DOC
EPA
United States
Department of
Commerce
National Oceanic and
Atmospheric Administration
Seattle WA 98115
United States
Environmental Protection
Agency
Office of Environmental
Engineering and Technology
Washington DC 20460
EPA-600/7-81 -088
May 1981
             Research and Development
             Effects of
             Experimental
             Oiling on Recovery of
             Strait of Juan de Fuca
             Intertidal Habitats

-------
   EFFECTS OF EXPERIMENTAL OILING ON RECOVERY OF
    STRAIT OF JUAN DE FUCA INTERTIDAL HABITATS
                   FINAL REPORT

                        by

  J.R. Vanderhorst, J.W. Blaylock, P. Wilkinson,
          M. Wilkinson, and G. Fellingham
     Battelle, Pacific Northwest Laboratories
            Marine Research Laboratory
             Sequim, Washington  98382
Prepared for the MESA (Marine Ecosystems Analysis)
     Puget Sound Project, Seattle, Washington
                 in fulfillment of

     EPA Interagency Agreement No. D6-E693-EN
           Program Element No. EHE625-A
      This study was conducted as part of the
      Federal Interagency Energy/Environment
         Research and Development Program
                   Prepared for

OFFICE OF ENVIRONMENTAL ENGINEERING AND TECHNOLOGY
        OFFICE OF RESEARCH AND DEVELOPMENT
       U.S.  ENVIRONMENTAL PROTECTION AGENCY
              WASHINGTON, D.C.  20460
                  SEPTEMBER 1980
             U.S. Srivlrotwental Protection  Agency

-------
                       Completion Report Submitted to
                 PUGET SOUND ENERGY-RELATED RESEARCH PROJECT
                     MARINE ECOSYSTEMS ANALYSIS PROGRAM
                   OFFICE OF MARINE POLLUTION ASSESSMENT
               NATIONAL OCEANIC AND ATMOSPHERIC ADMINISTRATION

                                     by

                  Battelle, Pacific Northwest Laboratories
                         Marine Research Laboratory
                          Sequim, Washington  98382
                                 DISCLAIMER
     This work is the result of research sponsored by the Environmental
Protection Agency and administered by the National Oceanic and
Atmospheric Administration.

     The National Oceanic and Atmospheric Administration (NOAA) does not
approve, recommend, or endorse any proprietary product or proprietary
material mentioned in this publication.  No reference shall be made to
NOAA or to this publication furnished by NOAA in any advertising or sales
promotion which endorses any proprietary product or proprietary material
mentioned herein, or which has as its purpose an intent to cause directly
or indirectly the advertised product to be used or purchased because of
this publication.

-------
                                  FOREWORD
     An anticipated increase in oil tanker traffic and proposals for
construction of subsea pipelines in the Strait of Juan de Fuca and northern
Puget Sound regions of Washington State are foreseen as part of the national
energy development plans.  These activities increase the opportunity for
spillage of crude oil into the marine ecosystems of the region.  The U.S.
Environmental Protection Agency has supported studies dealing with biologi-
cal characterizations, physical oceanography, trajectory modeling, pollutant
monitoring, and fate and effects of oil in the region.  These studies are
being administered by NOAA's Marine Ecosystems Analysis (MESA) Puget Sound
Project Office.  The research reported here deals with recovery of inter-
tidal and shallow subtidal communities in experimental habitats contaminated
with Prudhoe Bay crude oil.  The studies make comparisons in rate of
recovery by communities in experimental coarse and fine sand habitat and
hard substrate habitat, and a commercial clam bed.  They examine the role
of vertical distribution of habitat in the tidal zone, site, type of sub-
strate, season, and duration for recovery in field experiments.

-------
                                  ABSTRACT

     Experimental studies of the effects of Prudhoe Bay crude oil on the
recovery of intertidal infauna and epifauna were conducted in the Strait of
Juan de Fuca region of Washington State.  The studies experimentally evalu-
ated the effect of oil treatment, site, substrate type, season, and tide
level on the composition, density, and species richness of organisms colo-
nizing substrates which were initially free of organisms.  Significant
differences for some biological parameters were demonstrated for each of
the types of treatment contrasted (site, substrate type, season, tide level,
and oil).  Significant biological effects were demonstrated to be due to
oil treatments for 70% of 56 biological parameters evaluated in detail.

     Full recovery following contamination with oil was predicted for
sediment-borne infauna based on oil  retention time and recovery of infauna
in unoiled sediments.  Full recovery for epifauna on concrete substrates
could not be predicted from these studies because of the longer-lived nature
of dominant species and differing assumptions about what constitutes full
recovery.  Predicted full recovery for sand habitats at Sequim Bay and
Protection Island was 31 months following an initial oil treatment of
1,800 ppm.  Predicted full recovery for a commercial clam bed habitat was
46 months following an initial oil treatment of 2,500 ppm.  Density of the
principal species of interest on this clam bed (Protothaca staminea) was
significantly altered by the oil treatment during the first recruitment"
season.  Because of the longer-lived (compared to the general infauna com-
munity) nature of this species, it was predicted that effects on recovery
of the species may extend somewhat beyond that for the general infaunal
community.  "Best" and "worst" cases for chemical recovery of oiled concrete
substrates were three and 13 months.

     Effects from oiling on recovery is strongly related to feeding type of
infauna and epifauna but the influence is different depending on habitat.
For the sand habitats, detritivorous and herbivorous species were almost
universally influenced by the oiling.  Carnivorous species were about evenly
divided in their response to the oiling and, with one exception, no signifi-
cant effect was seen on the recovery of a suspension feeder.  For the commer-
cial clam bed, herbivores and suspension feeders were at least as sensitive
to the oil treatment as detritivores.  For the concrete habitat, detritivores
were not sensitive to the effect of oil treatment but herbivores and
suspension-feeders were highly sensitive.  Based on adjunct MESA studies of
trophic relationships, it appears that the severity of the influence on re-
covery of species in this study could be expected to have a deleterious effect
on important fish populations, and that the effect would extend somewhat
beyond the 15-month period studied in individual experiments in this program.

     Retention of oil differed depending on substrate type, tidal height,
and initial concentration.  Concrete substrates lost oil much more quickly
than sediments.  Oil was retained longer at higher tide levels than at lower
tide levels.  Proportionally more oil was retained in sediments initially
treated with higher concentrations of oil.

-------
                             TABLE OF CONTENTS

Foreword	iii

Abstract	  iv

List of Figures	 vii

List of Tables	xi

     1.  Introduction  	   1

     2.  Conclusions   	   6

          General  	   6
          Estimated Recovery Times 	   7
          Retention of Oil in the Experimental Substrates  	   8
          Significant Biological Effects From the Oil Treatments .   .   8
          Methodological Validation  	   9

     3.  Recommendations	10

     4.  Materials and Methods	12

          Study Sites	12
          Experimental Approaches  	  12
          Infaunal Studies 	  14
               Sediment Extraction and Chemistry 	  21
               Sediment Grain Size Analysis  	  21
          Epifaunal Recovery Studies 	  22
               Chemical Characterization of Bricks 	  23
               Biological Characterization of Bricks 	  23
          Grazer Manipulation Studies  	  23

     5.  Results	28

          Recovery on Hard Substrates	28
               A Perspective 	  28
               Hard Substrate Recovery -
                  Biological Data Presentation  	 ... 28
                    Monthly Experiments   	 29
                    Site Experiments	47
               Hard Substrate Recovery -
                  Total Oil Concentrations	63
                    Monthly Experiments 	 63
                    Site Experiments	63

-------
               Hard Substrate Recovery - Analyzed
                  Saturates and Aromatics 	   69
                    Monthly Experiments 	   69
          Clam Bed Recovery	74
               A Perspective	74
               Effect of Tide Level  on the General  Community  ....   78
               Analysis of Variance  for Taxonomic Groups  	   78
               Taxonomic and Trophic Composition  	   82
               Primary Species  	   86
               Species With Indicated Oil Treatment Effects 	   89
               Petroleum Hydrocarbon Data 	   89
               Sediment Grain Size	98
          Effects of Oil and Key Species Removal on Hard Substrate
            Communities and Community Recovery  	   98
               A Perspective	  .   98
               Taxonomic and Trophic Composition  	  104
               Treatment Effects - Field Exposure Time,
                  Oil and Grazers	104
               Total Oil on Treated  Bricks	115

     6.   Discussion

          Significant Effects from Oil on Initiation of Recovery  .  .  118
          Oil Treatment and Retention of Oil in Experiments 	  121
          Implications of the Oil Treatments for Overall Recovery .  .  124

References	128
                                    VI

-------
                              LIST OF FIGURES

Number                                                              Page

   1      Experimental  Sites 	    15

   2      Mean Densities for Mollusks in Numbers  per Square
            Meter by Month and in Numbers per Brick by
            Treatment	    33

   3      Mean Densities for Crustaceans excluding  Exosphaeroma  sp.
            in Numbers  per Square Meter by Month  and in
            Numbers per Brick by Treatment 	    36

   4      Mean Densities for Exosphaeroma sp.  in  Numbers
            per Square  Meter by Month and in  Numbers
            per Brick by Treatment	    37

   5      Mean Densities for Polychaetes in Numbers
            per Square  Meter by Month and in  Numbers
            per Brick by Treatment	    41

   6      Mean Numbers  of Species per Brick by Month
            for Taxonomic Groups and  All  Species	    43

   7      Mean Numbers  of Species per Brick by Tide Level	    44

   8      Mean Numbers  of Species per Brick by Oil  Treatment  ...    45

   9      Mean Densities for Suspension Feeders per Square
            Meter by Month and in Numbers per Brick by
            Treatment	    49

  10      Mean Densities for Detritivores in  Numbers per
            Square Meter by Month and in  Numbers  per
            Brick by Treatment	    50

  11      Mean Densities for Herbivores in Numbers  per
            Square Meter by Month and in  Numbers  per
            Brick by Treatment	    51

  12      Mean Densities for Carnivores in Numbers  per
            Square Meter by Month and in  Numbers  per
            Brick by Treatment	    52
                                  VI 1

-------
                              LIST OF FIGURES

Number                                                              Page

  13      Mean Densities of Individuals Within Taxonomic
            and Trophic Groups by Site	    53

  14      Mean Densities of Individuals Within Taxonomic
            and Trophic Groups by Month in Site Experiment ....    55

  15      Mean Densities of Individuals Within Taxonomic
            and Trophic Groups by Tide Level	    58

  16      Mean Densities of Individuals Within Taxonomic
            and Trophic Groups by Oil  Treatment	    59

  17      Number of Species per Brick by Site
            and Oil Treatment	    60

  18      Number of Species per Brick by Month and Tide Level   .  .    61

  19      Summary of Infrared Analyses in Terms of Individual
            Experiment Time Frames in Monthly Experiments at
            Sequim Bay	    65

  20      Monthly Mean Concentrations of Total Oil  on Top
            Surface of Experimental  Substrates at Sequim Bay   .  .    66

  21      Total Oil Concentration in Hard Substrate Site
            Experiments by Site,  Tide Level,  and Field
            Exposure Time	    67

  22      Total Oil Concentration in Hard Substrate
            Site Experiments at MLLW by Month of Experiment  ...    68

  23      Summary of Measured Saturate Compounds in Terms
            of Individual Experiment Time Frames in Monthly
            Experiments at Sequim Bay	    71

  24      Summary of Measured Aromatic Compounds in Terms
            of Individual Experiment Time Frames in Monthly
            Experiments at Sequim Bay	    72

  25      Comparison of Tide Levels  in Terms  of Measured
            Saturate and Aromatic Compounds 30 Days After
            Treatment at Sequim Bay	  .    73

  26      Comparison of Scraped and  Unscraped Bricks in
            Terms of Measured Saturate and Aromatic Compounds
            at MLLW Tide Level at Sequim Bay  30 Days After
            Oil Treatment	    75
                                 vm

-------
                              LIST OF FIGURES

Number                                                              Page

  27      Comparison of Sequim Bay and Commercial  Clam Bed
            at Discovery Bay in Terms of Species	    76

  28      Comparison of Sequim Bay and Commerical  Clam Bed
            at Discovery Bay in Terms of Natural Log of
            Numbers of Individuals/Square Meter 	    77

  29      Number of Species Within Taxonomic Groups in
            Commercial Clam Bed Experiment Distributed
            by Tide Level  	    79

  30      Natural Log of Number of Individuals/Square Meter
            Within Taxonomic Groups Distributed by Tide Level.  .  .    80

  31      Number of Species Within Trophic Groups  in
            Commercial Clam Bed Recovery Experiment  	    85

  32      Natural Log of Number of Individuals/Square Meter
            With Trophic Groups in Commercial Clam Bed Recovery
            Experiment	    87

  33      Time Series of Total Oil and Analyzed Saturate
            and Aromatic Compounds in Commercial Clam Bed
            Recovery Experiment at Discovery Bay,  May Through
            August, 1980	    93

  34      Vertical Stratification of Total Oil Concentration
            in Cores for Commercial Clam Bed Recovery
            Experiment at Discovery Bay, Spring-Summer
            Season, 1980	    95

  35      Mean Total Oil Concentration Due to Oil
            Treatment, Tide Level, and Vertical Stratification .  .    96

  36      Vertical Stratification of Analyzed Saturate and
            Aromatic Compound Classes in Relation  to Tide
            Level and Treatment From Commercial Clam Bed
            Experiment at Discovery Bay, Spring-Summer
            Season, 1980	    97

  37      Saturated and Aromatic Compounds from Commercial
            Clam Bed Experiment at Discovery Bay,  Spring-Summer
            Season, 1980	100
                                    IX

-------
                              LIST OF FIGURES

Number                                                              Page

  38      Mean Number of Species per Brick Related to
            Duration of Experiment 	   105

  39      Mean Number of Species per Brick Summarized by
            Main Effects of Oil  Treatment and Grazer Treatment
            of Experiments at Sequim Bay, May and June,  1980  .  .  .  106

  40      Mean Number of Species per Brick of Experiments
            Conducted at Sequim  Bay, May and June, 1980,
            Related to Tide Level	108

  41      Mean Number of Individuals per Brick Related to
            Duration of Experiments at Sequim Bay, May and
            June,  1980   	109

  42      Mean Number of Individuals per Brick of Experiments
            at Sequim Bay, May and June, 1980, Related to
            Tide Level	110

  43      Mean Number of Individuals per Brick of Experiments
            at Sequim Bay, May and June, 1980, Related to
            Grazer Treatment   	   Ill

  44      Mean Number of Individuals per Brick of Experiments
            at Sequim Bay, May and June, 1980, Related to
            Oil Treatment	113

  45      Mean Dry Weight of Algal Biomass and Density of
            Herbivores of Microalgae and Macroalgae  of
            Experiments at Sequim Bay, May and June,  1980   ....   116
                                    x

-------
                              LIST OF TABLES

Number                                                                 Page

   1      Schedule of Sampling for Oil Recovery Experiments 	    17

   2      Preparation of Units and Sampling Schedule for
            Experiments on Effects of Oil  on Recovery of
            Commercial Clam and Epifauna on
            Rocky Intertidal	24

   3      Species Composition and Analysis of Variance
            For Mollusk Density in Monthly Hard Substrate
            Recovery Experiments at Sequim Bay 	   31

   4      Species Composition and Analysis of Variance
            for Crustacean Density in Monthly Hard Substrate
            Recovery Experiments at Sequim Bay 	   34

   5      Abbreviated Analysis of Variance for Density
            of Exosphaeroma sp	39

   6      Species Composition and Analysis of Variance
            for Polychaete Density in Monthly Hard Substrate
            Recovery Experiments at Sequim Bay 	   40

   7      Analyses of Variance of Mean Number of Species Per
            Brick in Monthly Hard Substrate Recovery Experiments  ...  42

   8      Abbreviated Analyses of Variance for Densities
            of Trophic Groups in Monthly Hard Substrate
            Recovery Experiments  	  46

   9      Analyses of Variance for Density of Taxonomic and
            Trophic Groups in Hard Substrate Site Experiment  	  56

  10      Analyses of Variance for Numbers of Species in
            Taxonomic Groups in Hard Substrate Site Expeiments  ....  62

  11      Mean Monthly Concentrations of Oil on Bricks in
            Hard Substrate Recovery Experiments at Sequim Bay 	  64

  12      List of Saturate and Aromatic Compounds Identified
            by Gas Capillary Chromatography in Hard Substrate
            Recovery Experiments 	   70

-------
                              LIST OF TABLES

Number                                                                 Page

  13      Summary of Mean Numbers of Individuals and Species
            in Commercial Clam Bed Recovery Experiment  	    81

  14      Species Composition and Trophic Groups for Commercial
            Clam Bed Recovery Experiment	    83

  15      The Mean Density of Primary Species in Commercial  Clam
            Bed Experiment	    88

  16      Hypothesis tests for Density of Primary Species in
            Commercial Clam Bed Experiment	    90

  17      Species With Trophic Designation for Which Statistically
            Significant Oil Treatment Effects Were Computed 	    91

  18      Contribution to Taxonomic and Trophic Groups Overall
            and in Terms of Significant Oil Treatment Effects ....    92

  19      Taxonomic and Trophic Composition in Grazer Experiments .  .   101

  20      Composition of Herbivores Which Feed on Microalgae and
            Macroalgae in Sequim Bay Grazer Manipulation
            Experiments	114

  21      Total Oil Concentrations in Grazer Experiments
            at Sequim Bay (May-June, 1980)  	   117

  22      Summary of Tests for Statistically Significant Biological
            Effects from Prudhoe Bay Crude Oil Treatment  	   119

-------
                                 SECTION 1

                               INTRODUCTION

     The Strait of Juan de Fuca and northern Puget Sound regions of Washing-
ton  State  represent  some  of the  nation's  finest  inland  marine habitat.
Historically,  the fishery  and  shell  fishery  have  been prime  sources of
industry  and  recreation.   Noncommercial  organisms  in  shoreline habitats
represent  a source of  enjoyment  for residents of the  region  and give im-
petus to a  flourishing tourist industry.  Deep water in the channels of the
Strait  and the natural  harbors  provide a basis for a large  shipping in-
dustry.  Four major oil refineries in the north Puget Sound region have had
the  capacity  to   provide  petroleum  products   sufficient  for the  state's
needs.  Until  the OPEC embargo in 1973, the crude oil for these refineries
was  supplied  mainly  by pipeline from Canadian oil fields.   Since that time
the  state's refineries  have relied on tanker transport of crude oil, first
from  foreign  sources and,  since completion of the  trans-Alaska pipeline,
increasingly from Alaskan sources.  At the present time, the Strait of Juan
de  Fuca is being  considered  as  a throughpoint for  oil to  be  shipped from
Alaska to mid-western markets.  Such use would result in an order of magni-
tude  increase  in  the amount  of  crude oil  shipped in the  marine waters of
the Strait  of Juan de Fuca as well as to increase the risk from spillage at
off-loading  facilities.   There is public  concern  about potential environ-
mental  damage  resulting from  spillage  of crude oil  into  marine habitats.

     Against  the  foregoing background,  the  U.S.  Environmental  Protection
Agency, in studies administered by the  Marine  Ecosystems  Analysis  Project
Puget Sound, initiated a broad range of marine environmental studies in the
north Puget Sound region.   The major portion of these studies was aimed at
gaining an  inventory  of the marine biota in the Strait of Juan de Fuca and
the  northern  Puget  Sound  region, and  describing physical  transport pro-
cesses which  could aid  in  predicting movement  of  spilled  petroleum in the
Strait  of  Juan de Fuca-northern  Puget Sound  region.   As an  adjunct to
biological  inventories   and   studies  of  physical   processes,  the  present
studies were  undertaken  to  experimentally  measure potential recovery rates
of  organisms  and  communities within  especially  vulnerable  habitats should
they become impacted  by petroleum.   While direct studies of the effects of
petroleum on  specific organisms  and the biological  fate of petroleum have
been  and  will  continue to  be  addressed  by  other  agencies, the  studies
reported  here  involve the  experimental  application  of  petroleum to shore-
line habitat units to allow comparative evaluation of recovery under oiled
and unoiled conditions.

     Shoreline habitats are most vulnerable to  spilled oil  for at least two
major reasons.   First, at the interface  of water and land,  the non-soluble,
floating oil residue  comes  in direct contact with the  substrate.   Because
of  changing tide  and wave  action,   this  material   can  be  mixed with  the
substrate   in  the  case  of fine,  movable materials  (mud,  fine  and  coarse
sand, pea  gravel) and  repeatedly applied to  the  surface  in  the  case  of

-------
rock.   The  mechanisms of  effect  on marine  organisms are  not  necessarily
related  to  the  toxicity  of  the  petroleum  but rather  to such  things  as
physical   smothering  (rendering  substrate  unsuitable  for settlement)  or
interference  with  behavior processes.   A second  factor resulting  in  the
high  vulnerability  of  shoreline   habitats  to oil  relates to  the  hetero-
geneity  of  the habitats.   Unlike  open water or the  relatively smooth  ex-
panses of some bottom habitats, the shore consists of myriad irregularities
formed by cracks  and pools  and  mixes of  substrate sizes.  This  hetero-
geneity  allows  isolated  concentrations  of  petroleum  in  both water  and
substrate and  the  consequent  exposure  of populations for extended periods.

     The  habitat  heterogeneity  that  contributes  to the  vulnerability  of
specific  shore habitats  also  contributes  substantially to a high amount of
variation in the population spatial distribution within these habitats.   In
studies  of the MESA Puget Sound Project which had the goal of inventorying
shore  communities  (Nyblade, 1978;  Webber 1979),  difficulties  in measuring
community variables due  to  high  spatial variation were  mitigated by  a
classification of shoreline habitats into types relating to the predominant
substrate.  Nowhere do the types exist in a "pure" form.   Each of the types
contains  sub-parts  that  are clearly of each  of  the  other types.  The goal
of the present studies  has been to  test  hypotheses  about the influence of
Prudhoe  Bay crude  oil  on the  recovery of intertidal  communities related to
type.   To achieve this  goal, the  inherent  heterogeneity has  dictated  an
experimental approach using units of "pure" type.

     Three types  of habitat  were  chosen for recovery studies:   (1)  con-
crete; (2)  sand;  and (3)  a commercial  clam bed.  Even  in  the  "pure"  form
imposed  by the experimental approach used here, the  types  differ markedly
in their heterogeneity.   The  reasons  for selection  of  the  specific types
also differed.  Rock  is  the predominant type in the Strait of Juan de Fuca
and  San   Juan  Island portions  of  the northern  Puget Sound  region.   Its
expanse  alone gives sufficient  reason for  high  priority in  study of  its
recovery  potential.   In  addition,  based on inventory studies of the Strait
of Juan  de  Fuca  (Nyblade, 1978;  1979),  rock habitat is  perhaps the most
highly  diverse  and  productive  in  the region.   Unfortunately, from  the
standpoint of hypothesis  testing  studies,  rock also exhibits  the highest
degree of spatial  heterogeneity in  its populations.   Estimates of numbers
of individuals for a population at a given site typically exhibit a percent
coefficient of variation  from  150 to  several hundred percent  of the mean
number.   In  these  studies  concrete  was  used to  represent  rock.   Sand
habitats  are   also  quite  abundant  in more  or  less "pure"  form  in  the
northern  Puget Sound region.    Coarse, mobile sand  habitats  are among  the
lowest  in productivity  in the  Strait  of  Juan  de  Fuca  region (Nyblade,
1979).   However, where  sand is stable and finer components  are contained,
the  sand habitat  is extremely  important  for  fish  food  organisms.   The
primary  interest  in sand habitat  in this study  stems from the presence of
food organisms for bottom-feeding  fishes contained therein.  The habitat is
considerably  less  diverse  and  productive overall than  is  rock.   A highly
attractive  feature  of sand habitat for  experimental  studies  is  that  the
spatial   heterogeneity  within  units  amenable  for  experimental study  is
perhaps  at  its lowest for  intertidal habitat.  The third type  investigated


-------
here,  a  sand-pebble mixed  habitat associated with  a  commercial  clam bed,
was  chosen  primarily for its functional  role.   Commercial  shell  fisheries
are  an important  industry  to  the region.   The substrates  are  sometimes
"managed" in the  sense  that fine, highly  organic  materials are stabilized
with coarse sand and pebble-size rock.  Organism heterogeneity is less than
for  rock,  but  the  diversity and productivity of  these substrates exceeds
that  in  pure sand  habitat.   The  foregoing  characteristics of the target
habitats  had  direct bearing  on .the  experiments  planned and  the  specific
objectives deemed feasible for this investigation.
     In broad
encompass the
included:

          A.
          B.
scale, the studies were  divided
specific objectives.   The  tasks
into a number of tasks which
relating to the sand habitat
          C.
          D.
          E.
 Provide  a  time-series  survey  of  species  composition
 and  total  hydrocarbon  concentration  in experimentally
 prepared substrates.

 Measure  the  effect  from  Prudhoe  Bay  crude  oil  on
 reseeding  by early  colonizers  as related to  site
 and  substrate grain size during  a late summer-early
 fall  recruitment period.

 Measure  the  effect  from  Prudhoe  Bay  crude  oil  on
 reseeding  by later  colonizers  one year after initial
 colonization during the  late  summer-early  fall
 recruitment  period.

 Measure  at one experimental site the effect  from
 Prudhoe  Bay  crude oil  on reseeding by early
 colonizers during the  late  spring-early  summer
 season.

 Measure  at one experimental site the effect  from
 Prudhoe  Bay  crude oil  on reseeding by early
 colonizers as related  to tide  height.
     Tasks relating to the rock habitat included the following:
          F.




          G.


          H.
 Investigate the  suitability of  some  artifical  hard
 substrates for attachment  in the exposed  rocky
 intertidal zone  of the Strait of Juan de  Fuca  for
 future experimental  recovery studies.

 Investigate recovery of a  rocky intertidal community
 in terms of larval reseeding rate  for key species.

 Investigate recovery of a  rocky intertidal community
 in terms of mortality and  removal  of key  species.

-------
     A single task  dealt  with  the commercial  clam bed sand-pebble habitat:

          I.    Investigate recovery of a commercial clam bed in terms
               of larval  reseeding rate.

     Tasks A through E, which dealt with recovery of infauna populations in
fine and coarse  sand  habitats  were reported in detail in an interim report
on this project (Vanderhorst et al., 1980).   Priority was given those tasks
because:    (1)  background  information (Vanderhorst et al.,  1978)  permitted
the design of specific experiments; (2) the relative spatial homogenity for
contained  populations  ensured  cost  effectiveness;  and  (3)  important fish
food organisms were contained in the habitat.

     The objectives for Task A were:  (1) to  provide guidance in  selection
of an optimum  concluding  date  for early colonization experiments  in terms
of  hydrocarbon  retention  and  colonization  by  infauna;  (2) to  provide  a
seasonal  time series of species composition; and (3) to permit a continuing
evaluation of sediment retention of oil between early and late colonization
experiments.

     The objectives for Task B  were addressed  in a single experiment.  They
were  of  three kinds.   The first  objective was to  measure the  effect of
site, sediment source (particle sizes differed relating to source), and oil
treatment  on recovery  (species  density)  for seven  primary species  in  a
valid hypothesis-testing  framework.   The experimental  treatments  were oil
(two levels),  site  (two  sites),  sediment source (two  sources); these were
evaluated in a three-month late summer-fall  seasonal framework.   The second
objective  was  to measure  possible change in  hydrocarbon concentration and
composition  from initiation to  completion in  the above  experiment.   The
third objective was to measure  the effect of the experimental treatments on
recovery of all species (density, composition) in a descriptive statistical
framework.    For  this  latter objective,  the  same  type of  statistics were
used  as  for  the first  objective,  i.e.,  analyses of  variance.   However,
because of the  large  numbers of species  involved,  we cannot be certain of
the error probabilities.

     The  objectives  for   Task C  were  addressed  in  another  independent
experiment.   They  paralleled  the  objectives  for  Task  B  with  three
exceptions:  (1)   target  densities  and compositions  were  those  occurring
after 15 months  of  field exposure;  (2) a  single site was studied; and (3)
the number of primary species was increased to ten from the original seven.

     Task  D  and  E objectives were addressed in a third independent experi-
ment.  Again,  the  specific  objectives paralleled  those  for Task  B.   For
these tasks, however, the experimental treatments were oil (two levels) and
tide level  (two  levels).   These treatments were evaluated in a three-month
spring-summer seasonal framework.

     The investigation of recovery of epifauna  on  the  rock habitat  (Tasks
F,  G,  and  H)  had  distinctly different  objectives because  of:  (1)  the
severity of exposure conditions in which this type habitat normally occurs;

-------
(2)  the much  higher  degree of  heterogeneity  in  spatial  distribution for
species  as  compared to the sand  habitat;  and  (3) an expected much shorter
retention  time for oil  on the substrate.   Task  F  was  an investigation of
the  feasibility of using various  attachment  methods  for  different arti-
ficial   hard   substrates.    The  results  have  been  previously  reported
(Vanderhorst et al., 1980), and led to the use of the experimental approach
reported for Tasks  G and H.

