and Benson, A.A. 1972. PetroleumHydrocarbons: Uptake and Discharge by the Marine Mussel~Mtilus edulls. Sci. 1.77: 344-346.

Levington, J. 1972. Stability and Trophic Structure in Deposit-Feeding and Suspension-Feeding Communities. Am- Nat.106 950!.

� 42�

McAuliffe, C. 1966. Solubility in Water of Paraffin, Cyclo-paraffin, Olefin, Acetylene, Cycloolefin, and AromaticHydrocarbons. J. Phys., Chem. 7�!.

Nicholson, N.L. and Cimberg, R.L. 1971. The Santa Barbara OilSpill of 1969: A Post-Spill Survey of the Rocky Inter-tidal. In Biological and Oceanographical Survey of theSanta Barbara Channel Oil Spill 1969-1970. Pub. AllanHancock Fndn.:325-400.

Paine, K.P. 1953. Rate of Water Propulsion in ~M tilesCalifornianus ss a Function of Latitude. Biol. Bull.104:171-181.

Siegel, S. 1956. Non Parametric Statistics for the BehavioralSciences. Mc.Graw-Hill. Pub. New York. 159-161.

Smith, C.L. and MacIntyre, W.G. 1971. Initial Aging of FuelOil Films on Sea Water. In Proceedings of Joint Con-ference on Prevention and Control of Oil Spills. API,EPA, USCG. Washington D.C.: 457-462.

Smith, J.E. 1968. "Torrey Canyon Pollution and Marine Life".Cambridge at the Univ. Press: 133-134.

Steel, R.G.D. and Torrie, J.H. 1960. Principles and Pro-cedures of Statistics. McGraw-Hill Pub. New York. 106-107.

Straughan, D. 1972. Factors Causi.ng Environmental ChangesAfter an Oil Spill. J. Pet. Tech. Mar.: 250-254.

Whedon, W.F. l936. Spawning Habits of the Mussel ~M tiluscalifornianus Conrad. Univ. Calif. Pub. Zool. 41�!:35-44.

-43-