     The objectives of Task G were addressed in 12  independent experiments.
Ten  of  these  were  conducted at a single site at monthly intervals (October
1979 through July 1980).  Treatments were:  oil  (two levels) and tide height
(two levels).   The  similar objectives for each of these experiments were to
evaluate  treatment effects  on:  (1)  chemical   characteristics  immediately
after oil  treatment,   and  at  five and 30-day  intervals  after field place-
ment; (2)  suitability  of oiled substrates for colonization as indicated by
the  presence   of  marine  larvae  based  on  daily  observations  immediately
following field placement; and (3) hypothesis tests for differences between
treatment  and  control  after  30 days field colonization  for:   major group
densities  (i.e.,  taxonomic  groups =  polychaetes,  crustaceans,  mollusks;
trophic  groups = herbivores,  carnivores,  suspension-feeders, detritivores;
and species richness overall and for taxonomic groups.  Two further experi-
ments addressed  these  same objectives when an additional site was added as
a treatment factor.

     The objectives for  Task  H were based on  a  different premise than any
of  the  other   tasks,  including sediment tasks,  in  these  studies.   Whereas
all  other  tasks examined  the  rate  of recovery for oiled  and unoiled sub-
strates  which  were  initially  organism free,  in this  task experiments were
conducted on substrates which were allowed to colonize for a period of nine
months prior to application of treatments.   The specific objectives were to
evaluate treatment  effects from:   oil (two levels);  grazers  (two levels);
and  tidal  height  (two  levels).   The  effects  were  evaluated  immediately
after treatment, and   five  days and 30 days after  field  placement.   Three
independent experiments  were  conducted to  evaluate the  three  intervals.
The  end points were tests of hypotheses for treatment  differences  in dry
weight  plant   biomass, major  group  densities  and  species  richness.   An
additional  objective was to  compare  the partitioning  of  petroleum hydro-
carbon  concentration between  the  substrate itself and contained organisms.

     The final task, Task  I,  had the principal objective of evaluating oil
treatment  effects   on  reseeding  by  larval   littleneck clams  (Prptothaca
staminea).   A  single experiment,  of three  months duration, during the late
spring  and  summer  had exactly  the same  configuration and  objective  end
points as did  Task  B,  and ten primary species  were evaluated.

-------
                                 SECTION 2

                                CONCLUSIONS

GENERAL

     1.  Experimental  studies of  the  effects  of Prudhoe Bay  crude  oil  on
the recovery of  infauna  and epifauna of the intertidal  zone of the Strait
of Juan  de Fuca were  conducted over  a two-year period.   The studies in-
volved field  placement of  oil-treated  and untreated  substrates  (trays  of
sediment and concrete  bricks)  at four sites,  three tide levels, and in two
recruitment seasons.   For infaunal  experiments,  the  numbers  and kinds  of
animals  colonizing  sediments  after 15  months  of  field exposure  closely
resembled the numbers and kinds of animals reported from similar habitat at
adjacent baseline sampling  stations.   The  similarity extended beyond over-
all numbers and kinds of animals since the relative distribution of numbers
and kinds  within the  major taxa also closely  paralleled  that reported for
baseline stations.   From  those  data  it  is  concluded  that  the  15-month
control  sediments were fully recovered and that  they  reasonably  represent
what one would  find by sampling uncontaminated sites on the Strait of Juan
de Fuca  with  similar  habitat.   For the commercial  clam  bed, the  molluscan
herbivores   and  suspension-feeders  were at  least  as  sensitive to  oil  as
detritivores.   In the  rock habitat, detritivores were not sensitive to the
effect of  oil  treatment but herbivores and suspension-feeders were  highly
sensitive.

     2.  Effects from the oiling on recovery is strongly related to feeding
type but the  influence  is  different depending  on habitat.   For the sand
habitats, detritivprpus and herbivorous species were almost universally in-
fluenced by the  oiling.   Carnivorous species  were  about  evenly divided  in
their response to the oiling and, with one exception, no significant effect
was seen on the recovery of a suspension feeder.

     3.  Based on adjunct MESA studies of trophic relationships, it appears
that the severity of  influence on  recovery of  species in this study could
be expected to have  a deleterious effect on important fish populations, and
that this  effect would extend somewhat beyond the  15-month period studied
here.

     4.  Because  oil  was mixed  into  sediment  for infaunal  studies,  the
present  case may be considered a "worst" case  situation in terms  of treat-
ment severity.   However, the  sediment-borne  concentrations of both total
oil and analyzed aromatic and saturate compounds were well within  the range
of concentration reported for sediments exposed  to actual  spillages else-
where.    Initial target concentrations  of 2000  ppm for summer-fall, 3-month
recovery and 15-month recovery and 1000 ppm for spring-summer recovery were
obtained.  Reductions  in  total  hydrocarbons were about 35% in three months
for summer-fall  regardless  of sediment type.    The comparable amount for

-------
MLLW  in  the spring-summer period was 43%.  There was a slightly more rapid
loss  at  the lower tide level  (53%).  At  15 months, total hydrocarbons were
reduced  by  85 and 97% for  coarse  and fine sediments, respectively.  Based
on  rate  of  loss data between  3 months and 15 months, it is speculated that
background  levels would be  reached  in a total of 18.5 months.  An important
contribution  to  sediment  infrared  spectra  (most  likely due  to biogenic
materials and unrelated  to oiling) was  identified  in  analyses of control
sediments after 15 months.
ESTIMATED RECOVERY TIMES

     1.   For the sand  habitats investigated  at  Sequim Bay and Protection
Island,  a  full  recovery of infauna  from  the effects of an initial experi-
mental  oiling  of 1,758 ppm total Prudhoe Bay  crude oil was estimated to be
31 months.

     2.  Using the 15-month recovery of untreated sediments as a definition
of  full  recovery, 3-month  recoveries  in  similar sediment type,  site and
tide conditions were 69% for summer and 82%  for fall in terms of numbers of
species,  and only 11% for summer and  18% for fall  in  terms  of numbers of
individuals.   In 15  months,  oil-treated  sediments  had  recovered more than
90%  in  terms of numbers of species  but  had recovered only 48% in terms of
numbers  of individuals.

     3.  Although effects from oiling on recovery were found at each of the
tide levels  (MLLW and -2'[0.61 m] ), in  sand habitats at each of the sites
(Protection  Island and  Sequim  Bay), in each of  the sediment types (Sequim
Bay  native  and  Protection  Island   native),  and  in each  of  the  seasons
(spring-summer  and  summer-fall),   these   physical  factors,  nevertheless,
influenced gross density in both treated and control sediments.   Thus, much
higher  densities  were  found  in the summer-fall  season  than in  the spring-
summer   season;  much  higher  densities  were  found  at  Sequim  Bay  than
Protection Island; and  much  higher  densities  were  found  at -2'  below MLLW
as compared to MLLW tide level.  Density related to sediment type tended to
be  species   specific  and was  about equal   overall.   Studies  designed  to
elicit  effects  on recovery in an  actual  oil  spill event  will,  thus, need
some form of experimental control for these variables.

     4.   For  the commercial  clam  bed  habitat,  a full  recovery of  the
general  infauna  community  from  the  effects of  an initial  experimental
oiling  of  2,500 ppm  total  Prudhoe  Bay  crude oil, was estimated  to  be 46
months.  Recovery of  the primary species, the littleneck  clam  (Protpthaca
staminea) may be  slower  because of depth stratification of oil  as detailed
below.

     5.  For the experimental  concrete  habitats, a full  biological  recovery
cannot  be  predicted because of the lack of  a  usable  definition  of full
recovery.   Estimated   "best"   and  "worst"  cases  for reaching  background
concentrations  of total  Prudhoe Bay  crude  oil  at  an initial  treatment
concentration of 8.72  grams  per substrate unit (approximately 500 grams per
square meter) were three and  13 months, respectively.

-------
     6.   Predicted  differences  in  recovery time  for infauna  between the
sand and  commercial  clam  bed habitats related strongly  to  initial  concen-
trations of Prudhoe Bay crude oil and tide level.

     7.   Predicted  recovery time for epifauna  on  concrete  substrates did
not bear a direct relationship to the amount of oil applied to experimental
substrates.
RETENTION OF OIL IN THE EXPERIMENTAL SUBSTRATES

     1.  A  ranking  of  substrates in retention of  oil  from higher to lower
is  (1) the  commercial  clam  bed  habitat;  (2)  sand  from  Sequim  Bay  and
Protection  Island habitats; and  (3) concrete bricks used to represent rock
habitat.

     2.  Experimental substrates  of all  kinds which were  placed  higher in
the  intertidal  zone retained more oil than  those of  similar  kind placed
lower  in the intertidal  zone.

     3.  For  the commercial  clam bed substrate, more  oil  was  retained at
greater substrate depth.

     4.  For the sand substrate,  a high proportion of total initial concen-
tration of  oil  (about  50%)  was  lost  from experimental  substrates in three
months time.   Analyzed  saturate  compounds  were  lost from  substrates at
about  the  same rate as  total oil.  Analyzed aromatic  compounds  were lost
much more quickly.

     5.  For  the commercial  clam bed, a much Tower proportion of initial
oil  concentration   (about  13%)  was  lost from  experimental  substrates in
three  months  time.   Analyzed  saturate compounds were  lost from substrates
at  about  the  same  rate  as  total  oil.  Analyzed  aromatic  compound concen-
trations  had  not   changed  from  initial  concentrations  in  three  months.

     6.  For a combination of  the sand and sandy mud substrates, the reten-
tion of oil was closely tied to  the initial  concentration applied and tidal
height of field exposure.

     7.  For  the rock  habitat represented by concrete  bricks,  a high pro-
portion (84%) of total  oil on  bricks was lost in five days.

     8.  For  each  of the types  of  substrate,  the  total oil concentration,
and  concentrations  of  analyzed  saturate and  aromatic  compounds,  were well
below  concentrations reported  in actual oil  spill events.


SIGNIFICANT BIOLOGICAL EFFECTS FROM THE OIL TREATMENTS

     1.  The mean magnitude of 70% of 56 biological parameters estimated in
these  experiments was significantly reduced by the oil  treatments.

-------
     2.  Two species,  because  of their nearly ubiquitous occurrence within
the recovery  experiment and the north Puget Sound  region,  generally have
been  identified  as good  recovery  indicators when  used in  an appropriate
experimental  framework.    These  species  are  the crustacean,  Leptochelia
dubia, and  the polychaete,  Exogone  lourei.   A taxonomic  problem with the
former and  an  important oil-resistant congener of the  latter were identi-
fied.   The  species were verified  to  be useful  in the  commercial clam bed
habitat.   Exogone  lourei   should not  be used at  tidal  heights above MLLW.

     3.   In the  commercial  clam  bed  habitat,   and  in the  rock habitat,
mollusks  were  especially  abundant,  and sensitive  to oil  treatment.   The
magnitude of effects  from  oil  on mollusks often  exceeded  the magnitude of
effects due to tidal height in  these habitats.

     4.  In the  commercial clam  bed habitat, the density of the  littleneck
clam,   Protothaca  staminea,  was  significantly  reduced  by  oil  treatment.


METHODOLOGICAL VALIDATION

     1.  To protect the validity  of statistical  procedures, 13 species were
selected as "primary"  to  evaluate  effects on recovery  in  terms of indivi-
dual  species  density.   The  numbers  of individuals of  the primary species
comprised as very  substantial  proportion  of all  individuals  in this study
(78%)   as well  as  at the Beckett  Point baseline station (73%).  They repre-
sented 33% of all individuals at  the Jamestown baseline station for compar-
able conditions.   Three-month control  recovery for these primary species in
terms   of numbers   of  individuals  closely  paralleled  that  seen for  all
individuals (19% fall; 8% summer).   The primary species represent the three
major  taxonomic categories quite  well  (polychaetes,  crustaceans, mollusks).

     2.  The native and  artificial  substrates used in these studies proved
to  be  a  highly  satisfactory means for discriminating  between natural  and
pollutant effects in an experimental  framework.

-------
                                 SECTION 3

                              RECOMMENDATIONS

     The data  in  this  study verify the  utility  of experimental  approaches
to  measuring  recovery  of  intertidal  fauna  following insult  by sediment-
borne  pollutants.   There  are  some specific  studies using  the  approaches
which  have  high priority  and  are recommended  below based  on the present
findings.

     In  the  realm  of  oil  pollution  research,  the  following should  be
investigated:

     1.   Supplemental   studies  should  be  conducted  to characterize  the
organic and inorganic  constituency  of  substrates  used  in  these  studies.
Appropriately preserved samples are available.

     2.  Further sampling of substrates in place on the commercial  clam bed
evaluated  in  these  studies is  warranted.   Both  chemical  and  biological
evaluation  should continue  for a  period of  no  less than  three months.
Based  on  findings  from  those  samplings,   subsequent  sampling  may  be
indicated.
     3.  A  less  severe  treatment  with Prudhoe Bay crude oil should be used
to bracket potential  effects on recovery.  The reduction in severity should
relate to the  method  of applying oil  and not  necessarily  the total amount
used which, in  the present case,  was slight.   Thus,  an  approach where oil
is  layered  onto the  surface  of sediments,  either from  a  seawater surface
slick or direct surface application would be  appropriate.

     4.  Comparative  studies  of  the  effects  of  recovery  from processed
petroleum  products,  i.e.,  light  fuel  oils  and  residual  fuels should  be
undertaken  in  areas of  the north Puget  Sound  region especially vulnerable
to such spillage.

     5.  The  relative  severity on  recovery  effects  following oil  and oil
dispersant  combinations should be investigated using the  present  approach
to  assist  decision making regarding the application  of  dispersants should
spillage occur in our region.

     The methods used in this study appear particularly suitable for inves-
tigation of other sediment-related pollutant  problems  in  the Puget Sound
region. Specifically,  we recommend studies of the effects on recovery from:

       1.   Heavy metal contamination
       2.   Synthetic  organic contamination
       3.   Dredge spoil  contamination
       4.   Wood fiber and by-product contamination
       5.   Combinations of the above

                                   10

-------
Since  the  methodological  sensitivities  are  a  function of  the prevailing
seed  populations  and  the types  of  physical  factors  identified  in  this
study, there is good reason to believe that the approach used here would be
appropriate for these studies.

     The present  studies have  clearly identified an  urgent need  for two
further  types  of  investigation to  assist  in  the  interpretation of the
effects demonstrated:

     1.  Basic life  history studies  of especially  oil-sensitive primary
species  to  include studies  of  behavior in  response  to pollutant contami-
nation.  Two  questions need  to  be  addressed:   (1) what is  the zonal  dis-
tribution  of  the  seed  source;  and,  (2)  where do  organisms  go  that are
absent from oil-treated sediments?

     2.  Experimental  studies  of  feeding  relationships,  particularly  of
bottom-feeding   flatfishes.     The   experiments   should   have   a   field
orientation.
                                   11

-------
                                 SECTION 4

                           MATERIALS AND METHODS

STUDY SITES

     The  experimental  recovery studies  used the  four  sites identified on
Figure 1.   The  principal  site is near the mouth of Sequim Bay, Washington.
The  natural  substrate at  this site is  coarse  sand,  sparsely interspersed
with cobble.  The  beach  is east facing, well  protected from northwesterly
winds and ocean swell,  and  somewhat protected  from  the prevailing south-
easterly  winds  by  Travis Spit  (Figure 1).   Historically,  the  beach had
served as a commercial source for both littleneck (Protothaca staminea) and
butter (Saxidomus  giganteus)  clams.   For the past  12 years, the beach has
been  owned by  Battelle-Northwest and  harvest  of  clams has  been  largely
prohibited.  It is  now characterized by moderate populations of relatively
large individual clams (Vanderhorst and Wilkinson, 1979).  Parts of Tasks A
through E  and Tasks G and I were  conducted  at this site.  The site served
as  a  source  for  sand substrate  identified  as "coarse"  sand which was
evaluated  ijn  situ  and also  transposed  in  Task  B  to a  second  site  on
Protection  Island  (Figure  1).    The  Protection  Island  site  is  on  a
south-facing beach  of  fine sand.   The beach is  well  protected from north-
westerly winds  and  ocean  swell,  but was highly exposed to southerly winds.
The  fine  sand beach was  mobile during winter  months  and was used only for
Task  B  during  a   late  summer-early  fall  recruitment  period.   This  site
served as  a source for "fine" sand, evaluated jn  situ  and also transposed
to the Sequim Bay site for use in Tasks B and C.

     A  north  facing  beach at  Rocky  Point  (Figure 1)  served as  a  third
experimental site.   The natural  substrate  at this site consisted mainly of
cobble,  with a  mature,  highly diverse, exposed rocky shore community.   The
beach is exposed to northwesterly storm winds.   It was chosen as a compara-
tive site  for Task G during peak  spring and summer recruitment periods to
allow evaluation  of  site effects  on availability  of  rocky  shore larval
forms  at  Sequim   Bay.   Finally,  a  fourth  experimental  site  was on  an
actively managed commercial clam  bed (Figure 1).  The  beach at Carr Point
is  southeast  facing  and   subject  to exposure  from southerly  winds.   The
severity of exposure  to  these winds is less than for the Protection Island
site because  of much reduced  fetch.   Substrate  at this beach  is a mix of
fine and coarse  sand and gravel.


EXPERIMENTAL APPROACHES

     An  experimental  approach was adopted  to  investigate the  effects  of
Prudhoe Bay crude  oil  on  recovery of intertidal  fauna  principally  because
controlled experimentation seems to  us  the only practical way to discrimi-
nate between effects  on  fauna!  recovery due to  oil  and effects from other
                                   12

-------
factors.  The use of experiments in which controlled treatments are applied
to  the  substrate  itself permits a chemical  characterization  of the treat-
ment.   In the experiments reported here the substrates have been arrayed in
such  a manner  to  evaluate  the effects  from certain  other  environmental
variables  (e.g.,  tide  level,  site,  season,  substrate  type)  on a specific
aspect  of  recovery.   In studies of "natural" assemblages these factors are
invariably  confounded  with  effects  of a  contaminant of  interest because
true  replication  is  unattainable  in  nature.   The  specific experimental
approaches  used  in  these studies  also circumvent the problem  of depen-
dencies in time series  observations of the same biological material.  These
dependencies are inherent in any study of "natural" populations and violate
the  most  basic assumptions  of statistical  analysis  if estimates  of  true
error probabilities are needed.  In the present studies, time series obser-
vations  are  circumvented by   having  a  high  replication of  "blank"  or
organism-free  substrates  within  each  treatment  category, and  having the
field exposure  time  fixed and equal  for all treatment categories within an
experiment.   Our  approach makes the  assumption  that  the source organisms
are  equally  available  to  all  substrates (treated  and  untreated)  and  uses
the  high  replication  in  untreated substrates   to  test  that  assumption.
Thus,  differences  in the  mean kinds and  densities of  organisms  in or on
substrates at the conclusion  of  an experiment  should  relate  only to the
treatments applied.

     The experimental approach used  is  not without shortcomings.  Since it
is  based  on  the  assumption  that the  source  and condition  of organisms
available  to  colonize  substrates  are  unaffected  by  the  treatment  (a
function  of   the  treated   substrate  and not  surrounding substrate or the
water mass),  the  approach does not address perhaps  important questions of
recovery associated with that availability.   The experimental  approach  used
in  these  studies does  not allow  us  to discriminate whether  the  source of
organisms which colonize  substrates  is  the water mass  or surrounding  sub-
strate.   There  is  some  recent evidence  (Santos and  Simon,  1980)  which
suggests that trays  placed on the bottom are  colonized by larger, perhaps
adult,  organisms,  as  compared  to  trays of  sediment  suspended in the water
column.  These sorts of questions  do not interfere with the validity of the
experimental  approach we  have  used but they do have a bearing on the over-
all question of rate of recovery following an actual oil spillage.

     The approach in Tasks A through E  and Task I used trays of sediment as
experimental  units.   For  Tasks G  and H, concrete construction  bricks  were
used  for  a similar purpose.    In all cases  (with the exception  of Task H)
the  units  were:   (1)  initially free  of organisms;  (2) treated  with oil
(treated) or  not (control);   (3)  returned  to the intertidal  zone;  and (4)
allowed to  colonize  in a  selected array  of  natural   intertidal  habitat
conditions.    The  power of this approach  lies in:  (1) an equal  starting
point (organism free) for  treated  and  untreated substrates; (2) the use of
strict  random procedures  (as  opposed  to haphazard) for  allocation of  sub-
strate  units within  and between treatments; (3)  the inclusion of important
organism-controlling variables  (site,  tide level,  substrate  type, season)
as  treatment  categories (receiving both oil-treated and untreated units);
and (4) the use of a replication scheme in  the  case of Tasks  A  through  E


                                   13

-------
and  I  for  which methodological  sensitivity had  been  preevaluated (Vander-
horst et al.,  1978).

     Balanced  experimental  designs were  used in  all  the experiments  re-
ported  here,  and  independent controls  were   used  in  each  phase of  each
experiment;  thus,  given  the  inherent  assumptions  of normality,  common
variance,  and  additivity  of  the  statistical  model,   a  correct  use  of
analysis of variance  is indicated.   This is in sharp contrast to the  use of
analysis of  variance in field  surveys for descriptive purposes  in  which
time series data create dependencies between treatment categories and alter
error probabilities in an undefinable manner.


INFAUNAL STUDIES

     Details  of  the  experimental  methods  including   criteria  for  site
selection,  preparation and placement of substrates, chemical  and biological
characterization,  and sampling  rationale  and procedures  have  been  pre-
viously reported for  the infaunal  studies (Vanderhorst et al., 1979;  1980).
In  summary,  native substrate was  collected  from  three sites  (Figure  1),
brought to the  laboratory,  and  given a repeated freezing-thawing treatment
to  kill macrofauna.   Half  of the  substrate from each site was treated with
Prudhoe Bay  crude  oil by  mixing   in  a commercial  cement  mixer.   A  target
concentration of 2,000  ppm total  oil was  sought in Tasks A, B,  C,  and I,
and a target concentration of 1,000 ppm total  oil was sought in Tasks D and
E.  Total amounts of  oil  and concentrations of selected petroleum compounds
were measured  in treated  and untreated sediments  prior to  field  instal-
lation,  at intervals  between installation and  completion,  and  upon  com-
pletion of a  given experiment.   The other  half  of  the  substrate from each
site served as control.   It received the mixing just as  the oiled substrate
but  did not receive  an  application of oil.   The  prepared substrates  were
placed  in  PVC  trays  (30  x 15 x 15 cm).   The  bottom of the trays were pro-
vided  with  eight 2.5  cm  diameter  holes  for   drainage.    Experimental  sub-
strates were  retained in  trays by placing a  fiberglass  screen over these
holes.   For Tasks  A,  B,  C, and E,  trays were  buried with top surface flush
with the ground surface  at MLLW at each  of the sites.   For  Task D,  trays
were buried in  a similar fashion  at -2' below MLLW at the Sequim Bay site.
For  Task I  burial  of trays was at  MLLW and +2'  above MLLW.   Field instal-
lations for Tasks A,  B, and C were during August 1978.   Field installations
for  Tasks  D and E were  during  April  1979.  Task B terminated  in November
1978.   Tasks  D  and   E  terminated  in August  1979.   Task I  was initiated
during May 1980 and terminated late July 1980.  Tasks A and C terminated in
November 1979.

     A  relatively high amount of replication was used in both the placement
and  sampling  of  substrate  units   (Tables  1   and  2 list  sampling dates).
Based  on a predesign study using similar units (Vanderhorst et al.,  1978),
this  replication was  done to permit evaluation of  methodological  sensi-
tivity  and to  attain  a  quantitative measure  of  the  density of  individual
species  colonizing the trays.  Because of interest  in a  large number of
species and a limited number of independent units in even this rather large


                                    14

-------


-------
design, it was necessary to a priori select species of special interest for
hypothesis testing with valid probability statements concerning statistical
error.  For  these  species  we use a criterion of a ^ 0.01 to deem "signifi-
cant" effect on density.  The maximum real probability for Type I error for
any one of  the seven species (Task B)  was  7%,  and for  any  one  of the ten
species (C,  D, E, I) was  10%.   Task A data were  outside the experimental
framework.  To meet  the objectives of Tasks B,  C,  D,  E, and I,  four inde-
pendent experiments were conducted and a priori  selection of target species
was made.   These  species  were designated "Primary" species  and  consist of
the following:

For Experiment I

          (Task B):      Mollusks
                              Mysella tumida
                              Transennella tantilla
                              Lacuna sp.

                         Polychaetes
                              Platynereis bicanaliculata
                              Armandia brevis
                              Ophiodromus pugettensis
                              Capitella capitata

     For Experiment II
          (Tasks D and E)  and

     For Experiment III
          (Task C): and

     For Experiment IV
          (Task I):      Mollusks
                              Mysella tumida
                              Protothaca staminea
                              Lacuna variegata

                         Polychaetes
                              Platynereis bicanaliculata
                              Armandia brevis
                              Polydora social is
                              Exogone lourei

                         Crustaceans
                              Leptochelia dubia
                              Corophium ascherusicum
                              Photis brevipes

     The basis  for a priori  species  selection  for Task B rested  on:   (1)
results from the  predesign  study  (Vanderhorst  et al., 1978); and  (2)  for
two  species,  Lacuna  sp.   and Capitella  capitata,  reported  perturbations
following  oil  spills  elsewhere.   The basis for a priori selection of


                                   16

-------
Table 1.  Schedule of sampling for oil recovery experiments.
MONTH/YEAR
August/1978







September/1978






October/1978







November/1978





SITE/STATUS
TIDE LEVEL ( ) TASK
Preliminary ABC







Protection Is. A
(O1)


Sequim Bay A
(O1)

Protection Is. A
(O1)


Sequim Bay A
(O1)


Protection Is. B
(O1)




TREATMENT
STATUS
Oiled



Unoiled



Oiled

Unoiled

Oiled

Unoiled
Oiled

Unoiled

Oiled

Unoiled

Oiled





SUBSTRATE
TYPE
Coarse

Fine

Coarse

Fine

Fine

Fine

Coarse

Coarse
Fine

Fine

Coarse

Coarse

Coarse


Fine


SAMPLE
TYPE
Infrared
Gas chromat.
Infrared
Gas chromat.
Infrared
Gas chromat.
Infrared
Gas chromat.
Infrared
Biological
Infrared
Biological
Infrared
Biological
Infrared
Biological
Infrared
Biological
Infrared
Biological
Infrared
Biological
Infrared
Biological
Infrared
Gas chromat.
Biological
Infrared
Gas chromat.
Biological
NUMBER
TRAYS
3
3
3
3
3
3
3
3
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
3
3
5
3
3
5
NUMBER
CORES
9
3
9
3
9
3
9
3
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
9
3
35
9
3
35

-------
     Table 1.  (Continued)
00
MONTH/YEAR
November/1978





November/1978











December/1978







January/1979







SITE/STATUS
TIDE LEVEL ( ) TASK
Protection Is. B
(O1)




Sequim Bay B
(O1)










Protection Is. A
(O1)


Sequim Bay A
(O1)


Protection Is. A
(O1)


Sequim Bay A
(O1)


TREATMENT
STATUS
Unoiled





Oiled





Unoiled





Oiled

Unoiled

Oiled

Unoiled

Oiled

Unoiled

Oiled

Unoiled

SUBSTRATE
TYPE
Coarse


Fine


Coarse


Fine


Coarse


Fine


Fine

Fine

Coarse

Coarse

Fine

Fine

Coarse

Coarse

SAMPLE
TYPE
Infrared
Gas chromat.
Biological
Infrared
Gas chromat.
Biological
Infrared
Gas chromat.
Biological
Infrared
Gas chromat.
Biological
Infrared
Gas chromat.
Biological
Infrared
Gas chromat.
Biological
Infrared
Biological
Infrared
Biological
Infrared
Biological
Infrared
Biological
Infrared
Biological
Infrared
Biological
Infrared
Biological
Infrared
Biological
NUMBER
TRAYS
3
3
5
3
3
5
3
3
5
3
3
5
3
3
5
3
3
5
1
1
1
1
1
1
1
1
1
1
I
1
1
1
1
1
NUMBER
CORES
9
3
35
9
3
35
9
3
35
9
3
35
9
3
35
9
3
35
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1

-------
Table 1.  (Continued)
MONTH/YEAR
April 71979

April /1979

June/1979











July/1979











SITE/STATUS
TIDE LEVEL ( )
Sequim Bay
(O1)
Preliminary

Sequim Bay
(O1)
Sequim Bay
(O1)
Sequim Bay
(O1)
Sequim Bay
(O1)
Sequim Bay
(-21)
Sequim Bay
(-21)
Sequim Bay
(O1)
Sequim Bay
(O1)
Sequim Bay
(O1)
Sequim Bay
(O1)
Sequim Bay
(-21)
Sequim Bay
(-21)
TASK
A

D,E

A

D(A)

A

D(A)

E(A)

E(A)

A

D(A)

A

D(A)

D(A)

E(A)

TREATMENT
STATUS
Oiled

Oiled

Oiled

Oiled

Unoiled

Unoiled

Oiled

Unoiled

Oiled-

Oiled

Unoiled

Unoiled

Oiled

Unoiled

SUBSTRATE
TYPE
Coarse

Coarse

Coarse

Coarse

Coarse

Coarse

Coarse

Coarse

Coarse

Coarse

Coarse

Coarse

Coarse

Coarse

SAMPLE
TYPE
Infrared
Biological
Infrared
Gas chromat.
Infrared
Biological
Infrared
Biological
Infrared
Biological
Infrared
Biological
Infrared
Biological
Infrared
Biological
Infrared
Biological
Infrared
Biological
Infrared
Biological
Infrared
Biological
Infrared
Biological
Infrared
Biological
NUMBER
TRAYS
I
I
3
3
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
NUMBER
CORES
1
1
9
3
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1

-------
     Table 1.   (Continued)
ro
o
MONTH/YEAR
August/1979



August/1979





August/1979





September/1979



October/1979



November/1979






SITE/STATUS
TIDE LEVEL ( )
Sequim Bay
(O1)


Sequim Bay
(O1)




Sequim Bay
(-21)




Sequim Bay
(O1)


Sequim Bay
(O1)


Sequim Bay
(O1)





TREATMENT
TASK STATUS
A Oiled

Unoiled

D Oiled


Unoiled


E Oiled


Unoiled


A Oiled

Unoiled

A Oiled

Unoiled

C Oiled


Unoiled



SUBSTRATE
TYPE
Coarse

Coarse

Coarse


Coarse


Coarse


Coarse


Coarse

Coarse

Coarse

Coarse

Coarse


Coarse



SAMPLE
TYPE
Infrared
Biological
Infrared
Biological
Infrared
Gas chromat.
Biological
Infrared
Gas chromat.
Biological
Infrared
Gas chromat.
Biological
Infrared
Gas chromat.
Biological
Infrared
Biological
Infrared
Biological
Infrared
Biological
Infrared
Biological
Infrared
Gas chromat.
Biological
Infrared
Gas chromat.
Biological

NUMBER
TRAYS
1
1
1
1
3
3
5
3
3
5
3
3
5
3
3
5
1
1
1
1
1
1
1
1
3
3
5
3
3
5
254~
NUMBER
CORES
1
1
1
1
9
3
35
9
3
35
9
3
35
9
3
35
1
1
1
1
1
1
1
1
9
3
35
9
3
35
^788"

-------
 species  in  Tasks C, D,  E  and I were:  (1)  results  of Task B studies;  and
 (2)  examination  of baseline data  for nearby stations  (Nyblade,  1979).   The
 danger in a priori  selection  in any experiment  involving  field colonization
 is  that  selected species may not  occur in the future  experiment  or may be
 relatively unimportant constituents.   In general,  this  was  not the case  for
 the  species selected here.

      In  addition to the 13 species which  were  a priori selected to protect
 the  validity  of error probability statements,  quantitative data  were also
 collected  on  the  density  of more than 200 other species which  colonized
 trays.   Analyses of variance were computed  for these  data for  descriptive
 purposes.

 Sediment Extraction and  Chemistry

      Sediment cores collected for  chemical analysis were  frozen  immediately
 after collection.   The  frozen samples  were  thawed at  room temperature  and
 thoroughly  mixed  for  subsampling.   Twenty  grams  (wet)   of  sediment were
 placed  in a  250  ml  teflon-capped  bottle  with  20 g  of anhydrous sodium
 sulfate  and  thoroughly  mixed  to  hydrate  the water  from  the sediment.   To
 the  samples  for  I.R.  analysis, 50 ml  carbon  tetrachloride was  added,  and
 for  capillary G.C.  analysis,  25  ml   hexane was  added.   The  bottles were
 shaken overnight on a reciprocal  shaker.  The  solvents were decanted from
 the  sediment.   Carbon tetrachloride  was  extracted  into  a 25  ml   scintil-
 lation vial, and hexane  was extracted  into a 50 ml graduated cylinder.   The
 CC14  subsample was analyzed by  infrared spectroscopy (Simard et  a!., 1952).
 The  samples  for G.C.  analysis  were  extracted  for an  additional  two  hours
 with  25  ml  hexane and decanted into  the  graduated cylinder with  the  first
 extraction.   The sediment  sample  was then  extracted with  5  ml  volumes  of
 hexane and decanted until  the total  extract volume was 50 ml.   Twenty-five
 ml  of the  hexane extract were  concentrated  under  nitrogen to three ml  and
 separated  into  aliphatic   and  aromatic  fractions by  silica gel chroma-
 tography (Warner,  1976).   These fractions were concentrated  to  one ml  and
 analyzed  for  individual  hydrocarbons  by  capillary   gas  chromatography.
 Concentrations  were calculated by  using standard  addition and   internal
 standard compounds  with  comparison to  authentic standards.  Concentrations
 of  individual  aliphatic and  individual aromatic  compounds  were  summed  to
 represent the respective groups.

 Sediment Grain Size Analysis

      Following chemical  analyses  of  preliminary samples,  the remainder  of
 core  samples   for  these analyses  were  retained  in  a  frozen  condition.
 Twelve replicate cores  from Sequim Bay, 10 replicate cores from Protection
 Island,   and   11  replicate  cores  from  Discovery  Bay were  analyzed  for
 sediment  grain   size.   Frozen  cores  were thawed  at room  temperature  and
 dried in an oven at 100°C for 48 hours.  Following drying, individual cores
were  sorted to classes  of  particle size as  follows.   A series of  standard
 sieves,  mesh  sizes:  5.66  mm,  2.0 mm,  1.0 mm,  500 urn, 125 urn,  and 63  urn,
were  stacked  on an  Eberbach  shaker.   Sediment  from a  single  core was
 emptied  into  the top sieve.  The  shaker  was activated for a period  of  10


                                    21

-------
minutes.  The sediment retained on each sieve and in a pan placed below the
finest sieve, was weighed on an analytical balance.   The weight of sediment
in each  of the seven size  classes  for each individual core was  then com-
puted as a percentage of the total weight of sediment in that core.


EPIFAUNAL RECOVERY STUDIES

     Concrete construction  bricks  (19.5  x 5.5 x 9 cm) were used as  experi-
mental substrates for studies  of epifaunal recovery.  The bricks have been
shown to be suitable for colonization by a variety of typical  rock epifauna
in previous  studies  (Vanderhorst et al.,  1975;  Vanderhorst  and Wilkinson,
1977).   They have  two  additional  advantages  for  the present  studies  in
which  more  than  2,000  individual  substrate   units  were evaluated.   The
bricks were  readily  available  in large quantities with good  uniformity in
size,  shape,  and porosity.  Also,  they were amenable to  placement  on the
beach without the need for a physical support system to keep them in place.

     The experimental method involved four steps:   (1) preconditioning of
concrete construction bricks by  placement in a laboratory flowing-seawater
system for two  weeks;  (2) treatment of one-half of conditioned bricks with
Prudhoe Bay  crude oil;  the other half served  as controls; (3) characteri-
zation of  treatment  severity by  extraction of oil from bricks and chemical
analyses;  and  (4) field  evaluation of oil content  and biological  coloni-
zation  in  month-long experiments.   Independent experiments  were conducted
each  month commencing with October 1979  through  July 1980.   Additionally,
two identical experiments during June and July 1980 were conducted at Rocky
Point to test the effect at that site.

     Preconditioning of bricks  in laboratory flowing sea water was to leach
out any  foreign materials which  might have been  associated  with the manu-
facture  of the bricks  and to allow some chance  for colonization  with  a
microflora.  Prior  to the  treatment phase, all  bricks,   both  control  and
treated,  received a  thorough  washing  with a  high-pressure hose  both to
remove unseen but possible incidental occurrences of settled macroorganisms
and  to aid  in  equalizing the  effect of  preconditioning  on  control  versus
treated  bricks.   For treated  bricks,  treatment  lasted five  days.   During
this  time  the  control  bricks remained in the preconditioning tank.   At the
end of  the treatment phase, the control  bricks again received a wash with
the high-pressure hose.

     The treatment  of bricks  with oil was  designed to  simulate repeated
exposure of  intertidal  rock during shifts of  the  tide.   The procedure was
as follows.  Conditioned  bricks  were placed on the bottom of a rectangular
tank  (0.54  x  4.88  m).    The  tank  was  then  filled  with  sea water  (25 cm
depth).  Prudhoe  Bay crude oil (20 £) was poured on the surface of the sea
water  to provide a  slick thickness of  about  1  cm.   Continuous  inflow of
clean  sea  water was  provided (two  £/min).  Both the inflow and outflow of
the sea  water  were  subsurface  to prevent disturbing the  surface slick and
to  retain  it  in the tank.   Twice  each  day  the  inflow  of sea  water was
discontinued and the seawater level reduced to  a depth of  2 cm.  The period


                                    22

-------
of  this  reduced water  level  was  for two hours at  each treatment.   During
the periods  of  low water, the slick was in contact with the upper and side
surfaces of the bricks.  The twice-daily regime was repeated for five days.
At  the end  of five days, a surface outflow was provided, and the slick was
washed  into  the  laboratory  oil-treatment  facility.    Clean  seawater flow
(two £/min)  was  provided for a further  24  hours.   Bricks were then placed
in the intertidal zone for one month.

Chemical  Characterization of Bricks

     Routine  chemical  characterization of the treatment severity  has been
based  on  five   brick  subsamples  (see  Table  2  for  schedule).   Analysis
methods  for  both   infrared   spectrophotometry  and  capillary  gas  chroma-
tography  follow those  previously  reported  (Vanderhorst  et  al.,  1979).
Three  types  of extraction  procedures  have  been  evaluated  and  two  are
routinely used.   The first procedure involved washing whole wet bricks with
500 ml  CC14.  The  extraction efficiency was  poor.   The  second  procedure
involved air  drying bricks  for a period of 48 hours before extraction with
500 ml CC14.   This improved the extraction  efficiency approximately 100%.
To  provide  a  better  measure  of the  amount of  oil  actually "seen"  by
colonizing organisms, a two-part extraction procedure was adopted.   In this
procedure  the top  surface  of  bricks  was  washed with 200 ml CC14.   The
amount of oil in this extract was measured.   The bricks were then air-dried
for 48 hours and re-extracted with 500 ml CC14.   The amount of oil  in this
extract was measured.  Diluted samples of the extracts measured by infrared
spectrophotometry  are  computed  in terms of numbers  of grams of oil  per
brick.    Samples  analyzed by  capillary  gas  chrpmatography  are reported in
terms of number of milligrams per brick for individual compounds.

Biological  Characterization of Bricks

     For days one  through five following placement of bricks at the appro-
priate tide level in the intertidal zone, observations of all bricks during
extreme  low  tides  were  made  to  determine  the  presence or  absence and,
insofar  as  possible,  kinds  of organisms  colonizing bricks.   Thirty days
after field  placement  all  organisms  were scraped from the top of 15 bricks
in  each  treatment  category  (total  60 bricks).   Animal species were  identi-
fied  and  counted.   A  subsample  of  five   bricks  was  used  for  chemical
analyses  described  above.


GRAZER MANIPULATION STUDIES

     Three   experiments were  conducted to examine effects  from Prudhoe  Bay
crude oil,   grazer  manipulation,  and  tide level on epifauna and flora.  The
bricks  used  for  these  studies were randomly selected  from  a large  pool of
bricks  colonized at  MLLW and  +2'  above MLLW for  nine months preceding the
experiments (September 1979 through  May 1980).   The bricks were subdivided
into eight  treatment categories as follows:
                                   23

-------
     Table  2.   Preparation  of  Units and Sampling Schedule for  Experiments  on  Effects  of  oil  on
                Recovery  of  Commercial  Clams and Epifauna on  Rocky  Intertidal.1
     TASK/SITE
DATES
UNIT TYPE
PRELIMINARY   MLLW TIDE   +2 MLLW TIDE
ro
TOTALS
A/ Discovery Bay 5/80





6/80





7/80




8/80








Infrared
Trays 6
Cores 18
Capillary GC
Trays 6
Cores 6
Infrared
Trays
Cores
Biological
Trays
Cores
Infrared
Trays
Cores
Biological
Trays
Infrared
Trays
Cores
Capillary GC
Trays
Cores
Biological
Trays
Cores







2
. 2

2
2

2
2

2

6
18

6
6

10
70







2
2

2
2

2
2

2

6
18

6
6

10
70

6
18

6
6

4
4

4
4

4
4

4

12
36

12
12

20
140
                                Note:  Core profile on capillary GC and Biological Cores means that
                                discrete samples were 2X indicated number for 8/80 sampling.

-------
    Table  2.  (Continued)
     TASK/SITE
DATES
UNIT TYPE
PRELIMINARY    MLLW TIDE    +2 MLLW TIDE
TOTALS
B/Sequim Bay 9/79 Concrete
Infrared
Capillary GC
Biological2

10
10


10
10
30

10
10
30

30
30
60
     B/Rocky  Point
10/79 - 8/80


 5/80
The pattern for Task B was followed each month with exception of
Capillary GC, giving totals as follows:

Concrete
ro
en
Infrared
Biological2
6/80 Concrete
Infrared
Biological2
C/Sequim Bay 4/80 Concrete
Infrared
Capillary GC
Biological2
5/80 Concrete
Infrared
Biological
10


10


10
4
120



10
30

10
30

10
4
60

10
30
10
30

10
30

10
4
60

10
30
30
60

30
60

30
12
240

20
60
     1  Experiment  balanced,  i.e.,  for every treated sample a control  sample  is  also  indicated.
     2  Each  of  the biological  units with this notation  (with the exception of preliminary)  should  be
       multiplied  by 5  for  5 daily observations within  the month.  Appropriate  total  unit samples  are:
       Capillary GC, 60;  Infrared, 532; Biological, 3328.

-------
     OILED BRICKS

          MEAN LOWER LOW WATER

               (1)  Limpets Stocked
               (2)  Limpets Removed

          PLUS TWO FEET ABOVE MEAN LOWER LOW WATER

               (3)  Limpets Stocked
               (4)  Limpets Removed

     UNOILED BRICKS

          MEAN LOWER LOW WATER

               (5)  Limpets Stocked
               (6)  Limpets Removed

          PLUS TWO FEET ABOVE MEAN LOWER LOW WATER

               (7)  Limpets Stocked
               (8)  Limpets Removed

Bricks  receiving  oil  treatment  were  treated  exactly as  bricks  in  the
monthly  epifaunal  experiments described above.   The nonoiled  bricks were
placed on a  water table  and received a continuous flow of  seawater during
the  five-day  oil  treatment  period.   The  tide  level  treatment  involved
placing  bricks in  the intertidal  zone at  the  indicated tide  level  im-
mediately after completion  of the oil treatment  phase.   The  grazer treat-
ment involved  placing  ten  limpets (Acmaea spp.), with an approximate shell
diameter of two cm, on each brick in the grazer stock category and removing
all visible  limpets  from  the grazer removal  category.   This  process  was
repeated several  times each  day  throughout the  five-day treatment phase.
Stocked  limpets in the oil-treated portion of the experiment were all dead
after completion  of two  treatment cycles.   Stocked  limpets on  the unoiled
bricks tended  to  wander  off bricks and onto the  sides  of the water table.
In  no  case  did  we observe limpets  on  the  "limpets-removed"  category  of
substrates.   Grazer treatment ended  simultaneously  with oil  treatment  at
the time of field placement.

     A subsample  of ten  bricks from the oil-treated categories was evalu-
ated for total oil content at the completion of the oil  treatment phase.
The plant and  animal  material was scraped from  five of these bricks prior
to oil extraction  and  analysis.   The other five  bricks  were  extracted and
analyzed with organisms intact.

     Three independent experiments were based on samples of five bricks for
each treatment category (total 40 bricks) at:   (1) immediately after treat-
ment;  (2) five days after  field placement; (3)  30  days after field place-
ment.   Likewise,  subsamples of bricks were taken at each of  the intervals
for analysis for total  oil.

                                   26

-------
     The biological characterization  of  bricks  in the experiments was done
by completely  removing  living material by scraping.   Animals  were identi-
fied and  enumerated by  species.   Plants were  evaluated in aggregate,  by
brick,  in terms of dry weight biomass.  Plant material was oven-dried until
asymptotic weight was reached.
                                   27

-------
                                 SECTION 5

                                  RESULTS

RECOVERY ON HARD SUBSTRATES

A Perspective

     The dominant species  on  rock substrates are long-lived as compared to
the  infauna  of sand  and mud.   This  fact has  prompted Nyblade  (1979)  to
assign  relative  recovery  times  measured  in decades  for  rock  habitat  as
compared to  a few years for the  infauna communities.  Coupled with this
relatively long-term  biological  recovery  (possibly  involving successional
processes), rock, as  compared to mud and sand,  presents a  relatively minor
surface area  for accumulation and  retention of  spilled petroleum.   These
two factors in concert  tend to suggest that  the  primary effects of oil  on
recovery of  these  communities  will  be  the  degree  to which  the  complex
biological  communities  are broken  down  by the  effect  of  oiling.   A third
characteristic of the communities  of this habitat is  that the long-lived,
dominant species  tend to have infrequent  successful  recruitment.   Periods
between natural successful  recruitments  may  be  from three  to seven years.
Because of  this  factor,  even a  short-term  impairment of  substrate suit-
ability for settlement may have far-reaching effects on direction of future
recovery.   The  12 short-term experiments  in Task G were  designed  to test
the effect  of oiling organism-free hard substrates on the  suitability  of
those  substrates  for successful   settlement  and  survival.    Thus,  the data
provided here  do not directly address  the actual recovery  times  for com-
munities on hard substrates.  Rather, they establish effect of oiling on an
important first step in the recovery process.   The three short-term experi-
ments in Task H allowed for contaminant-free colonization  of substrates and
subsequent measurement of effects from oiling on the complexity of existing
communities.   Although these  experimental  communities  were quite simple as
compared  to   mature   rocky  shore  communities,   the  experimental  approach
nevertheless  provides  an  opportunity  to  determine  whether  or   not  oil
applied to these  communities is  selective  in  effect for  community com-
ponents and to make inferences concerning subsequent recovery.

Hard Substrate Recovery - Biological Data Presentation

     Two groups  of  experiments  conducted  to examine  the effects  of  oil
treatment  on  recovery   of  epifauna  on  hard   substrates   are  treated
separately.   The  first   group consisted  of 10 experiments,  one each month
during  the period  of  October  1979  through July 1980.    These were  all
carried out  with Sequim  Bay  field exposure  and are  designated "monthly"
experiments.   One  of the  primary  objectives of  these experiments  was  to
evaluate differences  in  recovery  related to season of field exposure.  The
other group consisted of two experiments, designated as "site" experiments.
These  were  conducted at  Rocky  Point during  the months of  June and July,


                                    28


-------
1980.   The  methods for  these two
"monthly" experiments  at Sequim
parisons between the two sites of
 site" experiments.
     Monthly Experiments.  Each
the  others  in the  group,  i.e.,
                                    experiments  exactly paralleled  the two
                                  Bay for  these  two months  only,  and corn-
                                  field exposure are made in the section on
                                of  these  10 experiments was independent of
                                 each was set  up  using preconditioned, but
uncolonized bricks.   Sixty  bricks,  15 in each of four treatment categories
(1 = oil-treated  MLLW;  2 = oil-treated +2'  above MLLW; 3 = untreated MLLW;
4 = untreated +2'  above  MLLW)  were used for biological analyses in each of
the 10  experiments.   Differing total  numbers of bricks  (70 or 80 bricks)
were used  in  individual  experiments depending on the schedule  for chemical
analysis.
     The first biological data coll
                                   ection in each of the 10 monthly experi-
ments involved examination  of  each of the  individual  bricks  once each day
during periods  of extreme  low tide  for  the first five days  of field ex-
posure.   The goal  of  these observations was to determine the earliest time
at which  marine  larvae  settled  on bricks.   The results  from these 3,000
individual brick examinations  (over the 10 experiments) consisted of a few
random occurrences of  mobile  adult organisms.  In no  case was an organism
which could  be  identified as  an  attached marine  larva  observed  on any
brick.  A  tabulation  of  calendar months  in  which  the  different kinds  of
organisms were observed is given below:
       TAXA
                                           CALENDAR MONTH
                                      1979                1980
     POLYCHAETES

          Polychaetes undet.

     CRUSTACEANS

          Amphipods  undet.
          Exosphaeroma amplicauda
          Exosphaeroma sp.
          Hemigrapsus 'sp.
          Hermit crab

     MOLLUSKS

          Lacuma sp.
          Limpet
          Mopalia lignosa
                                      10,  11
                                      12
                                      9,  10
1, 2, 3, 4
                                                        2,
                                                        3,  4
                                                        1,  5,  6,  7
                                                        3,  5
                                                        3
1, 2, 6, 7
2, 3, 4, 5
                                   29

-------
The  remainder  of biological  data for the  10  monthly experiments resulted
from  scraping  all fauna from  each  of the 60 bricks  in  each experiment at
the end of 30 days of field exposure (total = 600 bricks).

     Since the  monthly  experiments  were all independent, we deemed it most
appropriate  to  use a  single  analysis of variance  covering  all  10 experi-
ments for  each  response variable of interest.   The response variables were
evaluated in the experimental  design model:
     where:    Y = the response variable magnitude;

               M = the main effect on response variable magnitude
                   due to month of experiment (i = (1 = October 1979
                   through 10 = July 1980));

               0 = the main effect on response variable magnitude
                   due to oil treatment (j = (1 = oil -treated;
                   2 = untreated));

               T = the main effect on response variable magnitude
                   due to tide level of field exposure (k = (1 =
                   MLLW or 2 = +2'  above MLLW)); and

               E = random error.

The  interactions  between main effects  on  response  variable  magnitude are
also of interest and have been computed.  The tests for statistical signifi-
cance  concerning  interaction means  are,  however,  not  included  in  this
report  because  of  interpretive  difficulty  concerning the  interactions.

     The  response  variables  evaluated (Y's) in the  model  included numbers
of  individuals  within   taxonomic  and  trophic  categories  and numbers  of
specific  entities  within  taxonomic  categories.   For convenience  sake  we
refer  to  specific entities  as  "species" throughout this  report.   In fact
many of  the specific entities  are undetermined species within a genus  or
higher  taxa,  and  in  a  few  cases,  specific  entities  are  fragments.
Additionally, the number of specific entities per brick was estimated using
the model.   Hypotheses were  tested for non-zero differences  due  to month,
oil treatment,  and tide level.

     The  composition,  trophic classification,  and  an  abbreviated analysis
of variance for individual numbers  of mollusks are shown in Table 3.  There
were 20   species  of mollusks  identified  from  bricks  in the  ten monthly
experiments.  Of these,  9 (45%),  were herbivores,  7 (35%) were suspension-
feeders,   3  (15%) were  carnivores,  and 1  (5%) was  a parasite.  The trophic
classification  for  the  species comes from  Simenstad  et al.   (1979).   The
composition  indicated on  Table  3  includes many species  which are normally
associated  with  rock habitat  including several species  of grazing snails
and  limpets,  and the suspension- feeding  dominant mussel,  Myti 1 us edulis.


                                   30

-------
Table 3.  Species composition and analysis of variance for mollusk
          density in monthly hard substrate recovery experiments
          at Sequim Bay.
                               COMPOSITION
         SPECIES                                TROPHIC CATEGORY1
Acmaea digitalis
Acmaea pelta
Acmaea persona
Acmaea sp.
Alvania compacta
Alvania sp.
Chlamys rubita
Clinocardium nutallii
Cooperella subdiaphana
Doto sp.
Lacuna sp.
Littorina scutulata
Margan'tes sp.
Mopalia muscosa
Mysella tumida
Mytilus edulis
Nucella (Thais) sp.
Odostomia sp.
Protothaca staminea
Transennella tantilla
herbivore
herbivore
herbivore
herbivore
herbivore
herbivore
suspension
suspension
suspension
carnivore
herbivore
herbivore
herbivore
carnivore
suspension
suspension
carnivore
other
suspension
suspension
  Trophic classification after Simenstad et al. (1979).
                        ANALYSIS OF VARIANCE FOR
                      INDIVIDUAL NUMBERS PER BRICK
SOURCE        DEGREES OF FREEDOM      MEAN SQUARE       SIGNIFICANCE2
Month
Tide Level
Oil Treatment
Error
9
1
1
550
6.705
4.184
12.822
1.242
Yes
No
Yes

2 Probability for Type I error is equal  to or less than 0.01.

                              31

-------
It is apparent that several  of the species indicated in the composition are
ones  which  are  normally characteristic  of soft  substrates,  particularly
including several  species  of suspension-feeding  clams.   The  abbreviated
analysis of variance indicates significant (p = 0.01) effects on density of
mollusks due  to month  and  due to the oil  treatment.   Significant  effects
due to tide level were not demonstrated.

     Figure 2 shows the mean densities of mollusks by month,  oil treatment,
and tide level.   Peak densities of mollusks were in March,  followed  closely
by  June.   Numbers  of  mollusks  on  bricks  were  lowest  in  January  and
February.  The  maximum mean  difference in density  of  mollusks  was  between
March  and   January,  and  far  exceeds  mean  differences  due  to the  other
sources.  The significant effect  due to oiling (C-0,  Figure  2) is  greater
than the difference shown for tide level  (+2'-0).

     The species  composition, trophic groups, and  an  abbreviated  analysis
of  variance  for numbers  of  individual crustaceans  are shown  in Table 4.
There was  a greater  number  of crustacean species  as  compared  to  mollusks
(Table  3).   The trophic structure of the  group also differed,  with detri-
tivores  making  up  35%;  herbivores,  23%;  suspension-feeders,  16%;  carni-
vores,  16%;  and  the   "other"  category  contributing  10%.   Thus,  the  31
species were more  evenly distributed than mdllusks in trophic composition.
Although the  barnacles  were  represented  by  two species  (Table  4),  the
composition of crustaceans found associated with the bricks reflects mostly
transitory species.

     Because  a  high  proportion  of  individual  numbers of crustaceans  per
brick were represented in a single species, Exosphaeroma sp.  (a scavenging,
herbivorous isopod), the data on this species were removed from crustaceans
as  a  group for  the analysis  of  variance shown  in  Table  4.   The  analysis
reflects significant  effects  on  density  of  crustaceans  due  to month of
experiment  and  tide  level.    Significant  effects due  to the  oil  treatment
were not demonstrated.

     The largest  peak  in numbers of  individual  crustaceans  per  brick was
seen  in October 1979  (Figure 3).   A  much  smaller, but distinct,  peak in
numbers  was  seen  in  March  1980.   Mid-winter  and mid-summer  numbers  were
quite  low.   The significant  tide level  effect  on  numbers  of  crustaceans
indicates a much  higher number at the MLLW level.   This  difference due to
tide  level  is much greater  than  the difference due to oiling.   It is in-
teresting,  although not  statistically significant, that the apparent effect
of the  oiling was to reduce the number of individual crustaceans.

     The analysis  of  variance  for  Exosphaeroma  sp.  alone follows  exactly
the same pattern as did  that  for remaining crustaceans.  The data, in Table
5,  indicate  significant effects  on density  due to  month  of experiment and
tide  level,  but do not  demonstrate  a  statistically significant effect due
to oil  treatment.

     The peak in  density  for Exosphaeroma  sp. occurred in  the March 1980
experiment  (Figure 4).   High numbers were  observed  also in  October and
                                   32

-------
Table  3.  Species composition and analysis of variance  for mollusk
          density in monthly hard substrate  recovery experiments
          at Sequim Bay.
                               COMPOSITION
         SPECIES                                TROPHIC CATEGORY1
Acmaea digitalis
Acmaea pelta
Acmaea persona
Acmaea sp.
Alvania compacta
Alvania sp.
Chlamys rubita
Clinocardium nutallii
Cooperella subdiaphana
Doto sp.
Lacuna sp.
Littorina scutulata
Margarites sp.
Mopalia muscosa
Mysella tumida
My til us edulis
Nucella (Thais) sp.
Odostomia sp.
Protothaca staminea
Transennella tantilla
herbivore
herbivore
herbivore
herbivore
herbivore
herbivore
suspension
suspension
suspension
carnivore
herbivore
herbivore
herbivore
carnivore
suspension
suspension
carnivore
other
suspension
suspension
  Trophic classification after Simenstad et al. (1979).
                        ANALYSIS OF VARIANCE FOR
                      INDIVIDUAL NUMBERS PER BRICK
SOURCE        DEGREES OF FREEDOM      MEAN SQUARE       SIGNIFICANCE2
Month
Tide Level
Oil Treatment
Error
9
1
1
550
6.705
4.184
12.822
1.242
Yes
No
Yes

2 Probability for Type I error is equal  to or less than 0.01.

                              31

-------
It is apparent that several of the species indicated in the composition are
ones  which  are  normally  characteristic  of soft  substrates,  particularly
including several  species  of suspension-feeding  clams.    The  abbreviated
analysis of variance indicates significant (p = 0.01) effects on density of
mollusks due  to month  and due to the oil  treatment.   Significant effects
due to tide level were not demonstrated.

     Figure 2 shows the mean densities of mollusks by month, oil treatment,
and tide level.   Peak densities of mollusks were in March, followed closely
by  June.   Numbers  of mollusks  on  bricks  were  lowest  in  January  and
February.  The  maximum  mean  difference in density of  mollusks  was between
March  and   January,  and  far  exceeds  mean  differences  due  to the  other
sources.  The significant  effect  due to  oiling (C-0,  Figure  2) is greater
than the difference shown for tide level  (+2'-0).

     The species  composition,  trophic groups, and an  abbreviated  analysis
of variance  for numbers  of  individual crustaceans  are shown  in  Table 4.
There was  a greater  number  of crustacean species as  compared  to  mollusks
(Table 3).   The  trophic structure of the group also  differed,  with detri-
tivores  making  up  35%;  herbivores,  23%;  suspension-feeders,  16%;  carni-
vores,  16%;  and  the  "other"  category   contributing  10%.   Thus, the  31
species were more  evenly  distributed than mollusks in trophic composition.
Although the  barnacles  were  represented  by  two  species  (Table  4),  the
composition of crustaceans found associated with the bricks reflects mostly
transitory species.

     Because  a  high  proportion  of  individual  numbers of  crustaceans  per
brick were represented in a single species,  Exosphaeroma sp. (a scavenging,
herbivorous isopod), the data on this species were removed from crustaceans
as a  group  for  the  analysis  of  variance shown in Table  4.   The  analysis
reflects significant effects  on  density  of  crustaceans  due to  month  of
experiment  and  tide level.   Significant  effects  due  to  the  oil  treatment
were not demonstrated.

     The largest  peak in  numbers of individual crustaceans  per brick was
seen  in  October 1979  (Figure  3).   A  much  smaller, but  distinct,  peak in
numbers  was  seen  in  March 1980.   Mid-winter  and mid-summer  numbers  were
quite  low.   The significant  tide level  effect on numbers  of  crustaceans
indicates a much  higher number at the MLLW  level.   This  difference due to
tide  level  is much greater  than  the difference due to oiling.  It  is in-
teresting,  although not statistically significant, that the apparent effect
of the oiling was to reduce the number of individual  crustaceans.

     The analysis  of variance  for  Exosphaeroma  sp.  alone follows exactly
the same pattern as did that for remaining crustaceans.  The data,  in Table
5, indicate significant  effects  on  density due to month  of experiment and
tide  level, but do not demonstrate a  statistically  significant effect due
to oil treatment.

     The peak  in  density  for  Exosphaeroma sp. occurred  in the March 1980
experiment  (Figure 4).   High numbers  were observed  also in  October and
                                   32

-------
                                                 +2' = +2'  above MLLW

                                                  O1 = MLLW
         1979
                  MONTH
1980
                  TREATMENT
Figure 2.   Mean densities  for  mollusks  in  numbers  per square meter by
           month and  in  numbers  per  brick  by  treatment.
                                     33

-------
Table 4.
Species composition and analysis of variance for crustacean
density in monthly hard substrate recovery experiments
at Sequim Bay.
         SPECIES
                               COMPOSITION
                                      TROPHIC CATEGORY1
Ampithoe simulans
Ampithoe sp.
Aoroides columbiae
Balanus crenatus
Balanus sp.
Caprella laeviuscula
Caprellidae (undet.)
Cancer sp. (larval)
herbivore
herbivore
detritivore
suspension
suspension
herbivore
herbivore
carnivore
     Corophium sp.
     Exosphare'roma amplicauda
     Gnorimosphaeroma o.  oregonensis
     Hemigrapsus nudus
     HemigrapsTis" sp.
     HeptacarpUs nudus
     Idothea wosnesenskii
     Ischyrocerus sp.
     Jassa sp.
     Leptochelia dubia
     LeptocheTTa sp.
     Melita sp.
     Orchomene pacifica
     Pagurus sp.
     Parallorchestes  ochotensis
     Petrolisthes
     Petrolisthes
        sp.
        eriomerus
     Photis  sp.
     Pinnixia  eburna
     Pinnixia  faba
     Pinnixia  sp.
     Ppntogeneia  inermis
     Shrimp  fragments
detritivore
herbivore (scavenger)
herbivore (scavenger)
carnivore (scavenger)
carnivore
carnivore
herbivore (scavenger)
suspension
detritivore
detritivore
detritivore
detritivore
detritivore
detritivore
detritivore
detritivore
suspension
suspension
others-parasite
others-parasite
others-parasite
detritivore
carnivore
  Trophic  classification  after  Simenstad  et  al.  (1979).
                              34

-------
Table 4.   (Continued)
             ANALYSIS OF VARIANCE FOR INDIVIDUALS PER BRICK
SOURCE        DEGREES OF FREEDOMMEAN SQUARESIGNIFICANCE2
Month
Tide Level
Oil Treatment
Error
9
1
1
580
1724.0
3204.6
49.7
42.4
Yes
Yes
No

2 Probability for Type I error is equal  to or less than 0.01.   Analysis
  of variance excludes Exosphaeroma sp.  which is treated separately.
                             35

-------
        0
                                                +2'  = +2' above MLLW

                                                 0'  = MLLW
N
M
M
         1979
                MONTH
                     1980
°'LMONTH™*

TREATMENT
Figure  3.   Mean densities for crustaceans  excluding Exosphaeroma sp.
           in numbers per square meter by  month and in numbers per
           brick by treatment.
                                   36

-------
MEAN  NUMBERS PER SQUARE METER
     MEAN NUMBERS PER BRICK

-------
November,  1979,  and  much reduced  numbers in  other months.  The  statis-
tically  significant  tide level  effect  on  density appears  to  be  even
stronger and  in  the same direction as the tide level effect on crustaceans
as  a whole.   Mean differences  in  numbers  due  to oil treatment  favored
control bricks.

     The species composition,  trophic  groups,  and analysis of variance for
numbers  of individual  polychaetes are  shown  in Table  6.   There were  21
species of  polychaetes  overall  on bricks.  Of these, more  than half  (52%)
were  detritivores.   One-third  were carnivores  (33%).   The remaining  two
species were herbivores.  The composition somewhat reflects, and individual
brick observation confirms, that most of the polychaetes on bricks were not
attached; even the tube-building forms were transient in nature.

     The analysis of variance of numbers of polychaetes  per brick (Table 6)
shows  significant   effects  due to  month  of  experiment,  tide  level,  and
treatment  with oil.  A  graphic  display  of  the mean numbers  per  brick is
shown  in  Figure 5.  The  peak  number  of polychaetes was seen  in the first
monthly experiment  (October  1979).   A smaller but also  substantial  peak in
numbers  was  in  March  1980.   Mid-winter  and  mid-summer numbers of poly-
chaetes were  low.   A  greater  number of polychaetes occurred  on bricks at
the  MLLW  tide level.   Mean  difference  due to tide  level  far  exceeded the
difference  due to   oil  treatment,  although the  latter  effect was  statis-
tically significant.

     Analyses of variance were also performed on the numbers of species per
group, in total, and by taxonomic group (Table 7).   For  each of the groups,
and  as a whole,  statistically significant differences in number of species
per  brick  were demonstrated  for  the effect  of month  of  experimentation,
effect due to tide  level, and effect due to oiling.

     The mean numbers  of  species  per  brick  are graphically  displayed  by
month  of  experiment in  Figure 6.  Total  species  peaked  in October 1979,
March  1980,  and  May   1980,  and  were  closely paralleled  by  crustacean
species, the  major contributor (Figure 6).  Polychaetes also  followed the
same  trend with the  exception that the  winter decline ended by December
1979, as compared to January for crustaceans and total species.  Numbers of
mollusk  species  were lower  than  for the  other taxonomic  groups,  and the
maximum  number of   species occurred in June  1980.   The mollusks  did have
peaks  in  October  1979  and   March  1980  which  paralleled the  numbers  of
species for other groups.

     A comparison of mean numbers of species per brick between the two tide
levels  is   shown  in  Figure  7.   The  MLLW tide  level  exhibited  a  greater
number of  species  per brick for all groups with the exception of mollusks.
A similar type comparison of mean number of species between oil-treated and
untreated  bricks  is in  Figure  8.   For each of  the  taxonomic  categories a
greater  number of   species per brick is  indicated  for  the  control  bricks.
These  differences,  without exception,  were  statistically  significant.   In
an  attempt to evaluate whether or  not  the trophic  mode of species related
to  oil  treatment  effects, the species were  grouped as:   (1) suspension-
                                   38

-------
Table 5.   Abbreviated analysis of variance for density (individuals
          per brick) of Exosphaeroma sp.

SOURCES
Month
Tide
Oil
Error
DEGREES OF FREEDOM
9
1
1
550
MEAN SQUARE
6938.7
19773.0
1264.5
246.56
SIGNIFICANCE1
Yes
Yes
No

1 Probability for Type I error is equal  to or less than 0.01.
                              39

-------
Table 6.  Species composition and analysis of variance for polychaete
          density in monthly hard substrate recovery experiments at
          Sequim Bay.


                               COMPOSITION
         SPECIES                                TROPHIC CATEGORY1
     Anatides groenlandica                          carnivore
     Anatides sp.                                   carnivore
     Armandia brevis                                detritivore
     Armandi'a sp.                                   detritivore
     Ci'rratulus c. cirratus                         detritivore
     Eulalia s"p.                                    carnivore
     Exogone verrugera                              detritivore
     Exogone sp.                                    detritivore
     Halosynda brevisetosa                          carnivore
     Harmothoe' imbricata'                            carnivore
     Harmothoe sp.                                  carnivore
     Nereis ve'xillosa                               herbivore
     Nothria~e1egans                                herbivore
     gphiodromus pugettensis                        carnivore
     PlatynereTs bicanaliculata                     herbivore
     Polydora social is                              detritivore
     Polydora sp.                                   detritivore
     Sphaerosyllis call form'ensis                   detritivore
     Spionidae (.undet.)                             detritivore
     Terebellidae (undet.)                          detritivore
     Thelepus crispus                               detritivore
  Trophic classification after Simenstad et al. (1979).
             ANALYSIS OF VARIANCE FOR INDIVIDUALS PER BRICK
SOURCE        DEGREES OF FREEDOMMEAN SQUARESIGNIFICANCE2
Month
Tide Level
Oil Treatment
Error
9
1
I
550
72.227
198.45
29.05
2.58
Yes
Yes
Yes

2 Probability for Type I error is equal to or less than 0.01.
                               40

-------
                       MEAN NUMBERS PER  SQUARE METER
                                                                                       +
                                                                                       ro
                                                                                       +
                                                                                       ro
                                                                                       CU
                                                                                       cr
                                                                                       O

                                                                                       fD

                                                                                       3:
r^ ,
IV
tt>

^ Ci C
t-a CM o

1 ^v3 QO Ci "N3

Co
Oi
00
     Cn
                      tn
                                       an
                            MEAN  NUMBERS  PEP BRICK

-------
Table 7.   Analyses of variance of mean number of species per brick in
          monthly hard substrate recovery experiments.
SOURCE          DEGREES OF FREEDOM     MEAN SQUARE       SIGNIFICANCE1
1.




2.




3.




4.




Polychaete Species
Month
Tide
Oil
Error
Crustacean Species
Month
Tide
Oil
Error
Mollusk Species
Month
Tide
Oil
Error
Total Species
Month
Tide
Oil
Error

9
1
1
550

9
1
1
550-

9
1
1
550

9
1
1
550

7.97
23.36
6.95
0.25

31.69
106.63
15.73
0.86

2.21
2.80
1.78
0.25

81.21
190.18
62.05
1.71

Yes
Yes
Yes


Yes
Yes
Yes


Yes
Yes
Yes


Yes
Yes
Yes

1 Probability that Type I error committed is less than or equal to 0.01.
                              42

-------
    OJ
    3
    Q.
   fD
   O>     —i
    O
    rt-
    T3
    (D
    O

    n>
    V)
   CT
   fD



   °l
co
    O. -J.
-"• fD
3  -5
O
—" CT
c  -s
Q. ->•
fD  O
    O 3
    O rh
    3 3-
    r+
    -S -h
    O O
    Cu rl-
    3 Eu
    Q. X
       O
    O 3
    fD ->•
    Q. O

    CTU3
    -s -s
    _•. o
    O C
    V) (/>
              IQ
              00
              o
                          o

                          PO
                                 o
                                         p
                                         oo
                                                   INi
>


sj


C-|


^
                                             MEAN  SPECIES  PER  BRICK

                                               —•    —>      ro    ro     ro     ro
                                               ••••••
                                               cr>    oo      o    ro     j^>     CD

                                                1	1	1	*	(	»-
IN3

CO
co
•
o
co

ro
                                                                                                                      co
                                                                                                                                 co
co

00
                                                                                                                                                    IX)

                                     j-l
                            a

-------
OJ
(/
     ~s


     *vl
   CV> fD
     CU
   II 13
o  o» cr
o>  cr n>
— -o -5
   < c/)
  fD
-J. I — (/>
-h ST3
-i ---- fl>
n  •  o
o>    _i.
3    n>
n  s: w
o>  o
•   -s -a
   a. to
     -s

   =' cr

  -a -••
   Ql O
   -S 75-
   fD
   3 cr
   fD rh
   to -".
   fD Q.
   C/> fD
   3 fD
   O. <
   O  — «
   ED
   fD
   in
      O

      II
               1
               I
I
                      O




                      +
                      +
                      ro
                      ro
                                                O

                                                oo
MEAN  NUMBER OF  SPECIES  PER BRICK

     ,_,     i_i     i_i    H1     H"      ^     ^     ^

     ON3J>^°°°N:)'P"
                                                                                                               CO


                                                                                                               -I-
                                                                                                      U)
                                                                                                      •

                                                                                                      O

                                                                                                     —t-

-------
    3.2


    3.0-


    2.8-


    2.6
  U   '
  H
  C*
  « 2.2

  a;
     2.0
  CO
  w
  H
  O
  w
  I
  1.6


  1.4


I  1.2
)
4

-t  i.o-


  0.8


  0.6


  0.4-


  0.2-
                               (IES)
                                                           (YES)
                                   T
                                              T
               TOTAL
                         POLYCIIAETE
CRUSTACEA!:
MOLLUSK
Figure 8.   Mean numbers of species per brick by oil treatment (C = control;
           T = oil treated).  Word in parentheses indicates statistical
           significance.
                                     45

-------
Table 8.   Abbreviated analyses of variance for densities of trophic
          groups in monthly hard substrate recovery experiments.
SOURCE          DEGREES OF FREEDOM     MEAN SQUARE       SIGNIFICANCE1


1.    Suspension Feeders

     Month               9                  4.98             Yes
     Tide                1                  7.25             Yes
     Oil                 1                  9.26             Yes
     Error             550                  0.70

2.    Detritivores

     Month               9                955.97             Yes
     Tide                1                885.00             Yes
     Oil                 1                  6.00              No
     Error             550                 28.92

3.    Herbivores




4.




Month
Tide
Oil
Error
Carnivores
Month
Tide
Oil
Error
9
1
1
550

9
1
1
550
8010.9
5342.8
1631.5
265.26

6.31
15.03
7.56
0.76
Yes
Yes
No


Yes
Yes
Yes

1 Reject the hypothesis of zero effect at p = 0.01.
                              46

-------
feeders; (2) detritivores; (3) herbivores; and (4) carnivores.   Only two of
the  79 species  were not  accommodated  by this  grouping.   An  analysis of
variance was computed for the density per brick for each of the groups, and
abbreviated tables  are  shown  in  Table 8.  Statistically  significant  (p =
0.01)  effects  on  trophic group  density  from tide  level and  month  were
demonstrated for  every  group.  Statistically  significant  effects  from the
oil  treatment  were demonstrated  for  suspension-feeders  and  carnivores.

     The mean  densities of  suspension-feeders by month and  treatment are
shown in Figure 9.  Peak density for suspension-feeders occurred in June at
the time of mollusk peak density (Figure 2).   The statistically significant
effect of oil  treatment on suspension-feeders was about equal  to tide level
effects.  A greater number of suspension-feeders was at the +2'  tide level.
Effects from month  of  experiment  were much greater  than tide  level or oil
treatment effects.

     Mean densities  for detritivores  are shown in Figure 10.   Peak density
for this group  was in the October  1979 experiment.  A secondary peak in the
March 1980  experiment is  much lower than  that in October.  The difference
in  mean density  indicated  between  control   and  oil-treated bricks  (C-0,
Figure 10) is  negligible.   In contrast to the suspension-feeders, a greater
number of detritivores was at the  MLLW tide level.

     The mean density for  herbivores  is shown  in  Figure 11.   Peak density
was  in the October 1979  experiment,  and this  peak  was  about twice  the
secondary peak  in the March 1980 experiment.   The difference in density due
to oil  treatment  was not  statistically significant.   A greater density for
herbivores was  at the MLLW tide level.

     Mean number  of  carnivores  is  shown in Figure 12.   Peak density was in
the  March  experiment,  and a  secondary  peak  was in the May experiment.  A
greater  number  of  carnivores were at  the MLLW  tide  level.    The  statis-
tically significant  effect on  carnivore  density due to oil  treatment was
nearly as large as the effect indicated for tide level.

     Site Experiments.   Two  additional  experiments  were  conducted at Rocky
Point during the months of  June  and  July 1980 to evaluate the  effect of
field exposure  site  on  epifauna!  recovery.  We attempted to make the Rocky
Point experiments  identical  in all  respects  to  the Sequim Bay  "monthly
experiments conducted during those same  months.   For analyses of  data we
used the abbreviated model:


                      = si + MJ + °k + Ti  +  E

     where     Y  = magnitude of the response  variable;

               S  = the main effect of site on the response  variable;
                   i = (1  = Sequim Bay or 2 = Rocky Point);
                                   47

-------
               M = the main effect of month on the response variable;
                   j = (1 = June or 2 = July 1980);

               0 = the main effect of oil  treatment on the response
                   variable; k = (1 = oiled or 2 = unoiled);

               T = the main effect of tide level on the response
                   variable; 1 = MLLW or 2 = +2' above MLLW; and

               E = random error.

The response variables  evaluated  in the model were the same ones evaluated
in  the  monthly  experiments,  i.e.,  numbers  of  individuals per  brick  and
numbers of  specific entities per  brick.   These latter are  referred  to as
"species"  with the same qualifications previously given.
     A few  points of  clarification  are warranted.  First, we  are testing
hypotheses concerning differences due to site.   The data for the Sequim Bay
site is the  same  data used in "monthly" experiment analyses for the months
of  June  and  July 1980.   For rigorous  statistical  treatment,  these  data
should not  again  be  used  in the  present analysis.   However,  there  is no
reason to  suspect that  the use  of  these data  will  in  any  way  bias  the
outcome of the present analysis.   Second, the evaluation of the main effect
of month in these  experiments involves only two months,  i.e.,  June and July
1980.  Because of  this, differences in main effect means between months can
be expected  to  be much smaller than differences  shown  over the 10 monthly
experiments.   Also,  means  for a  response variable magnitude for  a  given
month  (June  or July, 1980) will  be different  than  those reported  under
monthly experiments because both  sites are used in arriving at these means.

     By way  of  orientation,  it is instructive to reexamine Figure 6,  which
presents numbers  of  species  within taxonomic groups  by  month of experiment
at Sequim Bay.  These data indicate  that the  months  chosen for these site
comparisons  (June and July,  1980) were months in which  the total number of
species and  number  of species  within  the  constituent  groups  (with  the
exception of mollusks) were  at near  minimal  values.  Thus,  in retrospect,
one can appreciate that peak recovery months  (March and  October) might have
been better for the site comparison.

     The sites  of experimentation are compared in terms  of density within
taxonomic and trophic groups in Figure 13.  Sequim Bay had a greater number
of  polychaetes, mollusks,  carnivores,  and suspension-feeders,  while  Rocky
Point had greater  numbers of crustaceans, detritivores,  and herbivores.   In
all  cases, with the  exception of mollusks and  herbivores,  the differences
in  density   attributable to  site  are  statistically  significant  (Table 9,
Figure 13).   For  taxonomic  groups,  crustaceans contributed  the greatest
number of individuals per  brick  at each of the sites.   For trophic groups,
detritivores were the predominant group at Rocky Point,  while herbivores
contributed  the greatest number  of  individuals  per  brick at  Sequim  Bay.
                                   48

-------
   45
   40-
Pi
8
CO
w
   35
   30--
   25"
   20-
   15
  10 •
   5-
                                              +2'

                                               0'
                  "2+ above  MLLW

                  MLLW
     SO    NDJFMAMJ
          1979   MONTH
1980
                TREATMENT
Figure 9.  Mean densities for suspension feeders per square meter by
          month and in numbers per brick per treatment.
                                  49

-------
                                                +2' = +2'  above MLLW
                                                 0' = MLLW
                                                                     ••14.00

                                                                     ••IS.13

                                                                     • '12. 25

                                                                     ••11.38

                                                                     ••10.50
                                                                            ^!
                                                                       9.63  H
                                                                            fc;
                                                                            nq
                                                                       5.75  ft;
                                                                            fe]
                                                                            R4
                                                                       7.55  03
                                                                            ?
                                                                            &q
                                                                       7.00  |!
                                                                            t>
                                                                            fe!
                                                                       6.13  ^
                                                                            "=U
                                                                            I
                                                                       5.25

                                                                       4.38

                                                                     •• 3.50

                                                                       2.63

                                                                       1.75

                                                                       0.88
         1979
                 MONTH
1980
                  TREA7MENT
Figure 10.   Mean densities for detritivores  in numbers per square meter
            by month and in numbers per brick by treatment.
                                    50

-------
                              +2' = +2' above MLLW
                               0' = MLLW
                                                                  -•2.63
                                                                         fe:

                                                                         fe:
          1979
                 MONTH
1980
                TREATMENT
Figure 11.  Mean densities for herbivores  in numbers per square meter
           by month and in numbers per brick by treatment.

                                  51

-------
                                        +2' '= +21  above MLLW

                                        0' = MLLW
        1979
                MONTH
1980
                                               J "Sw/ww™*  LEVeL
                TREATMENT
Figure 12.  Mean densities  for carnivores in numbers per square meter
           by month and in numbers per brick by treatment.
                                  52

-------
en
co
   11-


   10-


    9-
u
M
PS   8'

01
Wi   ~t •
t~*t   f



1   6'

&   5-




I
    3'
                                                                      (YES)
                                          (NO)
                                                        (YES)
                                                                                    (NO)
                                                                                              (YES)
           S.E.  R.P.     S.B.  R.P.     S.B.  R.P.      S.B.  R.P.     S.B. R.P.      S.B. R.P.     S.B. R.P.
          POLYCHAETES
                    CRUSTACEA
MOLLUSKS
                  TAXONOMIC
CARNIVORES   DETRITIVORES    HERBIVORES    SUSPENSION

         ,TROPHIC
    Figure 13.
           Mean densities of individuals within taxonomic and trophic  groups  by  site
           (S.B. = Sequim Bay; R.P.  = Rocky Point).   Data summarized over  two tide  levels
           (MLLW and +2' above MLLW); two oil  treatments (oiled and unoiled); and two
           months (June and July 1980).  Word in parentheses  indicates  statistical
           significance.

-------
     A comparison  of  taxonomic  and trophic group densities  between  months
is  shown  in Figure 14.   Greater  numbers  per brick were  in June for all
taxonomic and trophic  groups with the exception of suspension-feeders.   The
differences due  to the month of experiment were  statistically  significant
(Table 9, Figure 14) with the exception of mollusks  and suspension-feeders.

     Tide  level  effects on  group  density are compared in Figure  15.   All
groups with the  exception  of suspension-feeders exhibited a  higher density
at  the  MLLW tide  level.   Tide   level  differences  in  density were statis-
tically   significant   (Table 9,  Figure  15)  except  for  mollusks   and
suspension-feeders.

     The effect  of  oil  treatment on group density  in  the  site  experiments
is presented in  Figure 16.   In  all cases, a  smaller number  of  individuals
per  brick is  observable on  oil-treated bricks  as compared to  controls.
Statistically significant effects  from  the oil treatment  were demonstrated
for polychaetes, mollusks,  carnivores, and herbivores (Table  9,  Figure  16).

     The influence  of  site  and  oil treatment on the numbers  of  species per
brick in taxonomic groups for the site experiments is shown in Table  10 and
Figure 17.   A  comparison  of sites  (Figure  17A) reveals  a greater  total
number of  species per  brick and  greater  numbers  of  polychaete  and  crus-
tacean species  at Sequim Bay  than at  Rocky  Point.   These  indicated  site
differences  are  statistically   significant  (Table  10 and Figure  17).   A
greater mean  number of  species  of mollusks  per  brick is shown  for  Rocky
Point  (Figure  17), but the  difference  is  not statistically  significant
(Table 10).

     Numbers of  species per brick for  polychaetes,  crustaceans,  mollusks,
and overall, are shown in Figure 17B as related to oil  treatment.   In every
instance  a greater number of species  is shown for  control  bricks as  com-
pared  to   oil-treated  ones.   The  differences indicated are statistically
significant (Table 10, Figure 17) in every instance.

     Numbers of species  per brick  comparisons of month and  tide  level
effects are shown in Figure 18.   Total number of species per  brick, as  well
as numbers of species  in each of the constituent groups, was  higher in  June
than in July.   Differences attributable to month of  experiment were statis-
tically significant (Table 10,  Figure 18) except for mollusks.   There was a
larger number of species at the  MLLW tide level as compared to  the +2'  tide
level  in  every  instance.    For  mollusks, the tide  level   effect was not
significant.
                                   54

-------
en
en
        11-

        10-
    q..
S4
O
H

&   8*
     en
     &,
     KS
     a
              (YES)
                                                                    (YES)
                                          (NO)
                                                  (YES)
                                                                                   (YES)
                                                                                                  (NO)
          POLYCHAETES
6    7
                   CRUSTACEA
6    7
                                                       6   7
                                                                6    7
            MOLLUSKS
           CARNIVORES   DETRITIVORES    HERBIVORES
                                                                                               SUSPENSION
                   TAXONOMIC
                                                                    TROPHIC
      Figure 14.
            Mean densities of individuals within taxonomic and trophic  groups  by month
            in site experiment (6 =  May; 7 = June).  Data summarized over  two  sites
            (Rocky Point and Sequim  Bay); two tide levels (MLLW and +2  above MLLW);
            and two months (June and July 1980).  Word in parentheses indicates
            statistical significance.

-------
Table 9.   Analyses of variance for density of taxonomic and trophic
          groups in hard substrate site experiment.1
RESPONSE VARIABLE/
SOURCE OF
VARIATION       DEGREES OF FREEDOM     MEAN SQUARE       SIGNIFICANCE2


TAXONOMIC GROUPS

1.   Crustaceans

     Site                1               1991.5              Yes
     Month               1               4543.5              Yes
     Oil                 1                233.9               No
     Tide                1               4591.2              Yes
     Error             220                 68.44

2.   Polychaetes

     Site                1                  5.37             Yes
     Month               1                  8.89             Yes
     Oil                 1                  5.11             Yes
     Tide                1                  6.96             Yes
     Error             220                  0.31

3.   Mollusks

     Site                1                  0.31              No
     Month               1                  8.72              No
     Oil                 1                 18.06             Yes
     Tide                1                  2.94              No
     Error             220                  1.64
1 Analyses of variance performed on data from two sites (Sequim Bay
  and Rocky Point); two months (June and July 1980); and two tide levels
  (MLLW and +2'  above MLLW).   The densities of organisms on hard sub-
  strates were those which resulted after a 30-day field colonization
  period.

2 Probability for Type I error equal to or less than 0.01.
                              56

-------
Table 9.  (Continued)
RESPONSE VARIABLE/
SOURCE OF
OF VARIATION    DEGREES OF FREEDOM     MEAN SQUARE       SIGNIFICANCE2
TROPHIC GROUPS

1.   Carnivores





2.





3.





4.





Site
Month
Oil
Tide
Error
Detritivores
Site
Month
Oil
Tide
Error
Herbivores
Site
Month
Oil
Tide
Error
Suspension-Feeders
Site
Month
Oil
Tide
Error
1
1
1
1
220

1
1
1
1
220

1
1
1
1
220

1
1
1
1
220
3.60
2.95
3.29
2.84
0.26

2116.3
3530.1
44.2
2759.0
53.9

3.12
180.14
125.25
401.03
11.87

11.32
3.51
3.12
5.05
0.95
Yes
Yes
Yes
Yes


Yes
Yes
No
Yes


No
Yes
Yes
Yes


Yes
No
No
No

1 Analyses of variance performed on data from two sites (Sequim Bay
  and Rocky Point); two months (June and July 1980); and two tide levels
  (MLLW and +2'  above MLLW).   The densities of organisms on hard sub-
  strates were those which resulted after a 30-day field colonization
  period.

2 Probability for Type I error equal to or less than 0.01.

                              57

-------
en
CO
     O
     M
     &
     05
     fcl
     PI
     to
     f«
     M
    i
       ir

       10-
41
        2-
                                                       (YES)
                                                                                          (NO)
           -0'  +2'
         POLICEAETES
0'  +2'
                CRUSTACEA
                               O1  +2'
MOLLUSKS
O1  +2'
O1   +2'
                           CARNIVORES   DETRITIVORES
                                           O1   +2'
                           HERBIVORES
                                                                                                 0'   +2'
                            SUSPENSION
                  TAXQNOMIC

    Figure 15.   Mean  densities  of individuals within taxonomic and trophic groups  by tide  level
                (O1 = MLLW;  +2=2' above MLLW).  Word in parentheses indicates  statistical  significance.
                Data  summarized over two sites (Rocky Point and Sequim Bay);  two months  (June and July
                1980);  and two  oil treatments (oiled and unoiled).

-------
                                                      6S


                                                    MEAN NUMBERS  PER BRICK
to GO <
-"• O :
CO  i  '
3 Q.
_j. eu ;
-h*< <
      o ro
         O*
       II 3
      O

      O
o -h
cu -i. 3- <-+•
3 ro to -s
o —«    o
ro IT
             r-
             VJ
         (D
         o
         -h
   X 3
   •a ro
   o     ii
   to cu
   C 3 O -••
   -S O- -"• 3
   ro    —• a.
      r ,s	- —j.
   -s <<    a.
   -••    o c
   O i—' Cu  CU
   Q- IO c+ —'
      00 O»  to
   S O
   Cu	to  S
   tO ». £1  -^-

   c cu 3  3-
   to 3 cu  -j.
   ro ex -s  3
   Q_    -1.

      S ro  cu
      o a. x
   s:       o
   O c-h O  3
   -5 -•• <  o
   Q. a. ro  3
      ro -s  -«•

   3" —i <-f
      fD S  Cu
  •a < o  3
   cu ro    o.
   -s —i to
   ro to -»• (H-

   rf^^ro  o
         to -a
ro
to
ro
to
-•• 3
      o
   o»  o
      7*- -S
   a_<<  o
             -G
o ro o  to
QJ  — -J.

      3  .0-
   ro  cu
   to  cr
      o  cu  o
   to  <  3  -".
   <-t ro  a. —•


   -<• i— ro  -s
   to  i— jo  ro
   n- s c  eu

   0-33-
   cu        ro
   —•    DO 3
                   §
1
P3
S
                            O

                  Co
H
BIVOREL

                          o
                          O
                                                                 •)	•. ('
                          (-3
                         O
                         0
                         0
                         O

-------
                                                     (YES)
               /\-  POLYCHAETE    CRUSTACEAN    MOLLUSK
TOTAL
Figure 17A.  Number  of species per brick  by  site (S.B. = Sequim Bay;  R.P.  =
             Rocky Point).   Data summarized  over two months (June and July
             1980);  two tide levels (MLLW and  +2'  above MLLW); and  two oil
             treatments (oiled and unoiled).   Word in parentheses indicates
             statistical  significance.
                4 -
             H
             «
             CQ

             rt
             H
             ft
              U
              w
              ft
(YES)
                                 (YES)
                        (YES)
                                           (YES)
                  -  POLYCHAETE   CSUSTACEAS
                                           MOLLUSK
                                                      TOTAL
Figure  17B.   Number of species per  brick by oil treatment  (C  =  control;
              T = treated).  Data  summarized over two sites  (Rocky Point and
              Sequim Bay); two tide  levels (MLLW and +2' above MLLW); and two
              months (June and July  1980).  Word in parentheses  indicates
              statistical significance.
                                       60

-------
ea
c
-5
00
DO
_J,
3
CL
.u*
O
Eu
d-
fD
CO
CO
r+
EU
f-f-
_j.
CO
rt-

O
Cu
~-i

CO
— J*
tQ
3
— i*
-h
_j.
O
Eu
3
0
fD
•




























EU
3
CL
^
O

O
— J

<-+•
-5
fD
EU
c+
3
ro
3
c+
CO

^— ^
O

«~j
fD
Q.

EU
3
Q-

C
3
O
_j.
^^
fD
CX
*» — ^
•


g"
O
-s
CL
-„__!,
3

"^J
Qj
-s
fD
3
c+
3-
fD
(/)
CD
CO
f — ^
33
O
O
<

T3
ro
-s

cr
-•5
a
Sj
o^
jti

^
E
_j.
O
^*

O"
t^

f-f-
.^.
Q.
&
Q
S
§
is
i*J

ro

_^
ro
<
ro


*.^-*,
+


Q
g
tM


ro
_

n

ro

Cu
cr
0
<
ro

~^
(—
j —
s?^
•^^

                      MEAN  SPECIES PER BRICK

                     j        ro       U)

                     	,	h	
I/)
c+
CU
c-l-

l/>
rl-
_j>
O
Cu
«j

CO
_J.
CQ
3
.^.
-h

O
CO
3
O
ro
•




































o
o
3

rl-
o>

c
EU
-s
_j*
N
CD
CL

O
<
fD
-s

c-t



CO
•^»
f-j-
fD
CO

^-*^
PO
o
0
^"
<<

-o
o
— 1*
3
c+

Cu
3
Q.

co
fD
JD
C
_j.
3

-s
fD
1— »
00
3=>
C ^'''w
3 -J^-
cr ,
CD '
-s
O
-h

CO
•a
CD
1

s
1
s
o
«j*
CD
co a

•o
CD
-s

CT
•~v
Q
a
g

s
5a
-s
~J«
o
7^

,cr
t*^




§;
LM
t-i
§
8

3
rl-
3"

, — ^
Oi
II

C-i

^
|


i
c
3
fD
V •

--J

II

C_i
c
«J
D3<<
Eu
<<
N»— '
W •

x^^
•



                                                                                        MEAN  SPECIES PER BRICK

-------
Table 10.  Analyses of variance for numbers of species in taxonomic
           groups in hard substrate site experiments.1
RESPONSE VARIABLE/
SOURCE OF
VARIATION       DEGREES OF FREEDOM     MEAN SQUARE       SIGNIFICANCE2
1.   Polychaetes

     Site                1                  2.61              Yes
     Month               1                  5.73              Yes
     Oil                 1                  2.81              Yes
     Tide                1                  4.76              Yes
     Error             220                  0.19

2.   Crustaceans

     Site                1                 15.09              Yes
     Month               1                 80.14              Yes
     Oil                 1                  6.26              Yes
     Tide                1                 56.87              Yes
     Error             220                  0.74

3.   Mollusks

     Site                1            '1.38              No
     Month               1                  1.64              No
     Oil                 1                  2.37              Yes
     Tide                1                  0.72              No
     Error             220                  0.28

4.   Total Species

     Site                1                 47.11              Yes
     Month               1                164.50              Yes
     Oil                 1                 31.93              Yes
     Tide                1                118.74              Yes
     Error             220                  1.59


1 Data included two sites (Rocky Point and Sequim Bay); two months
  (June and July 1980); two oil treatments (oiled and unoiled); and
  two tide levels (MLLW and +2' above MLLW).   Number of species
  resulted from a 30-day field exposure.

2 Probability for Type I error is equal to or less than 0.01.
                              62

-------
Hard Substrate Recovery - Total Oil Concentrations.

     Monthly Experiments.   Infrared  analyses for  total  oil  and  capillary
gas chromatography  for  selected saturate and aromatic compounds  were per-
formed on  samples taken  immediately  after oil  treatment, and  at five and
30-day intervals after field placement of bricks.

     A time  course  of  total  oil  concentration  in the ten  monthly experi-
ments at Sequim Bay is in Figure 19.   A more detailed presentation of these
data is in Table 11.  Two types of data are represented on the figure.  The
extraction of all oil  from individual bricks resulted in  the data labeled
"W," for whole brick,  in Figure 19.  The extraction of oil  from the top, or
colonizing surface only, resulted in the data marked "T"  in Figure 19.  For
both types of data, a marked reduction  in  total  oil  between samples taken
immediately  post-treatment,  and   at  five  days  after field placement  is
apparent.   The 30-day  data declined slightly from 5-day data for both whole
brick  and  top surface  extractions.   For the top  surface  extractions, the
decline was somewhat less than for whole brick extractions between five and
30  days.   Initially,  the  top surface oil  content of bricks was slightly
more than  half  (56%)  of whole brick content (Table 11,  Figure  19).  For
five and 30-day samples the top surface of bricks has about 20% (18-21%) of
whole  brick  extractions (Table 11).   There was no distinct  trend in pro-
portion of top surface to whole brick extractions between five and 30 days.

     Data on  Figure 19  also show a consistent,  but  not  statistically sig-
nificant  (P  = 0.05),  higher  concentration  on  bricks placed  at  +2'  above
MLLW as compared to bricks placed at MLLW.

     The data on  Table  11 indicate much  higher  initial  oil  concentrations
in  March,  April,  May,  and June experiments  than in preceding months and
July.   These  high initial  concentrations undoubtedly relate to  a lack of
control of some  aipect of  the treatment  itself  and/or improvements  in
extraction  efficiency  noted   in  an  earlier report  (Vanderhorst et al.,
1980).   However,  it can be seen from the data  in Figure 20  that the high
initial concentrations  bear  littTe relation to  the concentrations of total
oil on the  top  surface  of bricks  after  five days of field exposure.   Data
in  Figure  20 also  permit  an  experiment-to-experiment comparison in total
oil concentration on  the top surface of bricks  among the several  treatment
categories.

     Site Experiments.  A  comparison  of  total oil concentration in the two
site experiments  involving hard substrates is given  in  Figures 21 and 22.
The overall  concentrations of  total  oil  are comparable to  the  data pre-
sented previously from  monthly experiments at Sequim Bay.   From Figure 21,
there  is  no consistent  trend  in  concentration  on bricks due  to  site.  At
the end of  five  days  of field  exposure,  the mean concentrations for whole
bricks at  respective  tide  levels  were slightly higher at  Rocky Point than
at  Sequim Bay.    The  converse  was true at  the  end of  30  days  of  field
exposure.   Concentrations  at the  MLLW tide  level  were  slightly, but not
significantly (P  = 0.05), lower  than  were those at +2'  above  MLLW.   A
comparison  of mean total  oil data by month of  site experiment  (June  or


                                   63

-------
Table 11.  Mean monthly concentrations of oil on bricks (grams/brick)
           in hard substrate recovery experiments at Sequim Bay.
TOTAL OIL (GRAMS/BRICK)
MONTH
OCTOBER

NOVEMBER

DECEMBER

JANUARY

FEBRUARY

MARCH

APRIL

MAY

JUNE

JULY

INITIAL
3.67
—
4.06
0.82
3.09
1.04
2.91
1.60
9.98
5.78
15.60
8.97
10.24
6.80
16.86
8.80
14.78
8.19
6.25
2.00
5-Day
+2' MLLW
-
-
4.80
0.89
3.64
0.07
2.01
0.32
3.47
0.45
6.70
0.98
2.56
1.03
6.19
1.32
4.02
1.15
7.73
1.28
_
-
1.76
0.49
1.97
0.07
2.86
0.08
3.16
0.89
8.02
1.38
3.36
0.93
5.05
0.96
3.18
0.72
6.70
1.22
N = 5 BRICKS/MEAN
30- Day
+2'
_
-
2.43
0.07
-
-
1.32
0.32
3.10
0.72
3.84
0.92
1.39
0.38
5.28
1.70
3.01
0.53
5.93
0.84
MLLW
0.49
-
3.41
0.03
-
-
1.92
0.39
1.81
0.65
3.42
0.57
2.38
0.51
4.62
0.77
1.49
0.27
6.50
0.86
EXTRACT1
W
T
W
T
W
T
W
T
W
T
W
T
W
T
W
T
W
T
W
T
OVERALL MEANS
WHOLE
TOP
STD. DEV.
WHOLE
TOP
% RATIO
TOP/WHOLE
8.72
4.89
DUE TO MONTH
5.53
3.50

56
4.57
0.83

1.94
0.45

18
4.01
0.75

2.14
0.46

19
3.29
0.69

1.67
0.50

21
2.89
0.51

1.83
0.27

19







1 W = whole brick extraction;  T = top surface of brick extraction.
                              64

-------
                                ^T-HTIZ	-+2    ~Vr
                                           "    — 	  	 • MT.T.W J '
            O
                                                         30
                                DAYS  (Post-Treatment)
Figure  19.  Summary of infrared analyses (IR) in terms of individual
           experiment time frames (N = 50 measurements per point) in
           monthly experiments at Sequim Bay (W = whole brick extractions;
           T = top surface extractions).

                                   65

-------
   9.0-


   8.5-


   8.0


   7.5t


   7.0


   6.5t


   6.0
   c c -
O  5.5
H
ffl
w
H

0
CO
H
g
5.0


4.5


4.0


3.5


3.0


2.5


2.0


1.5


1.0


 .5
         0
                                             ~ INITIAL
                                                  , —  5 day +2'
                                                   -  5 day MLLW
                                             v  x
                                             \/'/~SO day MLLW
                    D
MAM
          1979     MONTH
                                     1980
Figure 20.  Monthly mean concentrations  of total oil  on top surface of
           experimental substrates (N = 5 measurements per point) at
           Sequim Bay.
                                   66

-------
en
-vj
               11



               10



                9
             u
             H
             rt
             H
             eu
             H

             O
                6 "
4--
                3--
                        W T
                      INITIAL
                               W T


                               +2'
                                      W  T
                                   S.B.
                             W T    _W	T_


                              +2'	0'

                                 R.P.
W  T


+ 2'
W  T

  n'
    S.B.
W  T    W  T


 + 2'	0'

    R.P.
                                           5-DAY
                                                      30-DAY
      Figure 21.   Total  oil  concentration in hard substrate site  experiments  by site, tide level, and
                  field  exposure time (R.P. = Rocky Point; S.B. = Sequim  Bay;  W - whole brick
                  extraction; T = top surface extraction; 0' = MLLW;  and  +2'  = 2' above MLLW).

-------
u
s
m
H
H
O

CO
14



13  -



12  -



11  -•



10  -



 9  •-



 8  -•



 7



 6  t
    4 -•
    2



    1
              W
                 T
W
W
                                                       T
                                                                W
                6/80
                               7/80
                  6/80
                   7/an
                      INITIAL
                                                        SO-DAY
Figure 22.
         Total  oil concentration in hard substrate site  experiments at
         MLLW by month of experiment (W = whole brick extraction;
         T =  top surface extraction).  Data summarized over two sites
         (Sequim Bay and Rocky Point).

                                 68

-------
July)  is  in Figure 22.  These  data  show a slightly higher  initial concen-
tration  of total  oil  for whole brick  extractions  and top surface extrac-
tions  in  June as compared to July.  For the 30-day MLLW data, the opposite
is  true.   The data merely reflect the  normal variation associated with the
methods, which was more clearly shown in Table 11.

Hard Substrate Recovery - Analyzed Saturates and Aromatics

     Monthly  Experiments.  Samples analyzed by capillary gas chromatography
identified  the  compounds  listed on Table 12.    Example  concentrations of
individual  compounds  from samples taken  immediately  post-treatment and at
five  and  30  days  after field  placement in the  April  1980 experiment are
also shown  in Table 12.  These data and similar data from four other sets
of  replicates sampled during the April   experiment provide the basis for the
summed  saturate  and  aromatic  compound  concentrations  given  in  Figures
23-26.   The  total  saturate  compound   data  is  based  on  a sum of  the 19
saturate compounds  analyzed,  and the total aromatic compound data is based
on  a  sum  of the 12 aromatic  compounds  analyzed.   The sum of these classes
does not  and  should not add to comprise total oil, as measured by infrared
spectrophotometry  since it includes only a few of the  many components in
oil.

     Data  on  the time  course of analyzed saturate compounds  in the April
1980  experiment  are shown in Figure 23.   The data for MLLW follow a time
course which  parallels the  data on total oil for  the experiments overall
(Figure 19).  One should note that total oil concentrations are in grams of
oil per brick while total  saturate and aromatic compounds are indicated in
milligrams  of oil  per brick.   The  data for the +2' above  MLLW  tide level
show a much greater retention of total  analyzed  saturate  compounds during
the first  five  days of field exposure than for MLLW and the opposite at 30
days.

     The time  course  of analyzed aromatic compounds during  the  April  1980
experiment  is shown in Figure 24.  The  loss of these compounds from experi-
mental substrates  at the +2' tide  level  and MLLW was about  equal  at five
days.   A greater loss was indicated for the +2'  tide level  at the end of 30
days.   The  sample  sizes for computing these means were much smaller (N = 5
bricks per  mean) than for the total  oil data (N = 50 bricks per mean), and
the indicated discrepancy is within the methodological  sensitivity.

     Comparative data  for the  whole  brick and top surface extractions at
the 30-day time interval are in Figure  25.   The top surface extractions are
directionally  compatible with  whole  brick extractions  and  also  show  a
greater concentration  of  both analyzed saturate and aromatic  compounds at
MLLW as compared to +2'  above MLLW.   For  saturate  compounds,  the top sur-
face extractions represented  about  37% of whole brick extractions  at both
MLLW  and   at  +2'  above MLLW.   For aromatic  compounds,  the  top  surface
extractions  amounted  to  nearly  50%  of whole  brick  extractions.   These
proportions exceed  somewhat  the 20%  top  surface/whole  brick relationship
for total  oil  indicated in Table 11.
                                  69

-------
Table 12.  List of saturate and aromatic compounds identified by gas
           capillary chromatography in hard substrate recovery
           experiments.*
COMPOUND
SATURATES
Cl2
Cl3
Cis
Cie
Cl7
PRISTANE
Cl8
PHYTANE
Cl9
Cao
£21
C22
^23
C24
C25
^26
C27
C28
AROMATICS
NAPHTHALENES
NAPTHALENE
2 MN
1 MN
IE, 2E
2,6 2,7
1,3 1,6
1,7
1,4 2,3 1,5
1,2
PHENANTHARENES
PHENANTHARENE
Ci
C2
EXAMPLE WHOLE
INITIAL

8.80
13.67
28.05
27.45
29.69
19.57
26.17
12.05
27.29
23.59
21.06
18.81
16.37
15.01
12.21
10.51
6.18
4.00


0.12
4.38
3.63
0.95
4.98
4.27
4.66
2.88
1.44

0.95
0.35
1.04
BRICK CONCENTRATIONS
b-DAY

4.33
7.83
14.64
18.15
18.26
11.88
15.50
7.52
15.73
13.46
11.98
11.74
10.92
9.95
8.29
7.25
4.06
2.45


0.008
0.755
0.167
0.27
0.67
0.77
0.96
0.73
0.26

0.38
0.13
0.28
(mg/brick)
30-DAY

0.10
0.18
0.31
0.33
0.40
0.31
0.34
0.19
0.38
0.30
0.29
0.26
0.23
0.21
0.17
0.15
0.08
0.11


0.001
0.032
0.032
0.013
0.053
0.017
0.047
0.025
0.011

0.011
0.006
0.015
   These example data represent a single replicate from the April  1980
   experiment at Sequim Bay,  MLLW.

                              70

-------
                                                        MLLW

                                                       '+2 '  above MLLW
                          DAYS (Post-Treatment)
                                                      30
Figure  23.  Summary of measured  saturate compounds in  terms of individual
           experiment time frames  (N = 5 measurements per point) in
           monthly experiments  at  Sequim Bay.

                                  71

-------
                         DAYS (Post-Treatment)
Figure 24.   Summary of. measured aromatic compounds in terms of individual
            experiment time frames (N = 5 measurements per point) in
            monthly experiments at Sequim Bay.

                                   72

-------
to
Oi
      n>
      ro
      tn
     • o
      o
c  n -a
"^    Q*
Q) Ul -S

ro 3  t/>
m ro  o
>•• n»  3

3> c  o
   -s  -h
 n ro

ai ro  -"•
3 3  Q.
n» rt ro
—" t/i

N -a  ro
ro ro  <
Q. ~s  ro

o o~ t/>
-s P»
G>

Ct CO c+
-•• O fO
O    -S
W Q-3
-—*-Q) CO

•  <<
   CO O
      -h
   O)
   -h3
   c+ (D
   CD 01
   -S 
   -5 CD
   fD Q.
   O>
   rt- V)
   3 a>
   ro c-i-
   3 c
   c+ -S
      O)
   0» rt-
   r+ ro

   c/5 a>
   ro s
   x» Q.
   DO 3
   a>  o>
   -—-c>
   oo
      o
    II O

   a» -a
   3 o
   Oi C
   ro
   a.
                ro
                O
                m
                        CO
                                     MEAN MILLIGRAMS PER  BRICK




                                    f	1	1	1	1	h—|—
                                                                    -f-
J  N)  N)   U)  OJ



<	1	1	1	H
                        CD

-------
     The data on Figure 26 represent a preliminary attempt to partition oil
associated with  bricks at MLLW,  30-day field  exposure,  during  the  April
experiment between  organisms  on  bricks  and the  bricks  themselves.   The
comparison in  analyzed saturate  and  aromatic compounds  is  between bricks
which were scraped  free  of organisms and those with  normal  30-day coloni-
zation.    In   terms  of whole  brick  extractions,   the  difference  between
organism-free bricks and colonized bricks is proportionally quite small for
saturate  compounds  but may be appreciable  for  aromatics.   The proportions
attributable  to  organisms in  the top surface extractions  is  apparent and
large for both saturate and aromatic compounds.


CLAM BED RECOVERY

A Perspective

     This task was principally designed to measure effects from Prudhoe Bay
crude oil mixed in sediments on recovery by the littleneck clam (Protothaca
staminea), a  commercial   species.    In terms of longevity  and  frequency of
successful sets,  this  species is  more akin to dominant species in the rock
habitat.  Thus,  the three-month  experimental  framework directly addresses
the  question  of  sediment suitability for reseeding by the clam but not the
longer  term  full  recovery for this species.  The three-month experiment at
Discovery Bay  duplicated  an  experiment conducted during the summer of 1979
at Sequim Bay for MLLW with the exception of initial oil concentration, and
included  data  for nine other primary species and  the entire infaunal com-
munity,  as  well  as  data  on  the  littleneck  clam.   A comparison  of  these
summary  data  to  the Sequim Bay experiment  lends perspective to the present
experiment.   Data for  Sequim Bay are from an interim report on this project
(Vanderhorst et al., 1980).

     Figure 27 represents  the numbers of species within taxonomic groups in
this  experiment  (D.B.)  and  the  earlier  experiment  (S.B)  at  MLLW.   The
numbers  of  species represented by the  equal  sample sizes  in controls is
remarkably  similar,  with the  exception   of  mollusks.    Polychaetes were
represented  by  21 and 20  species; crustaceans by 13 and 14 species; and the
"other"  group, consisting of all other  species,  by 6  and 7  species for
Sequim  Bay and Discovery  Bay, respectively.  The mollusks were an exception
since they were only a minor  constituent at Sequim Bay (3 species) and were
nearly  as well represented as the crustaceans at Discovery Bay (12 species).
There  were  fewer  species for each  category  in  the  oiled  sediments  as
compared  to  unoiled controls for this task's experiment.  Mollusks were an
exception to  that trend in the previously reported experiment.   Polychaetes
were represented  by the most  species  followed by crustaceans, mollusks, and
all  remaining species.

     A  similar  comparison between the two  experiments for numbers of  indi-
viduals  is  in Figure  28.  The  natural  logarithm  of numbers of  individuals
per  square  meter is used  to  permit  plotting widely divergent numbers on  a
single  figure.   The  range in means was from  3  (21 individuals per  square
meter  (oiled, other  species,  Sequim Bay)) to  11  (about 50,000 per  square


                                    74

-------
en
      32

      30

      28

      26

    O 24

    « 22

    H 20
    /^ .
      18

      16

      14

    3 12

    g 10

    3  8-



       4

       2
        CO
        H
Figure 26.
                      WHOLE
                                  S_	A_
                                   TOP
                                                             A
                      A
MOLE
TOP
                             SCRAPE
                                                        NO  SCRAPE
                Comparisons of scraped and unscraped bricks  in terms of measured saturate and
                aromatic compounds at MLLW tide level  at Sequim Bay (N = 5 measurements per
                bar) 30 days after oil treatment.  Scraped bricks were organism-free when
                analyzed; unscraped  bricks had that complement of organisms colonized in 30
                days (S = analyzed saturates; A = analyzed aromatics).

-------
  22



  20  "



  18  •-



  16  •-



H 14  '•
H
U
W

w 12  t
  10
w
g  8
   6  •'


   4  -



   2  -
          S.B. D.B.
         POLYCHAETES
                             S.B.  D.B.
                            CRUSTACEANS
S.B.  D.B.
MOLLUSKS
S.B.  D.B.
  OTHER
Figure 27.
          Comparison of Sequim Bay  and commercial clam bed at Discovery
          Bay in terms of species.   (S.B. = Sequim Bay; D.B.  = Discovery
          Bay; number of species  is aggregate in 35 cores distributed  in
          5 replicates of 7 cores each for each condition.)  Samples
          were collected 3 months after oil treatment during  the spring-
          summer season.
                                   76

-------
11


10
I
o
CO
|

s  7
H
Q
PS
O  4-
O
   3
                                                            CONTROL
                                                            OILED
         S.B   D.B
                        S.B.  D.B.
        POLJCEAETES     CRUSTACEANS
S.B.  D.B.
                                         MOLLUSKS
                                                             S.B. D.B.
                    OTHER
Figure 28.   Comparison of Sequim Bay and commercial clam bed at Discovery Bay
            in terms of natural log of numbers  of  individuals/square meter.
            (S.B. = Sequim Bay; D.B. = Discovery Bay.)  Samples were collected
            3 months after oil treatment in  the spring-summer season.  Each
            mean is based on the natural logarithm of numbers of individuals
            in 35 cores
            condition.
                     distributed  in  5 replicates of 5 cores each for each
                                     77

-------
meter (control, crustaceans,  Discovery  Bay)).   Numbers of polychaetes were
about the  same  at the  two sites  as  were numbers  of individuals  in  the
"other"  species  group.   There were more  individual  crustaceans  and indi-
vidual mqllusks at  Discovery  Bay.   In every case  there  were fewer numbers
of  individuals  in oiled  sediments  as compared  to unoiled  controls.   The
highest  number of  individuals per square meter  was  for  crustaceans  at
Discovery Bay  followed  by polychaetes at both sites,  mollusks  in controls
at Discovery Bay, and crustaceans at Sequim Bay.

     These summary data indicate that the present experiment replicated the
Sequim Bay  experiment (Vanderhorst  et  al., 1980)  very  well and that the
overall  recovery  was  proceeding  at about the same rate as for that experi-
ment.

Effect of Tide Level on the General Community

     Since the  number of  species  represented at MLLW was  very similar at
Discovery Bay  and Sequim Bay, we have used data from the -2' tide level at
Sequim Bay  to help  illustrate tide  level  trends  in  community data.   The
number of  species from  each  of  three tide levels (-21,  Sequim Bay; MLLW,
Discovery Bay; +2', Discovery Bay) are in Figure 29.   For unoiled controls,
the  polychaetes,  the  crustaceans,  and the  "other" species  group  exhibit a
decreasing number of  species  from  low to  high  tide  levels.   Mollusks were
equally represented at  the two upper tide  levels,  and had fewer species at
the  -2'  tide level.   This  may very well  be an effect of  site  rather than
tide  level   because  of  the generally  low representation  of  mollusks  at
Sequim Bay.   At  the +2'  tide  level,  polychaetes were represented by equal
numbers of  species  in treated and  control  sediments.  The oiled sediments
at  this  tide  level  contained a greater number  of crustacean species than
did  controls.   At  all   other tide  levels and  for  the  other  groups,  the
number  of species  in  controls  exceeded  the number  in  oiled sediments.

     The  numbers  of  individuals  per square meter  (expressed as  a  natural
logarithm) as  related to tide level are shown in Figure 30.  The trend for
polychaetes  and  "other"  species  was from a higher number of  individuals at
the  lowest tide  to a lower number at the  highest tide.   This  is the same
trend  as  was  shown  for  number  of species in  Figure 29.   Mollusks had a
greater  number of individuals per  square meter at MLLW followed by  the +2'
tide  level.  Crustaceans  also  had the highest number of individuals  at MLLW
but  had  higher numbers  at the -2'  level than at  +2'.   Crustaceans at the
+2'  tide  level had slightly more  individuals  per  square meter in oiled as
compared  to  control  sediments.  In all  other  cases  the  numbers of  indivi-
duals in control  sediments  exceeded that for oiled sediments.

Analysis of  Variance  for  Taxonomic  Groups

      The  data on numbers of  individuals and species  in the  commercial clam
bed  experiment are  presented in  a different way  in Table  13.   The mean
numbers  of individuals  and species per  tray  are tabulated  in  terms of oil
treatment  and tide level.  Analyses  of  variance were computed on the data
resulting  in  these means  to distinguish  statistically  significant  (P =


                                    78

-------
                                                      6L
         IQ
         ro
         ro
                                                       NUMBER OF  SPECIES
3
(T>
3
ro
-S
ro

-h
o


co

3
O
3
r+
3-
en

EU
   O cr
•~J rt- cr
    _•. ro

o  ro
-s     o
ro  —> -h
to  ro
    <  m
fD  ro "o
oj  —i ro
o  •   o

       ro"
•a  :z in
ro  c
-s  3  2
    cr —'•
O  ro  r+
O  -S  3-

Q. O  3
-!•-»» ^
C"i     CT
—>. tn  BJ
o -a  x
3  ro  o
•   O  3
    -i. O
    ro  3
 ro  i     o
 -s  ro -i.
     - (/) IQ
 O        -S
 -<• II  £U  O
 — i   
       fD
    CTCQ  -i.
    ro  cu  3
-S
ro
a>
3 o ro
 ro
 3
 c+
3 s;
         o
         o
      03
      -h 3
         ro
   r~ oo -s
      en o
3 3
 -s

 3
ta


 c
ro
o
3
   O) o OJ
   c+ O — «
      ~i
   oo ro o
   ro c/> — i
   .a    fu
   s= a. 3
         cr
      c+ ro
   co -s o.
   OJ -i.
   << cr ro
   »  E X
i—» ro  ro
^o o. -^
oo    -••
0---3
•   3  ro
*•—^    3
    cn r+

D3 -s  O.
o  ro  -"•
rt-TJ  V>
       ro
         cr
         c
         <-*•
         ro
         a.
                O
                t-l
                CO
                 to
                 C!
                 CO
                fa
                Co
                is:
                o
                CO
                o
                1-3
                £d
                eg
                to
                                                                                       to
                                 U)
                                                                              00
                                                                                                           to


                                                                                                          H-
                                                                                                                   LO

                                                                                                                   O
                                                                                           t-i
                                                                                           fa

-------
                                                        08
          to
           £=
           -s
           CD

           oo
           o
   -S 2 rt- :z
   CD r~ -•• cu
  T3 I— Q- rt-
   —' s ro c

   o' Q. —' CU
   cu cu ro —i
   rt- rt- <
   ro cu ro —•
   in    —> o
o  o -s
3  Hi O '-NO
     3  I  -h
   en   ro
     o  - 3
   o -••    c:
   O in  H  3
   -so    cr
   ro o ro ro
   e/> <  - -s

  T3 -S cr O
   ro *< ro  HI
   -s    —•
     co o  -"•

   ro <<    Q.
  "O v»  ~g —I.
   o
   £U
   rl-
   (0

   O>
   -S
   CD
  -a
   -s
   fD
   c/>
   CD
   C/l


   CO
   Cu
   o
   3"
 Q. CU
 CU  —i
 rt- to
 CU  -N,
    in
 -h-Q

 O  Cu

    ro
 oo
 ro  3
 -O  ro
 C  rt-
 -j. ro
 3  -S

 D3 S
 CU  -i.
   O
   O
   N
   CU
   OO
   cn

   o
   o

   CD
 

 to cu
 >..  X
    o
 O 3
  - O
-J. Q.
O  -"•
3  in


   ~i

   CT
   d
   rt-
   ro
   ex
 3

 rt-

 ro

 in
T3
 -5

 3*
to
 i
        SETCQ

        CU  O
        3  CZ
        a.-a
           in

        ro CL
  II  rt-
    -S
 ro ->•
  - cr
    c
 CU  rt-
 cr ro
i o  a.

 ro  cr
                to
                §
                >*
                Co
                Co
                            MEAN NATURAL  LOG  NUMBER INDIVIDUALS/SQUARE  METER
                                                           Ul
                                                                                  00
                                                                                            tl
                                                                                               o
                                                                                               53
                                                                                               ^
                                                                                               to

-------
Table 13.  Summary of mean numbers of individuals and species in commercial  clam bed recovery
           experiment at Discovery Bay for a 3-month spring-summer period in 1980.
NUMBERS/TRAY
OIL EFFECTS
CONTROLS

INDIVIDUAL NUMBERS
POLYCHAETES
CRUSTACEANS
MOLLUSKS
00
^ SPECIES NUMBERS
POLYCHAETES
CRUSTACEANS
MOLLUSKS
TOTAL
Mean
7.65
31.91
2.91
2.92
2.06
1.87
7.15
(S.D.)
(0.43)
(2.82)
(0.24)
(0.14)
(0.23)
(0.15)
(0.38)
OIL
Mean
3.
12.
1.
1.
1.
0.
3.
71
31
05
31
68
84
90
ED
TS.D.)
(0.43)
(2.82)
(0.25)
(0.14)
(0.24)
(0.15)
(0.38)
TIDE LEVEL EFFECTS
MLLW +2'ABOVEMLLW
Mean
10.86
40.56
2.20
3.36
1.78
1.34
6.76
(S.D.)
(0.43)
(2.82)
(0.25)
(0.14)
(0.24)
(0.15)
(0.38)
Mean
1.29
3.49
1.77
0.88
1.96
1.37
4.30
(S.D.)
(0.43)
(2.82)
(0.25)
(0.14)
(0.24)
(0.15)
(0.38)
STATISTICAL
SIGNIFICANCE1
Oil
Yes
Yes
Yes
Yes
No
Yes
Yes
Tide
Yes
Yes
No
Yes
No
No
Yes
1 Deemed statistically significant with alpha probability less than or equal to 0.05.  Based on
  35 cores distributed in 5 replicates of 5 trays each per treatment.

-------
0.05) effects  due to  oil  and tide  level.   There  were  significantly more
individual  polychaetes,  crustaceans,  and  mollusks  per  tray  in  control
sediments as  compared  to oiled.   There were  significantly more individual
polychaetes and crustaceans  at  the MLLW tide  level  as  compared to the +2'
tide level.  A  slightly  higher  (0.43 individuals per tray)  mean number of
individual mollusks at MLLW  as  compared to +2'  tide level was not statis-
tically significant.

     Total number of species  per tray and numbers of polychaete species per
tray were significantly  higher  in control  substrates as compared to oiled,
and at MLLW  tide  level  as  compared to +2'.   Numbers of species of mollusks
per tray  were significantly  higher  in control  substrates as  compared to
oiled.   A slightly  higher  (0.03  per tray) mean  number  of species per tray
at the +2'  tide  level  as compared to  the  MLLW tide level for mollusks was
not statistically significant.

Taxonomic and Trophic Composition

     A total  of  70  species were sampled in the commercial clam bed experi-
ment (Table  14).  Crustacean species were best represented with 25 species
(36%); followed by  polychaetes  with 23 species (33%); mollusks, 16 species
(23%); and  all  other  species, six species (10%).   This  can be compared to
MLLW controls  at Sequim Bay (Vanderhorst et  al.,  1980)  where polychaetes
represented  48%;  crustaceans,  32%;  mollusks, 7%;  and all  other species,
14%.  The basic  difference in composition relates to the greater number of
mollusk species at Discovery Bay.

     The  overall  trophic composition (Table 14)  derived  from Simenstad et
al. (1979) shows a dominance of detritivores, 23 species (33%); followed by
carnivores, 17 species (24%); herbivores, 14 species (20%); and suspension-
feeders,  nine species  (13%).   Eight  species  (11%) had  either reportedly
varied  trophic  classification,  or  did not fit  into  our  chosen four basic
groups.

     The  trophic  classification  differed  between taxonomic  categories.
Polychaetes  were  dominated  by  detritivores,  12  species  (52%); and carni-
vores, seven  species  (30%).   There was one herbivore among the polychaetes
and  no  suspension-feeders.   Three  species  had varied  or  other  trophic
classifications.  The  crustaceans  also had a predominance of detritivores,
nine species  (36%),  but  had the most  even  trophic  distribution of all the
taxonomic groups:  carnivores, four species (16%);  suspension-feeders,  five
species  (20%);  herbivores,  six  species  (24%);  and varied  or other, one
species  (4%).   Mollusks  had more herbivores,  seven species  (44%),   than
other trophic  groups.   This  amounted to 50% of all  herbivore species.  The
other  trophic groups within the mollusks were:   suspension-feeders,  four
species  (25%);  detritivores,  two  species  (13%);  and a  single carnivore
species   (6%).   Two  species  did  not  fall  in  the  four  basic  trophic
categories we had adopted.

     The  composition  of trophic  groups  expressed  as numbers of species
within the tide level and oil treatments is shown in Figure 31.
                                   82

-------
Table 14.  Species composition and trophic groups for commercial clam
           bed recovery experiment.1
TAXONOMIC GROUPS/SPECIES
                                             TROPHIC GROUPS2
POLYCHAETES
     Armandia brevis
     Axiothella rubrocincta
     Boccardia~proboscidea
     Capitella capitata
     Cap i te11i d undet.
     Cirratulid undet.
     Exogone lourei
     Glycinde armigera
     Goniadid undet.
     Halosynda brevisetosa
     Hemipodus' boreal is
     Maldanid undet.
     Nothria elegans
     Notomastus (Clistomastus) tenuis
     Ophiodromlis pugettensis
     Owenia fuslformis_(= collaris)
                (Anaifides) maculata
Phyllodoce
PlatynereTs
                 bicanaliculata
     PolycflaeTe undet.
     Polydora social is
     Protodoryillea gr'acilis
     Spio filicornis
      pionid undet.
CRUSTACEANS
     Allorchestes angusta
     Ampelisca pugetica
     Amphipod undet.
     Anisogammarus confervicolus
     Aoroides columbiae
     Balanus sp.
     Caprella sp.
     Corophium ascherusicum
     Cumella v'ulgaris
     Eualus townsendi
     Exosphaeroma amplicauda
     Exosphaeroma sp.
     Gammaropsis
                                             detritivore
                                             detritivore

                                             detritivore
                                             detritivore
                                             detritivore
                                             detritivore
                                             carnivore
                                             carnivore
                                             carnivore
                                             carnivore
detritivore
cam'yore
detritivore
carnivore
herbivore
varied
detritivore
carnivore
detritivore
detritivore
                                             detritivore
                                             detritivore
                                             various
                                             herbivore
                                             detritivore
                                             suspension
                                             herbivore
                                             detritivore
                                             detritivore
                                             carnivore
                                             herbivorus scavenger
                                             herbivorus scavenger
                                             suspension
                              83

-------
Table 14.  (Continued)
TAXONOMIC GROUPS/SPECIES
                                                  TROPHIC GROUPS2
CRUSTACEANS (Continued)

     Gnorimosphaeroma o. oregonensis
     Hemigrapsus nudus
     HeptacarpTis' paludicola
     Heptacarpus sp.
     LeptocheTTa dubia
     Mebalia pugettensis
     Parallorchestes ochotensis
     Paraphoxus sp.
     Pfiotis brevipes
     Photis sp.
     Pugettia gracilis
     Upogebia pugettensis

MOLLUSKS
     Acmaea sp.
     Alvania compacta
     Caecum occidentale
     Lacuna van' egata
     Littorina scutulata
     Littorina' sitkana
     Macoma inquinata
     Macoma sp.
     Margarites pupil 1 us
     Mysella fumida
     Myti1 us edulis
     Flassarius~mendicus
     Notoacmea' persona
     Protothaca stami nea
     Solariella sp.
     TransenneTla tantilla

OTHER SPECIES

     Amphipholis sp.
     Leptosyna'pta clarki
     Nemertea undet. (sp.  A)
     Nemertea undet. (sp.  B)
     Paranemertes peregrina
     Sipunculid undet
                                                  herbivorous  scavenger
                                                  carnivore
                                                  carnivore
                                                  carnivore
                                                  detritivore
                                                  suspension
                                                  detritivore
                                                  detritivore
                                                  suspension
                                                  suspension
                                                  herbivore
                                                  detritivore/suspension


                                                  herbivore
                                                  herbivore

                                                  herbivore
                                                  herbivore
                                                  herbivore
                                                  detritivore
                                                  detritivore
                                                  herbivore
                                                  suspension
                                                  suspension
                                                  carnivore
                                                  herbivore
                                                  suspension

                                                  suspension
                                                  carnivore
                                                  carnivore
                                                  carnivore
                                                  carnivore
1 Experiment at Discovery Bay, MLLW and +2' above MLLW with 3-month
  colonization during spring and summer, 1980.

2 Trophic classification from Simenstad et al. (1979).

                              84

-------
00
en
      CO
         20  •



         18  --



         16  ••


         14  •
      to
      w
      u  12
      w
    10


     8  -


     6  --



     4  -


     2  --
                                                                          CONTROL
                                                                               OILED
                           + 2
                                      + 2
        + 2
   0'   +2
                  DETRITIVORES
                               CARNIVORES
HERBIVORES
SUSPENSION
Figure 31
                 Number of species within trophic groups  in  commercial clam bed recovery experiment.
                 Number of species is aggregate in 35 cores  distributed in 5 replicates per condition.
                 (0 = MLLW;  +2' = 2' above MLLW.)  Experiment  was at Discovery Bay for 3 months during
                 the spring-summer season, 1980.

-------
Detritivores were  much better  represented  at MLLW,  in controls,  than  at
+2'.   There were also eight fewer species in oiled sediments as compared to
controls at MLLW.   The number of species of detritivores in oiled sediments
at +2'  exceeded that  for  controls by  one  species.   Carnivores  were also
much better represented at MLLW than at +2'.  There  were  fewer species in
oiled sediments at  each of the tide levels.   Numbers  of species of herbi-
vores in  controls  were equal  (seven  species) at each  of  the  tide levels.
There were  fewer  (four species)  in oiled  sediments  at MLLW  and one more
species in  oiled  sediments than controls at  +2".   Suspension-feeders were
represented by  a greater  number (total  six species) of species at MLLW and
in control sediments at both tide levels.

     The density of trophic groups related to oil treatment and tide level,
expressed as  the  natural   logarithm of  numbers  of individuals per square
meter, is shown in Figure  32.  Detritivores  had higher density at MLLW than
at +2'  tide level  and higher density in control  sediments compared to the
respective  oiled  sediments.   The  same  trend was  true for  carnivores  and
suspension-feeders  although  for  carnivores  the  tide  level  difference
between controls was  minimal.   Herbivore densities were roughly equivalent
in all of the  tide level  and oil treatment categories.

Primary Species

     Tray  densities  for   the  ten  primary  species which were  a  priori
selected for testing of oil effects hypotheses are shown in Table 15.   With
the exception  of Lacuna sp. (MLLW, +2'  tide levels), Corophium ascherusicum
(+2'  tide level),  and Platynereis bicanaliculata (MLLW), all mean densities
were equal to  or higher in controls than in oiled sediments.

     In control sediments,  there  were  higher densities at  MLLW than at +2'
for all species except the littleneck clam (Protothaca staminea).   In oiled
sediments there were  higher densities  at the +2' tide level for Protothaca
staminea. Corophium  ascherusicum, and  Leptochelia  dubia.   Photis brevipes
never occurred in oiled sediments.

     In  control  sediments,  _L.  dubia  far  exceeded all other species  in
density.  Exogone lourei  was the second highest species in  density at MLLW.
These two  species  were identified as  particularly  good  subjects  for  the
experimental indication of oil  treatment effects on recovery of infauna in
our region  based on Sequim Bay and Protection  Island  data (Vanderhorst et
al.,  1980).   Their high  density  and the  differential in  density between
oiled  and  unoiled  sediments tends to  confirm  their  value for  such use.

     The overall density  for primary  species was much higher (498 individ-
uals per tray) for this experiment as compared to the equivalent experiment
at Sequim Bay in 1979 (54.4 individuals per tray).  Most of this difference
is  attributable to  the  density  of Leptochelia dubia although  there  was
higher  density in  MLLW  controls  for  each of  the primary  species except
Photis brevipes and Polydora social is.
                                   86

-------
        to
        c
        -s
        fD

        GO
        ro
a cr o :z
—'• CU —* Qi
to  m 01 c-f-
O   3 C
o  a.    -s
<    cr QJ
fD  O fD —•
-S  3 O-
DO in
OJ
<< O
   O
-h -S
O fD
~S tf>

OO Q-

3 

3 -S

3- cr
in c.

a. fD
c a.
      -s o
      CD ua
      o
      o o
      < -h
      fD
      ro cr
      x tt>
     "a -s
      fD
      -s o
        -h
      3 3
      c+ D-
3
(O

rt-
3- -S
n> a>
  •a
(/> — '
-a -j-
-s o
     o c

      H  —'


     f— in
in
   TJ
   CD
n>
      + fD
      ro
      II Cf
        fD
     ro -s
   O
   o

   CL
c/>
fD
O>
C/l
O
3 O  S O
»  3  I~T3
      O 3"
      <
      fD cf
                          MEAN  NATURAL LOG INDIVIDUALS/SQUARE  METER
                                     en
                                                     oo
                 to
                 H
                to
                fa
                to
                55!
                H
                 to
                     Ni
                 to
                 H
                 ^!

                 §
                 to
                 C3
                 Co

                 a
                 H
                 Q
                                                                      O
                                                                      H
                                                                      t-i
                                                                              Q
                                                                              O
                                                                              53
                                                                              ^3
                                                                              to
                                                                              O
                                                                              in

-------
Table 15.  The mean density of primary species in commercial clam bed
           experiment (tide level, oil, three-month recovery summary).
                                      MEAN NUMBERS/TRAY
                                 MLLW             +2'  Above MLLW
PRIMARY SPECIES Control
MOLLUSKS
Mysella tumida
Protothaca staminea
Lacuna sp.
CRUSTACEANS
Corophium ascherusicum
Photis brevipes
Leptochelia dubia
POLYCHAETES
Armandia brevis
Exogone lourei
Platynereis
bi canal iculata
Polydora social is

12.9
2.9
1.8
11.2
0.3
405.3
10.5
47.6
4.2
1.7
Oil
4.0
0.1
2.2
0.4
0
6.8
3.5
19.4
6.4
0.6
Control
4.0
3.3
0.2
0.0
0.2
137.8
0.0
0.6
0
0
Oil
1.2
2.6
0.2
1.4
0
11.2
0.0
0.3
0
0
Means based on 5 replicates per condition (7 cores per replicate tray),
Analyses of variance for tide and treatment effects on Table 16.
Experiment was at Discovery Bay during a 3-month spring-summer period,
1980.
                              88

-------
     Hypothesis  tests for  the  main effects  attributable to  oil  and  tide
 level are shown  in Table 16.  Statistically significant effects  due  to  tide
 level were  demonstrated for seven of the ten species.  The exceptions  were
 Protothaca  staminea,  Corophium  ascherusicum,  and Photis  brevipes.   This
 indicates  a  much  larger  tide  level effect  in this  experiment where the
 comparison  was  between MLLW and +2' as  compared to the  equivalent  experi-
 ment  at Sequim  Bay  (Vanderhorst  et a!.,  1980) where the  comparison was
 between  MLLW and -2'.   In that experiment  significant  tide level  effects
 were demonstrated for  only  four of ten species.

     Significant  effects  due  to oil  treatment were  demonstrated for the
 densities of four  of the ten primary  species:   Mysella  tumida, Protothaca
 staminea,  Leptochelia dubia, and  Armandia brevis.   For  comparison, there
 were also  four  of the 10  species  having significant effects due  to oiling
 in the equivalent Sequim Bay experiment  (Vanderhorst et al., 1980).

 Species With  Indicated Oil  Treatment Effects

     For reasons  stated  in the methods  and in Vanderhorst et al.(1980), we
 used a  priori selected  species  for hypothesis  tests to  ensure a  conser-
 vative estimate of oil treatment effects.  However, analyses of  variance of
 the density for  all  70 species  in  this  experiment were  computed.   Statis-
 tically  significant  (P = 0.05) effects  on  density attributable to  the oil
 treatment were  indicated  for nearly a third  (21)  of the 70 species (30%).
 These species, with  attendant trophic  designation  are listed  on Table 17.

     It  is  of interest  to compare the  contribution  of  species to  trophic
 and taxonomic groups  as  related to oil treatment.  We  have  done this in
 index form  as follows.  For  the  denominator  of the  index we  use the per-
 centage  contribution  based on numbers  of species  in  taxonomic and  trophic
 groups  (Table 18).   For the  numerator  of the  index, the  percentage  the
 trophic  or  taxonomic group contributes  to the total of 21 species on Table
 17 is used.   An  index value of 1.0 indicates that the taxonomic or  trophic
 group was influenced by oil treatment on par with all species.   An index of
 greater  than  1.0  indicates that the group was  more susceptible to the oil
 treatment than  were all  species.   The  appropriate percentages and ratios
 obtained are shown  in Table  18.    In   terms  of  composition,  polychaetes
 (index 1.41) and moll usks (index 1.09)  were more severely influenced by the
 oil treatment than  all  species,  and crustaceans  (index  0.71)  and "other"
 species  (index  0.50)  were  less  severely  influenced  by oiling.   For  the
 trophic  groups,  detritivores  (index 1.45)  and suspension  feeders  (index
 1.54)  and "other" species (index 1.25)  were more severely influenced by the
 oil treatment, and carnivores (index 0.63) and herbivores (index 0.25) were
 less severely influenced by the oil treatment.

 Petroleum Hydrocarbon Data

     Time series  data on total  oil  measured  by IR,  and  initial  and final
concentrations of compound classes  analyzed  by capillary GC  are  shown in
Figure  33.    The  compound  class  composition is identical  to that previously
described for the hard substrate experiments (Table 12).


                                   89

-------
Table 16.  Hypothesis tests for density of primary species in commercial
           clam bed experiment at Discovery Bay during a 3-month period
           in 1980 (spring-summer, tide, oil).


                       PROBABILITY FOR ERROR IN REJECTING THE HYPOTHESIS

PRIMARY SPECIES                    Tide Level1            Oil2
MOLLUSKS
Mysella tumida
Protothaca staminea
Lacuna sp.
CRUSTACEANS
Corophium ascherusicum
Photis brevipes
Leptochelia dubia
POLYCHAETES
Armandia brevis
Exogone lourei
Platynereis bi canal iculata
Polydora social is

0.000*
0.054
0.002*
0.115
0.082
0.000*
0.000*
0.000*
0.004*
0.003*
0.000*
0.014*
0.700
0.134
0.561
0.000*
0.000*
0.018
0.540
0.135
1 The tide level hypothesis is:  Density in trays at MLLW  is equal to
  density in trays at +2' above MLLW.

2 The oil hypothesis is:  Density  in trays receiving oil treatment is
  equal to density in trays not receiving oil treatment.

* We reject the  indicated hypothesis with a maximum real probability
  for error of 10%.
                               90

-------
Table 17.  Species with trophic designation for which statistically
           significant (P = 0.05) oil treatment effects were computed.1
TAXONOMIC GROUP/SPECIES
                                               TROPHIC GROUP2
POLYCHAETES
     Armandia brevis
     Axiothella rubroci ncta
     Boccardia proboscidea
     Exogone Tourei
     HemTpodus boreal is
     Notomastus (ClisTomastus) tenuis
     Owenia fusiformis
     Protodorvillea gfacilis
     Spio filicornis
CRUSTACEANS
     Gnorimosphaeroma o. oregonensis
     Hemigrapsus nudus
     LeptocheTTa dubia
     Photis brevipes
     Upogebia pugettensis
MOLLUSKS
     Caecum occidentale
     Macoma sp.
     Mysella tumida
     Protothaca staminea
     Transenella tantilla
OTHER SPECIES
     Leptosynapta clarki
                                                detritivore
                                                detritivore

                                                detritivore
                                                carnivore
                                                detritivore
                                                detritivore
                                                carnivore
                                                detritivore
                                                herbivore
                                                carnivore
                                                detritivore
                                                suspension
                                                detritivore
                                                detritivore
                                                suspension
                                                suspension
                                                suspension
1 Statistical  significance computed from analysis of variance.   The
  number of analyses precludes rigorous statistical  evaluation.   See
              text for explanation.   Experiment was  at Discovery Bay
methods and
during a 3-month spring-summer period,
                                         1980.
  Trophic classification from Simenstad et al.  (1979).
                              91

-------
Table 18.   Contribution to taxonomic and trophic groups overall and
           in terms of significant oil treatment effects.1' 2
TAXONOMIC/TROPHIC GROUP
TAXONOMIC
Polychaetes
Crustaceans
Mollusks
Other Species
TROPHIC
Detritivores
Carnivores
Suspension Feeders
Herbivores
Other Species
PERCENT
Oil Treatment

45
25
25
5

45
15
20
5
15
CONTRIBUTION
Overall

32
35
23
10

31
24
13
20
12
Ratio

1.41
0.71
1.09
0.50

1.45
0.63
1.54
0.25
1.25
1 The index of severity is computed as follows:   Percentage contribution
  of taxon or trophic group to those species having "significant" oil
  treatment effects (Table 17) is divided by the percentage contribution
  of the taxon to numbers of individuals as a whole.

2 Based on a 3-month experiment at Discovery Bay during a spring-summer
  period, 1980.
                              92

-------
       3000 -
    a
    a,
CO
       2000
       1000 -•
         Q-
         Q.
                                                              §
                                                              H
         EH
         S5
         H
         O

         g
         O
         8
         o
                                                              o
                                                              u
                       TOTAL OIL

                         IR
                                                     NO TIDE
                                                     MLLW
+2'  TIDE
                                                                 110 T
                                                                 100
                                                                  90 •-
                                                                  80 "
             70 -•
60"
                                                                  50"
                                                                  40--
                                                                  30"
                                                                  20 --
                     May   Aug.

                     SATURATES
                       May   Aug.

                       AROMATICS
                   GC
    Figure 33.  Time series of total oil and analyzed saturate and aromatic  compounds in commercial
                clam bed recovery  experiment at Discovery Bay, May through August,  1980
                (IR = infrared spectrophotometry; GC = capillary chromatography).

-------
     The mean initial  concentration of total  oil  in this experiment (nearly
2500 ppm) was more  than  twice as high as  the  equivalent summer experiment
at Sequim Bay, and  400 ppm higher than the  fall  and long-term experiments
in that  group (Vanderhorst et  al.}  1980).  Final concentrations  in  total
oil were 80% at MLLW and 95% at +2'above  MLLW of  the  initial  concentrations
(Figure 33).  This  is  in sharp contrast  to the findings for  Sequim Bay and
Protection Island sediments in  which total oil concentrations were reduced
by half  in  a similar  period.   The intermediate concentrations  in  June and
July (Figure  33)  indicate higher than initial concentrations.   These con-
centrations were based on a single core analysis  at each of the tide levels.

     The analyzed saturate  compound  concentrations shown in  Figure 33 were
also higher  initially  than  for previous  experiments, and  the loss was less
over the three-month period (26% for MLLW,  and 20% for +2').   In Sequim Bay
and  Protection  Island sediments,  the  loss of saturates was  about  half in
three months  and was  not dependent on  initial  concentration  for  rate of
loss.  The analyzed aromatic compounds had  an initial mean concentration of
17 ug/g.   During the  three months,  no reduction  in  mean  concentration of
analyzed aromatics  occurred at  MLLW.   The  reduction  at +2' was  20%.   In
earlier  experiments  at Sequim  Bay  (Vanderhorst  et al., 1980)  the  loss of
analyzed aromatics ranged from 80 to 92% in three months.

     Mean  total   oil   concentrations  at  the  conclusion of  the clam  bed
experiment are given by the tide level  and  vertical core stratum, in Figure
34.  For oil-treated   sediments,  higher  mean  concentrations  are shown for
the  +2'  tide level as compared to MLLW.   Higher  concentrations are  shown
for the bottom half of cores as compared to the top half.  In contrast, the
untreated sediments show higher background  CCU extractable organic concen-
trations at  MLLW  as compared  to +2' and in  the  top half  of  cores  as com-
pared to the bottom half.

     "Main  effects" from  oil  treatment,   tide level effect,  and  vertical
core  strata, are shown  in Figure  35.   Analyses  of  variance  of  the data
resulting  in these means  indicate that the effects of  oil  treatment and
vertical  distribution  in cores were statistically significant  (P  - 0.05),
and  that the  effect  of  tide  level  was  not a  statistically  significant
effect.

     The  tide level  and  vertical  core stratum   distribution  of  analyzed
saturate  and  aromatic compounds  in  cores  is  shown  in  Figure  36.   For
saturate compounds  in  oil-treated sediments, the tide level distribution is
not consistent in the  top and bottom half of cores.  For the bottom half of
cores,  a higher  concentration is shown at +2'  as compared to MLLW, and for
the  top  half of cores, a higher concentration is shown at MLLW as compared
to +2'.  At  both tide  levels, however, a higher mean concentration is shown
in the bottom half  of  cores as compared to the top half.  Analyzed aromatic
compounds in oil-treated  sediments also show slightly higher concentrations
in  the  bottom  half  of   cores  as compared  to top.   The  concentration of
analyzed aromatic compounds are  higher at MLLW in the top half  of cores and
about equal  between tide  levels  in the bottom  half of cores.
                                    94

-------
en
     a
     o
                                                        NOTE:
                                                  •Different Scales
                                            B
                                                               TOP
                                                                      10--
                                                               BOTTOM
                                                                              T
                                            B
B
                    MLLW
+ 2
                                                                               MLLW
                                                          + 2
                              OILED
                                                                                       CONTROLS
          Figure 34.   Vertical  stratification of total oil concentration in cores  for commercial  clam
                      bed  recovery experiment at Discovery Bay, spring-summer season, 1980.   Samples
                      were taken in August 1980 at the conclusion of the experiment and  measured  using
                      infrared  spectrophotometry (Top = upper 5 cm of core; Bottom - remainder  of core),

-------
H
o
1
2200 -•


2000


1800 '•


1600 '•


1400 "


1200 -'


1000 -


 800 '•


 600 '•


 400 '


 200 •-
                                        OIL
                  (YES)
   T =  Treated
   C =  Control
TIDE

   0'=  MLLW
  +2'=  +2' above  MLLW

VERTICAL

   T =  Upper 5  cm of Core
   B =  Remainder
                                                  (YES)
                                 (NO)
                                     + 2
                                                    B
                  OIL
                                TIDE
       VERTICAL
Figure 35.
                               TOTAL OIL
        Mean total  oil concentration due to oil treatment, tide level,
        and vertical stratification.  Word in parentheses indicates
        statistical significance.
                                 96

-------
                                      Z6
      (Q
      n>
      co
      cr>
                      MEAN  HYDROCARBON CONCENTRATION
o
    CU O>
  I/)
  fD
                      .MEAN  HYDROCARBON  CONCENTRATION
                            O
                            *

                            to
 o
 *
 Ul


-f
         O

         555
            +
            to
                                                             O
                                                             en
                      n
                      «+
                      
                      09

-------
     It should be noted that the data on untreated controls shown in Figure
36  are  on  a  scale  two  orders  of magnitude  lower  than  the data  for
oil-treated sediments.  The  data reflect the normal background variability
for these compound classes.

     A  summary  of analysis  for saturate and aromatic  compound  class data
is  shown  in  Figure  37.   For saturate compounds,  the  effect of treatment
was, of  course,  significant  (P = 0.05).  The slightly  higher mean concen-
trations for the  +2'  tide level and bottom  half  of cores compared to MLLW
and top  half  of  cores were not significantly different.  Analyzed aromatic
compound concentrations were significantly higher in oil-treated and bottom
half of cores  compared to unoiled and top half of cores.  The difference in
mean concentrations at the two tide levels was not significant.

Sediment Grain Size

     Analyses  of  sediment grain size were performed on 11 cores taken from
experimental trays at the Discovery Bay Clam Bed (MLLW).   Additionally, 10
cores  from  experimental   trays  at  the  Protection Island  site  and 12 cores
from trays  at  the Sequim Bay experimental site (MLLW, Vanderhorst  et al.,
1980)  were  analyzed.   Results  from these analyses are tabulated below in
terms of percentage weight contribution by grain size fractions:
GRAIN SIZE FRACTION
SCREEN SIZE (mm)
GREATER THAN
5.66
2.00
1.00
0.50
0.125
0.063
PAN
SEQUIM BAY
Mean (S.D.)
11.18 (3.49)
19.54 (3.03)
17.35 (3.96)
25.38 (2.72)
24.09 (5.10)
2.05 (0.64)
0.40 (0.26)
PROTECTION ISLAND
Mean (S.D.)
0.87 (0.53)
3.59 (0.83)
3.27 (1.00)
6.53 (1.58)
83.71 (2.99)
1.83 (0.27)
0.21 (0.05)
DISCOVERY BAY
Mean (S.D.)
52.53 (3.33)
15.08 (1.43)
7.99 (0.73)
10.30 (0.71)
12.30 (0.89)
1.32 (0.17)
0.47 (0.07)
EFFECTS OF OIL AND KEY SPECIES REMOVAL ON HARD SUBSTRATE COMMUNITIES
AND COMMUNITY RECOVERY

A Perspective

     The  data presented  in this section differ  from  those in the two pre-
ceding   sections  in  that  bricks  were allowed to colonize  for  a period of
nine  months  (September  1979  through May 1980)  before  any treatments were
applied.  The three independent experiments reported had the primary objec-
tives  of evaluating  the effect of  oil  treatment on  existing associations
                                    98

-------
and the recovery of those associations, and the effect of a differential in
grazing  pressure  on  the associations.   For  experiments in  the preceding
sections,  the  nature  of the effects due oil treatment was clear, i.e., the
effect  of the oil treatment,  if any,  was to alter  the  suitability of the
substrate  itself.   This  was measured  by  comparing treated  and untreated
substrates given  equal  field exposure conditions (through randomization of
multiple  units,  Figure  37).   In an attempt to  distinguish  between effects
on  the associations  from the  treatments,  and  effects  related to  a dif-
ferential  in  recovery   between treated  and  untreated  bricks, the  three
experiments  reported  in this section were treated in analyses of variance.
Using  duration  of field exposure  after  oil   or   grazer  treatment  as  a
"treatment"  in the experimental design model:


                Yijkl  = Di + °j + Gk + Tl + E

     where     Y = the response variable magnitude;

               D = the main effect due to duration of field exposure
                   (i  = 0, 5, or 30 days);

               0 = the main effect due to oil  treatment (j = oil-treated
                   or untreated);

               G = the main effect due to grazers (k = limpets stocked
                   or removed from associations);

               T = the main effect due to field exposure tide level
                   (1  = MLLW or +2' above MLLW); and

               E = random error.

The response variables  evaluated in analyses  of variance were  of  the same
types as  those  in  the other hard  substrate experiments,  and,  in addition,
included dry weight biomass of algae.   This latter response variable was of
particular interest because  of an expected relationship  between algae and
grazers.

     The  species  composition  for these  experiments  overall  is  shown  in
Table 19.  All  experiments  were conducted at Sequim Bay.   For  the present
experiments,  a total  of 95  species are shown  in Table  19.   For the monthly
experiments overall,  including  the site experiments, a total  of 79 species
were shown.  Thus, the  number  of species for  these  precolonization experi-
ments  is  higher than the accumulated  total  for the 12  other  experiments.
In fact,  this  is  a  very conservative  estimate of the  true difference  in
structure because the aggregate sample size contributing to the 95 species
in the present  experiments was  a total of 120  bricks  giving an average of
0.8  species  per brick,  and  the  aggregate  sample size  in  the  12  monthly
experiments was 840 bricks  for  an average of  0.09 species per  brick.   This
approximate order of  magnitude  difference  in  numbers of species per  brick
reflects the importance  of  community structure in the present  experiments
since the  total  available colonization  time  (10 months) and the  specific
10-month period were  identical  for each of the two cases.

                                   99

-------
                                               001
                                MEAN  CONCENTRATION   (ppm)
      fD
      CO
 O to CO
 ai ~a 01
•a -s ct
 —•3
 — HQ
 D>
   (/> fD
   C O.
O> fD 3
V) -5 D.

O t/1 Oi
3" fD -S
-S Oi O
O  3
3 O 01
Ol 3 cH-

O    o'
IQ I—>
-s to o
O> CO O
-003
3- •  XJ
*<    O
      c
   CO 3
   O) Q.
S 3 to
o -a
-s —• -h
a. fD -s
   to o
   fD O
•a -s o
O> fD 3

fD r+ fD
3 O> -S

3- fD -••
fD 3 O)
to    —i
fD -••
tn 3 O
Co
!fc>
•-a
c:
to
       n
                      to
                     T
                     -h
                     fD

                     n>
                     CO
                     O
                     O)

                     fD
                          to
                          o
                                                co
                                                o
                                                                Ul
                                                                o
en
o
CO
o
                                 MEAN  CONCENTRATION   (ppm)
                                                        CD •«
                                                        3t)
                                                        PJ fB
                                                        H- H

                                                        DJ Ul
                                                        (D
                                                        H Q
                                                           3

                                                           O
                                                           f-h

                                                           n
                                                           o
                                                           1-5
                                                           (D
                                                                           CO
                                                                                  to o
                                                                              H   - -
                                                                          i-3
                                                                          H
                                                                          D
                                                                              s
                                                                                  II  II

                                                                                  +  S
                                                                                  cr
                                                                                  o
                                                                                  <
                                                                                  fD

                                                                                  g
                     O

                     II

                     O
                     O

                     rt

                     O
                                                                                                   O
                                                                                                   H
                                                                                                   IT"
        II

       "-3

       (D

       ft
       (D
                                                                O   H   to  U)   *•   Ui
 -•• c cr
 o en fD
 o> CH- o.
 0+
 fD I—' fD
 W W3 X
   c»-a
 to o fD
 n-   -s
 a* a* ->•
 c+ 3 3
 -1. Q. fD
 to    3
 n- 3 c+
 -J. fD
 o o> o>
 O) «/> r+
 —" C
   -s a
 W fD -"•
 -J- Q. to
03    O
 3 C O
 -"• V) <
 -h -"• fD

 o'lQ <<
 O>
 3    CO
 O    O>
to
O
is:
^
    o

-------
Table 19.   Taxonomic and trophic composition in grazer experiments.
SPECIES/TAXONOMIC GROUP
TROPHIC GROUP2
POLYCHAETES
Anai tides groenlandica
Anai tides williamsi
Anai tides sp.
Armandia brevis
Axiothella rubrocincta
Boccardia proboscidea
Capitella capitata
Cirratulus cirratus
Dorvillea sp.
Eulalia nigrimaculata
Eulalia sp.
Exogone lourei
Halosynda brevisetosa
Harmothoe imbricata
Lumbri nereis sp.
Maldanidae undet.
Nephthys caecoides
Nereis vexillosa
Nothria elegans
Opheliidae undet.
Ophiodromus pugettensis
Phyllodocidae undet.
Platynereis bi canal icul ata
Polychaeta undet.
Polydora social is
Polydora sp.
Protodorvillea gracilis
Spio filicornis
Spionidae undet.
Thelepus crispus
Thelepus sp.
CRUSTACEANS
Amphipoda undet.
Ampithoe simulans
Ampithoe sp.
Anonyx sp.
Aoroides columbiae
Balanus cariosus
Balanus glandula
Balanus sp.
carnivore
carnivore
carnivore
detritivore
detritivore
suspension
detritivore
detritivore
carnivore
carnivore
carnivore
detritivore
carnivore
carnivore
herbivore
detritivore
carnivore
herbivore
herbivore
detritivore
carnivore
carnivore
herbivore
varied
detritivore
detritivore
carnivore
detritivore
detritivore
detritivore
detritivore
varied
herbivore
herbivore
detritivore
detritivore
suspension
suspension
suspension
                              101

-------
 Table  19.   (Continued)
SPECIES/TAXONOMIC GROUP
TROPHIC GROUP2
CRUSTACEANS (Continued)
Caprella laeviuscula
Caprella sp.
Ceradocus spinicaudus
Corophium sp.
Expsphaeroma sp.
Fabia subquadrata
Hippplyte sp.
Idothea wosenesenski i
Ischyrpcerus sp.
Jassa falcata
Leptochelia dubia
Leptochelia sp.
Maera sp.
Melita californica
Nebalia pugettensis
Pagurus hirsutiusculus
Pagurus sp.
Parallorchestes ochotensis
Paraphoxus sp.
Photis sp.
Phoxocephalidae undet.
Pinnixia faba
Pinnixia tubicola
Pinnixia schmitti
Pinnixia sp.
Pontogeneia inermis
Pugettia gracilis
MOLLUSKS
Acmaea digitalis
Acmaea persona
Acmaea sp.
Alvania sp.
Bittium sp.
Caecum occi dental e
Cyanoplax hartwegii
Lacuna sp.
Littorina sp.
Margarites pupi 1 lus
Margarites sp.
Mitrella sp.
Mopalia muscosa
Mysella tumida
My til us edulis
herbivore
herbivore
detritivore
detritivore
herbivore
carnivore
herbivore
suspension
detritivore
detritivore
detritivore
detritivore
detritivore
suspension
detritivore
detritivore
detritivore
detritivore
suspension
detritivore
other
other
other
other
detritivore
herbivore
herbivore
herbivore
herbivore
herbivore
carnivore
herbivore
herbivore
herbivore
herbivore
herbivore
carnivore
suspension
suspension
                              102

-------
Table 19.  (Continued)
SPECIES/TAXONOMIC GROUP                           TROPHIC GROUP2


MOLLUSKS (Continued)

     Odostomia sp.
     Protothac'a staminea                          suspension
     Searlesia~dira                               carnivore
     Searlesia sp.                                carnivore
     Transenne'lla tantilla                        suspension

OTHER SPECIES

     Amphipholis sp.                              detritivore
     Emplectoneiiia gracile                         carnivore
     Evasterias trpschelli                        carnivore
     LeptasterTas hexactis                        carnivore
     giigocottus sp.                              carnivore
     Paranemertes peregrina                       carnivore
     Paranemertes sp.                             carnivore
     Pholis laeTa
     Pholls sp.


1 This composition results from analysis of all bricks used in the grazer
  experiments (N = 12).   Bricks had been field exposed at Sequim Bay
  from September 1979 through June 1980.  Both MLLW and +2 MLLW tide
  levels are represented.

2 Trophic categories from Simenstad et al.  (1980); (-) indicates no data.
                                  103

-------
     Thus, although the compositional  list (Table 19) is simple in terms of
other  sampled rocky  intertidal  communities  (Nyblade,  1979)  reported  in
excess of 900  species  for a combination of several  Strait of Juan de Fuca
stations), effects from  treatments  resulting  in a reduction of  numbers of
species or individuals  in the  present experiments will reflect a breakdown
in community complexity, a stated task objective.

Taxonomic and Trophic Composition

     A breakdown of  the  data on  Table 19 indicates the 95 species were: 31
polychaetes  (32%);  35   crustaceans  (38%);  20  mollusks  (21%);  and  nine
species  not  belonging  to the  major taxonomic  groups (9%).  For  trophic
categories overall (using Simenstad  et  al. (1979) for classification), the
95   species   were:    29  detritivores  (31$);  23   carnivores   (24%);  11
suspension-feeders (11%);  20 herbivores (21%); and  12  species  not fitting
any of these categories (13%).

     The  taxonomic  groups  differed  in  trophic  composition:    among the
polychaetes,  13  species  (42%) were  detritivores;  12 species  (39%)  were
carnivores;  four  species  (13%) were  herbivores; and  one  species each (3%
each)  were   suspension-feeders  or species  not  fitting the trophic  cate-
gories.  For crustaceans, 16 species (44%) were detritivores; seven species
(19%) were herbivores;  six species  each (17% each) were suspension-feeders
or did  not  fit the trophic classification; and only one species (3%) was a
carnivore.   Moll usks had no  detritivores;  nine herbivores (45%); four each
of  carnivores and  suspension-feeders  (20%  each);  and three  species not
fitting the  trophic  categories (15%).   The species outside the major taxo-
nomic  categories   included  six carnivores (67%);  two  species  outside the
trophic  classification   (22%);  one  detritivore  (11%); and  no  suspension-
feeders or herbivores.

Treatment Effects - Field Exposure Time, Oil  and Grazers

     The  main effects  means  for  numbers of  species  in  taxonomic groups
related  to  duration of  post-treatment  period  are  shown  in  Figure 38.  A
statement  of  statistical  significance  (P =  0.05)  of  differences between
these means  is indicated  in parentheses over each comparison.  Total number
of  species  and numbers  of species  in each of  the  taxonomic groups except
crustaceans  were  less  in the  0-day field exposure  experiment  than in the
30-day  field exposure.   For polychaetes and total  species the differences
between  field exposure  periods were  significant  and indicate colonization
during  the  30-day post-treatment  period.  The  intermediate  5-day   field
exposure  experiment  showed  fewer mean  number  of  species  than 0-day  field
exposure  for  total  species,  crustaceans,   and  mollusks.   Although the
statistical  analyses  used  do not  permit assignment of  probability for
significance of  these  differences,  they do not appear  to  be of a magnitude
which would  be deemed significant if treated independently.

     Mean  number  of  species per brick  for oil  and grazer main effects are
shown  in  Figure  39.   Statistical significance of differences between  group
means  (P = 0.05) is  shown in parentheses above the means compared.


                                   104

-------
      o
      H
      CO
      H
      H
      O
      W
      CM
      03
   12 +



   11



   10



    9
      «  8 -•


      Ai
          7 •-
    6 •-
w   5


    4 •



    3 -


    2 •


    1 -
                   (YES)
                                  (YES)
                                                                (NO)
                                                  (NO)
                   0  5 30
                           0  5  30
                                                0  5  30
                                                        0   5 30
                    TOTAL      POLYCHAETES   CRUSTACEANS   MOLLUSKS
Figure 38.   Mean  number of species per brick related to duration of
            experiment.  Data summarized over two tide levels  (MLLW and
            +2' above MLLW); two oil treatments (oiled and  unoiled);
            and two grazer treatments (limpets stocked and  limpets removed)
            (N =  120 bricks.)
                                    105

-------
14 f   (YES)



13


12


11


10
 PQ

 K
 W
 H
 H
 u
 H
 04
 CO
 PQ

 g
    9  ••
7 -•


6 ••


5 '


4


3


2


1
                (NO)
                         (YES)
                                                             (YES)
                                            (YES)
                                                   (NO)
                                                                    (NO)
          C  T  S  R

         Oil  Graz.
            TOTAL
                        C  T  S  R

                        Oil  Graz.

                       POLYCEAETES
                                         C  T  S  R

                                        Oil  Graz.

                                       CRUSTACEANS
 C T  S  R

Oil   Graz.

 MOLLUSKS
Figure  39.
        Mean number of species per brick summarized by main effects  of
        oil treatment and  grazer treatment of experiments at Sequim  Bay,
        May and June, 1980 (C = control; T = treated; S = limpets
        stocked; R = limpets removed).  (N = 120  bricks.)  Word in
        parentheses indicates statistical  significance.

                               106

-------
Significant  effects  on   numbers  of  species  due  to  oil  treatment (C  =
control; T =  oil-treated, Figure 39) for each  of the  taxonomic groups and
total number  of  species  were demonstrated.   In no case was a statistically
significant  effect  demonstrated for  the  grazer treatment  (S =  limpets
stocked; R = limpets removed).

     The mean  number of  species  for tide level main effects  are  shown in
Figure 40.  There  is a slightly higher mean number of species per brick in
all  categories  at the MLLW  tide  level.   In no case was  the  difference in
number of species per brick due to tide level deemed significant.

     A comparison of mean numbers of individuals per brick within taxonomic
and  trophic  categories related  to  duration of field exposure  is  shown in
FigureK41.  A  statement  of statistical significance (P = 0.05) is shown in
parentheses  above  each   comparison  of  means.    There  was  a  higher  mean
density  at  30 days  post-treatment  compared to  immediately post-treatment
for  each of the  taxonomic and  trophic  groups.   Differences  between  mean
densities due to  duration of period post-treatment were demonstrated to be
statistically  significant   for   crustaceans,   mollusks,   herbivores,   and
suspension-feeders.  It  is  of  interest that the largest mean difference in
group density (detritivores, about 40 individuals per brick difference) was
not  significant.   This is either due to  high error  variance  or some other
factor  in  the  model contributing  to  large differences  in the  number of
detritivores.    The  intermediate (5-day  post-treatment)  mean numbers  of
individuals per  brick were higher than immediate post-treatment  for crus-
taceans, mollusks, herbivores,  and detritivores.   Although the significance
of the  differences in these means  is  not tested in the  model,  it appears
that the magnitude  of difference in mean number of mollusks and herbivores
would be significant if independently tested.

     The mean  densities   for main effects due  to tide level  are  shown in
Figure 42.  Statistically significant  tide  level  effects were demonstrated
for  polychaetes, crustaceans,  and mollusks  (P,  C, M, over taxonomic cate-
gories,   Figure  42); and  suspension-feeders  and  herbivores  (S and  H  over
trophic  categories,  Figure  42).   Significantly  higher densities at MLLW as
compared to +2'  above  MLLW are shown for crustaceans,  mollusks, and herbi-
vores.   Significantly higher densities at the +2'  above MLLW tide level  are
shown  for polychaetes  and  suspension-feeders.   Tide  level   effects  for
carnivores and detritivores were not deemed  significant.   It is of interest
that the mean difference  for density of detritivores between tide levels is
practically nil,  and,  thus, does not provide for the  high nonsignificant
mean differences related  to duration seen earlier (Figure 41).

     Main effects  mean  numbers   of  individuals  per  brick  due to  grazer
stocking and  removal  are shown  in Figure  43.   Significant  effects  were
demonstrated  for  differences  in  density  of  polychaetes, mollusks,  and
herbivores.   The  manipulated species  (Acmaea  spp.) belong to  mollusk and
herbivore groups.   Thus,  the significant  difference indicating higher  mean
density   on  grazer-stocked versus grazer-removed  (S versus R,  Figure  43),
are  in part a reflection that  the manipulation itself  was detected by the
sampling approach.   In  the case  of polychaetes,  a significantly  higher


                                   107

-------
o
CXI
      H
      tf
en
w
H
o
H
frl
CO
rt
W
PQ
12
11
10
 9
 8
 7
 6
 5
 4
 3
 2
 1
                                           (NO)
                                                                                     (NO)
                                                                (NO)
                          +2'
                                        +2
                                                          + 2
      +2
                      TOTAL
                                 POLYCHAETES
                                                   CRUSTACEANS
MOLLUSKS
    Figure 40.   Mean number of species per brick of experiments conducted at Sequim Bay,  May  and  June,
                1980, related to  tide level.  Data summarized over two oil  treatments  (oiled  and
                unoiled); two grazer treatments (limpets stocked and limpets removed);  and  0-,  5-,
                and 30-day field  exposure bricks (O1 = MLLW; +2' = 2'  above MLLW).   (N  =  120  bricks.)
                Word in parentheses indicates statistical significance.

-------
o
10
     W
     W
50


45


40


35


30


25 +


20


15


10


 5
                                           (YES)
                             (YES
                                                            (NO,
                                                                                                        (YES)
                                                          0   5 30

                                                         Carnivores
                                                                   0   5 30

                                                                 Herbivores
 0   5  30

Detritivores
 Figure 41.
  0 5  30        0   5  30       0  5  30

Polyohastes    Crustaceans     Mollusks

               TAXONOMIC                                          TROPHIC
Mean number of individuals per  brick related to duration of experiments at Sequim Bay, May
and June, 1980.  Field exposure main effects summarized over two treatments (oiled and
unoiled); two tide  levels  (MLLW and +2' above MLLW); and two grazer treatments (limpets
stocked and limpets removed).   N = 120 bricks.)  0 = immediately post treatment; 5=5 days
post-treatment; 30  = 30 days  post-treatment.)  Word in parentheses indicates statistical
significance.
  0   5  30

Suspension

-------
u
    55  -•


    50  -


    45  "


    40  --
  tf
  W
  ft 35  -f-
  CQ
  a
  B 30  t
  H
g  25  t
  | 20  t

  K
  iz 15  •-


    10  -


     5  -
                             (YES)
                     (IES)
           (.YES)
                                              (YES)
                                      (NO)
                                                       (NO)
                                                                (YES)
          D'4-21
                              0'+2'
                              M
0' +21  0'  +2
0'+2'   0'  +2'
                                                 H
                TAXONOMIC
                                                   TROPHIC
Figure 42.
          Mean number of individuals  per brick  of  experiments at Sequim
          Bay, May and June,  1980,  related  to tide level.   Data summarized
          over two oil treatments  (oiled and unoiled);  two  grazer treat-
          ments (limpets stocked and  limpets removed);  and  0-, 5-, and
          30-day field exposure  periods  (0' = MLLW;  +2' = 2' above MLLW;
          P = Polychaetes;  C  - Crustaceans; M = Mollusks in the Taxonomic
          Groups and C = Carnivores;  H = Herbivores;  D  = Detritivores;
          S = Suspension Feeders in the  Trophic Group).  Word in paren-
          theses indicates  statistical significance.
                                  110

-------
                        c:

                        fD
                       co
O

Q.
      d- -S
      3- fD
 3d7

•o -••
 CU d-
 -S -"•
 fD <
 3 O
 d- -S
 3- fD
ro
-"•  II
3
Q. CO
Cu "O
d- fD
fD 3
      0<  fD

      O
      3
      O
            -h d-


            —' fD
                 to CO
      O

      CD II
           -X <
            •a n>
            o —<
            to to

           3 -S *~N
            fD 2
            CU
           1 ^  2 3

            N'<< i
            n>    cr
            O. CU fD
               3 -s
            o  a.
            <    o
            fD  C_i -h
            -s
         <<
      CU O
      3 3-
      Q. CU
            -i. cu
            O  3
            Q. Q.
               PO
                  S  »   Q.

                     I—" <
                  O  VD -••
                  -J. CO o.
                  —' O E
         fD
       II 10
to O
d- 3
      Cu
      -S
ct-~n<
to fD
d- CL
-i. fD
             II Cu fD fD
               CT ~
            —• O

            3'S
            -a
            fD 2:
            d- Cu  fD
            3  d-  -S
            fD  fD
            3  Q.  O"
            d-    -S
Oi
to
   3

ta d-
 3 3"
 _i. fD
 -h
 -•• -H
 O -S
 Cu o
 3 -a
 O 3-
 fl> -••
 •  O
      fD to
      to d-
         O
         Cu
         3
               S
            l/> —
            d-»-
            O    -^ Cu
            O Cu fD Nl
                    OO
            O IQ
            -i. -s  O
   O
fD
-S 2
cr
-"•  II
<
O 2
-s o
fD —i
(/) —i

   to
   7T
   to
            fD Q.
            a.
            ». o
                I
               en
               1
            3 cu
            •a 3
            fD Q.
            d-
            to co
               o
                I
               Q.
               CU
               <<
   -5  X
CU   73
3 d- fD
a. -s  -s
   n>  -"•
c cu  3
3 d- fD
033

—' 3  in
fD d-
O. •   Cu
                     o co
                  d- Cu  fD
                  S d-JD
                  O Cu  C
                                                                 MEAN  NUMBER  INDIVIDUALS  PER BRICK
           "^
           O
           £g
                       _..   o
                             H
                             c^j
                                      CO
                                     co
                             H
                                 o
                                     CO
                                     en
                                                en

                                               -4—
                                                                  en

                                                                 -H-
                                                                          ro
                                                                          o
                                                                                   ro
                                                                                   en
                                                                                            co
                                                                                            o
                                                                                              co
                                                                                              en
                                                                                                                      en
                                                                                                             en
                                                                                                             o
en
en
                                                                                                                    N
                                                                                                                    tg
                                                                                                                    03

-------
density  is  shown for  grazer-removed  bricks (R, Figure 43).   This  may re-
flect competitive interaction between polychaetes and the limpets; however,
the methods used  do  not allow such discrimination.   Higher mean densities
(not statistically  significant differences)  on grazer removal  bricks are
also shown for detritivores and crustaceans.

     Significant main  effects  on  density due to oil  treatment were demon-
strated within each of the taxonomic and trophic groups tested (Figure 44).
The  magnitude  of the mean  difference  in density  for detritivores  far
exceeded  that for  other  trophic  groups and  taxonomic groups.  The  mean
difference  between  control  and  oil-treated  bricks  for  polychaetes  and
crustaceans,  the  major contributors  to the detritivore  group (Table  19),
was also quite large.

     To  examine  the  possible effects  from  the  grazer  stocking and  removal
treatment in  more  detail,  the herbivore species were subdivided into those
which  feed  on  microalgae  and  those  which  feed  on macrpalgae  as  per
Simenstad  et  al.  (1979).    The  composition  of the  algae  itself  was  not
measured;  however,  the sea  lettuce,  Ulva  sp.,  a  macroalgal  type,  clearly
dominated  the plant biomass  and  appeared  to  completely  cover every brick
used  in  the  experiments.    The  composition  of macroalgae  and microalgae
herbivores  is   shown   in  Table  20.   The   microalgae   herbivores  were
principally moll usks and  two crustaceans.   They included the species which
were manipulated, Acmaea  spp.,  as well as several  herbivorous snails.   The
macroalgae herbivores were polychaetes and crustaceans.

     Main effects means  from analysis of variance for dry  weight of total
algal  biomass,  microalgae  herbivore  density, and   macroalgae  herbivore
density  are  shown   in Figure  45.   There  were statistically significant
effects  on  mean  dry  weight  total  algae  due to   duration   (day)  post-
treatment,  and  due  to  the oil treatment.   There  was  a  higher dry weight
algal biomass  in  the experiment 30 days post-treatment as  compared to the
immediate  post-treatment  experiment.   There  was  a  decreased  dry weight
algal biomass in oil-treated  bricks  as compared  to  controls.   The grazer
and  tide  level  treatments  did  not result  in significant  effects  on dry
weight algal biomass.

     There  were  statistically  significant  effects   from  duration  (day)
post-treatment,  grazer stock-removal  treatment, and tide  level  on micro-
algae  herbivores.   The  tide  level  effect appears  most  important.   The
difference  between   oil-treated  and  control  bricks  was  not statistically
significant for microalgae herbivore density.   There  was a higher density
of  microalgae herbivores  in  the  30-day post-treatment experiment  as  com-
pared to the  immediate post-treatment experiment.   There was a higher mean
density  of  microalgae  herbivores  on bricks receiving a stock of limpets^as
compared to those where limpets were removed.   There was a higher density
of microalgae herbivores at MLLW as compared to +2' above MLLW.

     There  were significant differences  in density  for  macroalgae herbi-
vores due to day post-treatment,  and due  to  the  effect  of oil treatment.
For this  group there was a lower density 30 days post-treatment as  compared


                                    112

-------
                   CQ
                   n>
                                                    MEAN NUMBER  INDIVIDUALS  PER  BRICK
CO
                           O
                           O
                           t/o
                              O
                              O
                              O
o


(-3





O


(-3





O


t-3





O


i-3
        01

       H-
                                                        ro
                                                        o
                                ro
                                en
co
O
CO
en
                                                                                       en
en
O
en
en
                                                                                                               en

-------
Table 20.   Composition of herbivores which feed on microalgae and
           macroalgae in Sequim Bay grazer manipulation experiments.1
MICROALGAE FEEDERS                           MACROALGAE FEEDERS
CRUSTACEANS                                  POLYCHAETES
     Caprella laeviuscula                         Nereis vexillosa
     Caprella sp.                                  Platynereis bicanaliculata
MOLLUSKS                                     CRUSTACEANS
     Acmaea digitalis                             Ampithoe simulans
     Acmaea persona                               Ampithoe sp.
     Acmaea sp.                                    Pugettia gracilis
     Alvania sp.
     Cyanoplax hartwegii
     Lacuna sp.
     Littorina sp.
     Margarities sp.
1 Trophic designation from Simenstad et al. (1979).
                                   114

-------
to immediate post-treatment.  This  is in contrast to  the  dry weight algal
biomass  and  density  of microalgae  herbivores.   There  was  a  lower  mean
density on oil-treated  bricks  as  compared to control  bricks.   Two further
points of elaboration about the data in  Figure  45  are that the scales for
density  of  the  two herbivore  groups shown in  the figure  differ.   Thus,
there was a much higher overall density for microalgae herbivores than for
macroalgae herbivores.  Second, the  mean density for microalgae herbivores
exceeds  by  at least a  factor of  four,  the numbers of  limpets stocked per
brick on half of the bricks.

Total Oil on Treated Bricks

     Mean concentrations of total  oil on bricks for  the  three experiments
are  shown in  Table  21.   It was apparent  at  the  time treatment was applied
that  a high proportion of the oil  present adhered  to vegetation  on the
colonized bricks.   This is reflected by  comparing  the mean concentrations
for  bricks  extracted immediately post-treatment with  vegetation intact to
bricks which were scraped to remove vegetation.

     For the 5-day post-treatment experiment, bricks were scraped to remove
algae  before extraction of oil  and chemical  analysis.  Mean concentrations
were  higher  at  both tide levels  at the  conclusion of  the  5-day post-
treatment experiment as compared  to the  immediate  post-treatment experi-
ment.   Since  no oil was added  to  bricks in the field,  these mean concen-
trations  obviously  reflect  uncontrolled variability  in  the  extraction
methods.

     The  mean  total oil concentrations  for the  conclusion  of  the 30-day
post-treatment  experiment  were based on samples from MLLW  only,  and in-
volved  extraction of bricks with  algae intact.   A substantial reduction in
mean  concentration  from 38.2  grams  per  brick  to 0.79 grams per brick is
apparent.  Caution  must be used in interpreting this as a large difference
since  the standard  deviations  are large and, as mentioned above, appear to
relate to the amount of vegetation present on bricks.
                                   115

-------
*  I
EC  ^
C5  Co
H  S
X
P5
Q
CO
o
&
w
EG
8
o

       4 -


       2 -


       2


       1


       0
      50


   *  40
      30


      20


      10
        0
co £  2.0
H CQ
  i—i     •. -

  I  1
  o  §  1-0


     §  0.5


         0




Figure 45.
              (IES)
                                                    (YES)
                             (NO)
                                        (NO)
                                        (YES)
                (YES)
                            (YES)
                                                    (NO)
                                                                       in
                                                                       0)

                                                                       
-------
Table 21.   Total oil concentrations (grams/brick) in grazer experiments
           at Sequim Bay (May - June 1980).


EXPERIMENTAL CONDITIONS                           TOTAL OIL/BRICK (Grams)1
                                                        Mean   S.D.
IMMEDIATELY POST-TREATMENT
  BRICKS SCRAPED TO REMOVE ALGAE

     Top Surface                                        0.07  (0.09)
     Whole Brick                                        0.25  (0.29)

  BRICKS EXTRACTED WITH ALGAE INTACT

     Top Surface                                       28.45 (23.70)
     Whole Brick                                       38.20 (34.32)

FIVE DAYS POST-TREATMENT
  BRICKS SCRAPED TO REMOVE ALGAE

     MLLW
     Top Surface                                        0.08  (0.03)
     Whole Brick                                        1.03  (0.34)

    +2' ABOVE MLLW

     Top Surface                                        0.17  (0.04)
     Whole Brick                                        1.32  (0.46)

THIRTY DAYS POST-TREATMENT
  BRICKS EXTRACTED WITH ALGAE

    MLLW

     Top Surface                                        0.01  (0.00)
     Whole Brick                                        0.79  (0.42)


1 N = 5 bricks for each mean reported.   Standard deviation  in parentheses.
  Total oil measured by IR Spectroscopy.
                                   117

-------
                                 SECTION 6

                                DISCUSSION

SIGNIFICANT EFFECTS FROM OIL ON INITIATION OF RECOVERY

     These studies  have demonstrated statistically  significant  effects on
density and species richness of taxonomic and trophic groups and individual
species  density  due  to oil treatment  under controlled  experimental  con-
ditions.   An  abbreviated summary of  these  effects  is  shown  in  Table 22.
From Table 22, 70% of the 56 parameters  estimated  have been significantly
reduced by the oil  treatments.   In only one case, Mysella tumida, was there
an  inconsistency  in  effects  shown.    For  that  species, a  significantly
higher density was  shown in oil-treated substrates  compared to controls in
the  sand habitat  (Vanderhorst et  al.,  1980),  and a  significantly lower
density was shown for the species in oil-treated as  compared to controls in
the  Discovery Bay  clam bed  habitat.   The  evidence indicates  that where
significant  effects  on density  were   not  shown for  two  of the  primary
species  (Exogone  lourei',  on the  clam  bed;  and Protothaca  standnea,  in
sand), it was  a  function of methodological  sensitivity  and  not  due to the
absence  of  an effect.   In  the  case  of  Exogone lourei,  the  +2'  above MLLW
tide  level   was  an   inappropriate  habitat  for  the  species.   The  highly
significant  and  extreme effect  of  tide  level  masked the  substantial  but
smaller  effect due to oil  treatment.    In the  case  of Protothaca starninea,
the density  of the species in the sand habitat  (Vanderhorst et al., 1980)
was far too low for valid comparison.

     In  addition to the 56 categories  listed on Table 22, effects from oil
treatment were  indicated  from descriptive  analysis of  variance  for the
density  of nearly  a third of all other species in the clam bed habitat and
for  dry  weight algal  biomass  in the  colonized epifauna  experiments.   We
believe  the  evidence  is  overwhelming  that oil  treatment,  as  applied in
these  studies,  is  a  sharp detriment to  the initial stage  of recovery in
each of the habitats investigated here.

     Further  perspective  concerning  the magnitude  of  the oil  treatment
effects  can  be  gained  by considering those  effects  in  light of other
environmental variables investigated.

     Season  was  by far  the most important  environmental  variable  for the
initiation of  recovery  based  on  data  in these  experiments.   For the sand
habitat,  dependencies  in data  did  not  allow us to  statistically test for
"significant" seasonal  differences.  However,  for experiments conducted in
the fall  versus  ones  conducted in the spring, there was an order of magni-
tude  difference  in density of primary  species  and  substantial differences
in both  composition and species richness.  The commercial clam bed habitat
was  investigated  in  a single  season.   The  densities   and  richness  of
taxonomic and  trophic  groups  were  significantly affected by  month  in the


                                   118

-------
Table 22.  Summary of tests for statistically significant  biological effects
           from Prudhoe Bay crude oil treatment.
CONDITION
1. DENSITIES
POLYCHAETES
*Armandia brevis
xExogone lourei
Platy nereis
bicanalicuiata
X0phiodromus pugettensis
xCapite!la capitata
xPolydora social is
CRUSTACEANS
*Leptochelia dubia
xPhotis brevipes
xCorophium ascherusicum
Exosphaeroma sp.
MOLLUSKS
*Mysella tumida
Protothaca staminea
Lacuna sp.
Transenella tanti 1 la

SAND2
X
X
X
0
0
X
X
0
X
X
0
0
0

CLAM
BED
X
X
0
0
0
X
X
0
0
X
X
X
0
HABITAT
BRICK COLONIZED EPIFAUNA
X X
0 X
0
X X
OTHER SPECIES (TAXONOMIC)

DETRITIVORES
CARNIVORES
HERBIVORES
SUSPENSION-FEEDERS
OTHER SPECIES (TROPHIC)

2.  SPECIES RICHNESS

POLYCHAETES
CRUSTACEANS
MOLLUSKS
TOTAL SPECIES
        0
        X
        0
        X
X
0
X
X
X
X
X
X
                 X
                 X
                 X
                 X
X
X
X
X
1 X = Statistically significant effect (P = 0.05).  Where * appears, species
      was a priori selected (P = 0.01).
  0 = Statistically significant effect not demonstrated by methods.
  - = Parameter not estimated.
2 Sand Data from Vanderhorst  et al. (1970).
                                   119

-------
seasonally independent experiments conducted in the hard substrate habitat.
The  range  of  mean densities  over  season was  often twice  (and  sometimes
three or more  times)  the range due to oil treatment and tide level.   Thus,
it can be expected that the magnitude of effect on initial  recovery will be
strongly  affected  by season  regardless  of  what  perturbation  initially
removes species from substrates.

     Substrate type has  a  special  role in recovery since it can ultimately
determine  whether a  species can  survive at all,  depending on  needs  for
attachment or  burrowing.   Our  investigation of substrate as an independent
experimental   variable was  inadequate  in  the  present  studies.   The  only
statistically  valid comparisons  made  related to effects on primary species
densities in sand habitat (Vanderhorst et al., 1980).  In those comparisons
the  texture   of   the  sand  (fine  versus  coarse)  resulted  in  significant
effects on density for two primary species, both small bivalves, during the
initiation of  recovery  in  the  fall (first 3 months).  A significant effect
due  to  substrate was  not  demonstrated for  these species  after  15  months
recovery.   The  effect  of  oil   treatment  was  much  more  severe   in  those
experiments than was substrate.  Two other sources of data in these studies
bear  on the  importance of  substrate in  recovery.   The first relates to
compositional  comparisons  between  the substrates.   We have  not formally
made  these comparisons  but it  is quite apparent  that mollusks were a more
important  contributor to  density and composition on the rock substrate and
in the  commercial  clam  bed substrate than they were to sand (Vanderhorst
et al.,  1980), both in terms of absolute  numbers and as a percentage con-
tribution.   The   other  source  of  data on the  importance  of  substrate in
recovery relates  to the  retention of contaminants and  is  addressed in the
next  section.

      The  importance  of  tidal  height in  the initiation  of  recovery  for
particular  groups  is  well  demonstrated  in  these  studies.   For infauna,
lower tide  heights often  resulted  in  significantly higher  densities  and
species  richness.   For epifauna, generally   higher  densities and richness
of crustaceans and polychaetes were at lower tide levels,  and the reverse
was   true  for mollusks.   The  effect  of  the  oil  treatments on initial
recovery   of   polychaetes  and  crustaceans  in  the   epifauna was  usually
slightly  less  than the  effect  of tide  height difference of two feet.  For
mollusks,  in  the  epifauna,  however,  oil  treatment  effects  resulted in
higher   differences   in   density   and  richness  than  did  tide  height
differences.

      We  described the geographical  location and physical attributes of the
four  study sites in the methods section.   Statistical  comparisons of  site
effect  were  made between  Sequim  Bay  and  Protection  Island  for infauna
recovery and  between  Sequim Bay and Rocky  Point for  epifauna  recovery.  For
both  habitat  types  statistically  significant effects  on  density  and/or
species  richness  were  demonstrated  to be  due  to  site.    For particular
species   (Platynereis  bicanaliculata;  Armandia  brevis),   in  the infauna
experiments  reported  in Vanderhorst et al. (1980), the effect of oil treat-
ment  was as  important as the effect of site  .  For mollusks  in the epifauna
the  effect  of oil treatment  was  more important  than  the  effect of  site
                                    120

-------
difference  between  Rocky  Point  and  Sequim  Bay.    Because  of  untenable
assumptions, statistically valid comparisons of site effects between Sequim
Bay and Discovery  Bay  were  not  possible in these studies.  However, mean
data  on  species   richness  (Figure  27)  and  taxonomic  and trophic  group
densities   (Figure  28)  showed  remarkable  similarities  between  the  two
sites.


OIL TREATMENT AND RETENTION OF OIL IN EXPERIMENTS

     Total oil concentration  at  the initiation of experiments with infauna
ranged from  a  low  of about 1,000 ppm  for the Sequim Bay summer experiment
(Vanderhorst et  al.,  1980) to a high  of  about 2,500 ppm total oil  in  the
commercial  clam  bed  experiment.   Fall  and  long-term  experiments  at Sequim
Bay (Vanderhorst et  al.,  1980) were intermediate  in  initial  concentration
with  about  1,800 ppm  total  oil.    Although  the mixing  of  oil  in sediment
does  not  occur in every  oil   spill  incident, the  total  amount  of  oil  in
experimental sediments, even in the highest concentration case (2,500 ppm),
was no  more than  25%  of  the  approximate 10,000 ppm  measured  in  sediments
following the  AMMOCO CADIZ spill.   There was one  anomaly  in  the analyzed
saturate  and aromatic  compound data  noted  for  the summer experiment  at
Sequim Bay in that the initial concentration of analyzed aromatic  compounds
was quite low  as  compared to other  experiments.   The experimental  designs
were  such  that we cannot  attribute differences  in biological effects  to
differences  in the initial  treatment concentration of oil.   In each set of
experiments,  the  oil  treatment per se,  as  documented  and verified by  a
large number of  measurements  of  tolaToil and saturate  and  aromatic com-
pounds,  was evaluated as a two-level factor (treated or untreated).

     The  distribution  of  oil in  time and  space within the  experimental
designs could  be expected to  relate to the pattern  of  faunal  recovery  in
time and space.  The  most striking difference in oil retention by  sediments
was the  difference between Sequim  Bay sediments  and those from the com-
mercial  clam bed on  Discovery Bay  in  3-month  summer  experiments.  For the
former  case (Vanderhorst  et  al.,  1980),  initial  oil  concentration  was
reduced by  48% on the  average;  while  for the commercial clam  bed experi-
ments losses  amounted to only 12.5% on the average.   This sort  of a dif-
ference is  of great interest  because,  based on the  evidence we  have pre-
sented,  it will directly affect the initiation of recovery.

     A part  of this differential  in  retention  can be  explained by tidal
height differences.  In the  summer  experiment at  Sequim  Bay,  the -2' tide
level  had 47% of  initial concentration;   the MLLW tide  level  had  57%  of
initial  concentration.   At Discovery  Bay,  the MLLW tide level had 80%  of
initial  concentration  and the +2'  tide  level  had 95%  of  initial  concen-
tration.

     At  least  three  other factors  may  contribute to  the difference  in
retention seen in  these  studies:   (1) wave  exposure;  (2)  initial  concen-
tration; and (3) sediment  texture  and  organic content.   Wave  exposure  was
                                   121

-------
not measured  in these  studies,  but we  judge that  for  the period  of the
experiment,  the two  sites  were roughly  equivalent and  received moderate
wave activity  for inland waters.   In terms  of  initial  concentration, the
concentration at Discovery  Bay  was nearly 2.5 times as high as the concen-
tration  at  Sequim Bay.   We have  MLLW  data  showing a consistent trend of
percentage  retention  to initial concentration  in  these  studies; however,
the data  are  confounded by  season, and statistical  comparison is inappro-
priate.  At Discovery Bay there was an initial concentration of about 2,500
ppm and  a retention  of 80% in  three months;  at Sequim Bay, in a fall 1978
experiment  of three months  duration,  there was an initial concentration of
1,758 ppm and a retention of 66%;   in the  Sequim Bay summer experiment, as
stated  above,   there  was   an  initial   concentration  of  1,069  ppm  and  a
retention of 57%.  The  thesis behind this observed trend could, of course,
be that the more oil  present, the greater the proportion  retained.

     It   appears   that  tide  level  and   initial   concentration  factors
adequately  account for  the differences in  proportional  retention  of oil
seen  in  these  studies.   However,  very   clear differences   in  sediment
retention time  for oil  have been  shown by  other investigators relating to
sediment  particle  size and  organic content, with  smaller  grain  size re-
taining  oil much  longer than  coarser  ones.  Thus,  a smaller  grain size
dominance at  Discovery Bay  would  be consistent with  the longer retention
times  seen  in these  studies.   In  fact, from  the limited data presented on
grain  size  patterns  from experimental  sediments in the  present studies   ,
the  grain  size  patterns at  the  two  sites  differ  chiefly in  the  higher
contribution of a gravel (greater than 5.66 mm diameter)  phase at Discovery
Bay.

     From   the  Sequim  Bay/Protection  Island  data  on   the   sand  habitat
(Vanderhqrst et al.,  1980),  we were able  to develop a  model  of oil loss
from sediments  to  predict a background level of total oil  in 18.5 months.
The model was developed from experimental  data which showed a high initial
rate of  loss  followed by a much lower  rate of loss thereafter.  The lower
rate was  fairly stable  between three and 15  months after initial contami-
nation.   If that lower rate (71 ppm per month) is applied to concentrations
of  total  oil remaining at  Discovery Bay  at the end of  three  months, then
complete  depuration of those sediments would occur by the end of 31 months.
From the  available data it is  impossible to  judge  if this is a liberal or
conservative estimate.  Further experimentation is warranted.

     Data  from  capillary gas chromatography  confirm the  general  trends in
retention  time  established by total oil  concentrations above.   One very
distinct  difference  between  Sequim  Bay experiments and  the  Discovery Bay
experiment  relates to analyzed aromatic compounds.  In both summer and fall
experiments  at  Sequim  Bay,  the   loss  of  analyzed  aromatics  was  propor-
tionally  much more rapid than the loss of saturates.  Only 14% of analyzed
aromatics remained in  the Sequim Bay fall  three-month experiment.   In the
summer  experiment, analyzed  aromatics  were at  background  levels in three
months    (Vanderhorst  et al., 1980).   In sharp contrast, analyzed aromatic
concentrations  were unchanged from initial  concentrations in the Discovery
Bay experiment  after 3 months of field exposure.


                                    122

-------
     In addition  to the  differences  in retention time due  to  tide level,
sediment  types,  and  initial  concentration  noted above,  the  experimental
approach also differentiated the effect of sediment depth on oil retention.
This factor is important to recovery since larval or young organisms settle
in  surface  sediments,  and  penetrate  deeper into  sediments as  they  grow.
Our approach  involved  a  simple  separation of the  top  and bottom halves of
cores.   Core  length was equivalent to  tray  depth (15 cm).  At  the end of
three  months  there were  significantly  higher concentrations of total  oil
and analyzed  aromatic  compounds  in the bottom half of cores as  compared to
the top.   Based on  the  three-month time frame  of  the  present  experiment,
this significant  differential  in concentration  probably  had little effect
on  observed  recovery.   We base  this  thesis  on  the  fact  that  the  same
general magnitude  of  effects  were  seen in  Sequim Bay  experiments  where
total   oil concentration  in whole cores was  less than half that  seen in the
Discovery  Bay experiment.  The  longer term  implication to recovery  may,
however, be great.   Oil  deep  in sediment strata degrades more  slowly than
in the upper strata because of less oxygen and mechanical action (Westlake,
1980).

     The  oil  treatment characterizations used  for  hard  substrates do  not
relate  easily to  real-world spill  situations;   and,  thus,   there  is  heavy
reliance  on  the   method   of  application  in making judgments  about  the
severity  of treatment in  these studies.  We  do  have  data on  the parti-
tioning of  oil in  these  experiments which  aid in  making  such judgments.
The  chemical   characterization   of  substrates  provides  insight  into  the
effect of tide level, site, and duration on the retention of oil within the
experiment  itself.   The top surface,  whole brick partitioning  in the  ex-
traction procedures provides a basis for best case and worst case retention
of oil.

     Both  total  oil  and  analyzed  saturate  and aromatic  compound classes
showed  a  rapid loss  of  oil from  the  top surface of bricks.   Only 16% of
total   oil concentration  on the  top surface remained five days after treat-
ment (computed  from  Table 11).   A further reduction to  12% of the initial
concentration at the  30-day post-treatment  period was only  slight and  not
significantly  different  from  five-day concentrations in  spite  of the fact
that 50 replicate bricks were used in each part of the comparison.  Coupled
with the  fact  that the  amount  of biological  colonization in  this  same
five-day period was, over all  experiments, negligible as compared to 30-day
colonization, we infer  that effects attributable to the oil treatment were
due to  rather small  amounts of oil (less than  one gram per brick).  It is
also pertinent, from Figure 20,  that excessively high concentrations of oil
immediately post-treatment bear no relation to five-day top surface concen-
trations.   Thus, if  the  treatments applied initially to bricks  (except for
the grazer  experiments  involving  colonized  bricks) were  unrealistically
severe,  the  actual  amounts  of  oil   to  which  colonizing  organisms  were
exposed, were not.

     The whole  brick  extractions of oil indicated that  a reservoir of oil
may exist  in the  pores  of the  brick substrate  for  extended periods.   For
the total  oil concentrations,  this amounted to 49% of initial concentration
                                   123

-------
for five  days  and  34%  of initial  concentration  for 30 days.  We  have no
evidence  that  this  reservoir  influences  the top  or  biological  exposure
surface  concentration.   This  characteristic of  retention  is, of  course,
unique  to the  substrates used.   While  it  would probably  be a safe  as-
sumption that the concrete bricks used in these experiments  fall within the
range  of porosity  of natural  substrates of  the Strait of Juan  de  Fuca
region, we have no evidence concerning this.

     The tide  level  distribution  of oil retained on  bricks  over  one-month
periods  is  defined  by  total  oil  data which  indicate  a slightly  greater
retention of oil  at the  +2'  tide level  as compared  to  MLLW.   The analyzed
saturate and aromatic compound data showed the reverse of  this tide level
trend  at  30 days  post-treatment.   However,  the  replication used  for  ob-
taining  those  means  was  limited to  five bricks  per condition  (a single
month)  and  the  month chosen  for  these  detailed analyses  was  atypical.

     There were  not significant  differences  in total oil  concentration for
bricks  related to  the site of experimentation between Sequim Bay and Rocky
Point.

     Using  the total oil  data on Table 11 as a base for calculation,  and
the rate  of loss between  five and 30 days as a rate  which may be represen-
tative  of longer term losses,  we calculated a top surface  background con-
centration  would occur  in  3.73  months from the  termination  of treatment,
and  a  whole  brick  background concentration would  occur in  2.93  months.
Obviously,  the rate of loss for whole bricks is  at the end  of  30 days a
much  higher  rate (grams per brick) than  the  top  surface  rate.  This whole
brick  calculation  can be  taken as  a  "best"  case.   If we use  rate  of loss
seen  for the  top  surface of  bricks  to compute  loss of the  amount of oil
remaining in the whole  brick at the end of 30 days,  it is much slower than
would  be expected, and represents, perhaps, a "worst" case for retention of
oil.   The calculation predicts  a  total  loss  of oil  from  bricks  in 13.5
months.   Thus,  even  using  the worst  case,  the  retention  of oil  on hard
substrates  is  of substantially shorter duration than for the sand and mud
substrates  (18.5 and 31 months) discussed previously.

     Although  the  retention  and distributon  of total  oil   in  the grazer
manipulation  experiments  did  not  differ substantially from  the patterns
shown  for other  hard substrate experiments,  it should  be noted that since
the  preponderance  of  organisms  were  present on bricks  at  the  time the
treatments  were  applied,  the  severity of the  treatments relative^ to the
organisms measured was  tremendously greater (orders  of magnitude in terms
of oil  concentration).


IMPLICATIONS OF  THE  OIL TREATMENTS  FOR OVERALL RECOVERY

      The  implications to  overall recovery from the effects of  oil discussed
previously  differs depending on the habitats and specific task objectives.
                                   124

-------
     Perhaps the simplest case relates to the infauna community in the sand
habitats which  has  been  previously  discussed (Vanderhorst  et a!.,  1980).
In that case  the  15-month recovery experiments were of sufficient duration
for control substrates to gain full recovery in terms of individual species
and  aggregate density,  and  species  richness.   Thus,  comparisons derived
directly from the experiments allowed expression of effects on recovery in
terms of  densities  and  species  richness  as a percentage of  the  fully re-
covered controls.   In that case, species richness was essentially recovered
in  15  months while  individual  and  aggregate   densities   lagged control
densities by  50%  at  the end  of  15 months.   By using data on oil  retention
in  those  studies,  an oil-free  habitat could be  predicted to occur 18.5
months after the initial treatment of about 1,800 ppm total  oil.  By adding
the time to  predicted total oil loss  and  demonstrated time to recovery in
controls (15  months) we  arrived at  a total recovery  time  of 33.5 months.
In that particular set of experiments, initiated in late summer, and termi-
nated  in  November  the  following  year,  the only  extrapolation  involved,
i.e., a further 3.5  months for total  oil  loss,  was  during  a winter low
recruitment  period.   It  is  probably not  critical  if  the  extrapolation is
not entirely  accurate.   The rationale for predicting recovery assumes that
a total depuration  of oil from  sediments  is required  before full recovery
can  be  realized.   That  assumption has not been tested  in these studies.
However, partial  recovery  can  occur even  when  substantial  amounts  of oil
remain  in  sediments  as indicated by the  compositions  and  densities in the
oil-treated  sediments.   Thus,  our estimates are worst case for the experi-
mental conditions used.

     The  overall  similarity  in species  richness  and density  between the
Sequim  Bay  site  and the Discovery Bay commercial clam bed site controls at
the  end of three months, shown  in  Figures 17 and  28,  suggests to us that
the "fully recovered" condition  in 15-month controls for Sequim Bay experi-
ments will  also  fairly represent conditions for controls at Discovery Bay.
Thus,  using  the  same  worst case  computations  as  applied  above and the
longer  oil   retention  time  estimates  for  Discovery  Bay  (31  months),  we
predict a full recovery  from the 2,500 ppm  oil treatment in 46 months (31 +
15  = 46).    Obviously,  because  of the necessary extrapolation of rate of
loss  of oil  from  sediments  beyond  the experimental  time  frame, there is
considerably  less  assurance  in this  estimate  than  for  the Sequim Bay
habitat.

     The  principal   focus  in  the  commercial  clam  bed experiment was the
effect  of  oil on  recovery of  the  littleneck clam (Protothaca  staminea).
Fortunately,  1980  was  a  typical  settlement  season  for  the species  in
Discovery Bay.  A  statistically significant effect from oil on the density
of  this species  was demonstrated.   From the mean density data presented in
Table 15, the magnitude  of mean  difference  is much greater at the MLLW tide
level  than at  the  +2'  tide  level.   We  reviewed  the data  for  individual
cores  and  trays  to  assure  ourselves  that this  mean difference was not
attributable  to  high  or low density  in  one or  a few cores,  or a single
tray.   This  was  not  the  case.   For  this particular species,  the demon-
strated effect  is an  especially conservative estimate  since the recovery
period  covered  only  about  half of  the species'  normal  recruitment period


                                   125

-------
and the concentrations  of  oil  were still extremely high at the termination
of this experiment.

     We presented data indicating a vertical  stratification of retained oil
concentration after three months in the commercial clam bed habitat.  These
data suggest that the surface  sediments where young clams initially settle
may be free  of  oil  at an  earlier  time  than  the projected 31 months; while
sediments at greater depth, but within 15 cm of the surface, may retain oil
for appreciably  longer periods.   The  effects of this oil  on  larger clams
which penetrate deeper into the sediment has  not been investigated in these
studies.

     The  significant  effects  on organisms colonizing bricks  in  the month-
long experiments with hard substrates clearly  establish  an important role
that oil  plays  in  the initiation of recovery  processes.   The  duration of
these experiments represents such a small proportion of the time reportedly
required  for development  of  a "mature"  rocky intertidal  community that
meaningful  recovery  rate  predictions are  out  of  reach.   Two  prevailing
views of  oil pollution effects  on recovery in  the  rocky intertidal are at
odds.

     The  first view,  subscribed to in our region  by  Nyblade (1979), holds
that  the  rocky  intertidal  communities  consist  chiefly  of  long-lived
organisms  with  infrequent successful   recruitment.   From  this view  our
experimental design has a demonstrated relevance.  Clearly, oil,  as applied
in these  experiments,  interferes with the initiation of  the recovery pro-
cess.    The  end  result  of these  recovery  processes  may  require  decades.
Unfortunately,   despite decades  of  imaginative experimental  research  and
biological survey,  quite  apart  from oil pollution research, there is not a
clear  single definition  of a  fully  recovered  rocky  intertidal  community
(there are many).   Indeed, a  demonstration  of the  relative importance of
only a few of  the many dependencies  inherent in a fully recovered concept
is  only   beginning.   Among all  studies  of  the rocky  shore,  catastrophic
physical  processes,  principally  from wave  action,  play  a large  role in
restructuring  communities   by  completely  denuding  sections of  the shore.
The second view  of  the role of oil in the recovery of the rocky intertidal
assigns oil  as  just another catastrophic event in  many.   Thus,  the effect
of  oil  in recovery lasts  only  so  long  as the  oil  is  present.   In concert
with the  second view, our data  indicate  that while the preponderance of oil
may be washed from substrates isolated in a large matrix of clean substrate
and  oil-free sea  water within a  period of  five  days,  a  small  residual
amount  adheres  to  the  substrate.   Within  a  month-long  time  frame this
residual  amount  is  sufficient to  produce  significant effects  on recovery.
The "best"  and  "worst" cases  for  total  loss  of this oil are approximately
three and 13 months.

     The  grazer  manipulation  studies identify  the  potential  of  _oil treat-
ment to  reduce  species richness and density, and algal biomass in existing
rocky shore  communities.  The effect of  experimentally shifting the balance
in grazer density was  shown to  significantly  increase moll usk and herbivore
densities, groups to  which  the  manipulated species belonged, and to


                                   126

-------
significantly decrease the  density  of polychaetes.   Within the 30-day time
frame  of the  experiments,  the  oil  treatment  resulted  in a  significant
decrease  in  algal  biomass  and  the  experimental  stocking and  removal  of
algal grazers did  not.   The absence of an  effect  from grazer manipulation
undoubtedly  relates  in part  to  the fact that the  principal  algal  biomass
was  macroalgae  while the grazers  manipulated were  feeders on microalgae.

     During  the 30  days  post-treatment,  in the grazer experiments overall,
there  were  significant  increases  in density  for   crustaceans,  mollusks,
herbivores,  and suspension-feeders,  and  significant increases  in  species
richness  for polychaetes.   There was a significant  increase  in algal  dry
weight biomass.  The  studies  were not of long enough duration to evaluate
the  possible effects  from  upsetting the  structure  in  these  communities.
The  loss  of  oil  from the colonized  substrates proceeded at a rate compar-
able  to   the  rate  obtained for  monthly experiments  previously  discussed.
                                   127

-------
                               REFERENCES CITED
Nyblade, C.F.   1978.   The intertidal  and shallow subtidal benthos of the
     Strait of Juan de Fuca, spring 1976 to winter 1977.  NOAA Tech. Memo.
     ERL MESA-26.   Marine Ecosystems  Analysis Program, Boulder, Colo., 156 pp.

Nyblade, C.F.   1979.   The Strait of Juan de Fuca intertidal  and subtidal
     benthos.   DOC/EPA Interagency Energy/Environment R&D Program Report
     EPA-600/7-79-213, U.S. Environmental Protection Agency, Washington,
     D.C., 129 pp.

Santos, S.L. and J.L.  Simon.  1980.  Marine soft-bottom community establish-
     ment following annual defaunation:  Larval  or adult recruitment?
     Marine Ecology-Progress Series,  2:235-241.

Simard, R.G.,  I. Hasegawa, W. Bandaruk, and C. Headington.   1952.  Infrared
     spectrophotometric determination of oils and phenols in water.  Analyt.
     Chem.. 23:1384-1387.

Simenstad, C.A., B.S.  Miller, C.F. Nyblade, K. Thornburgh,  and L.J. Bledsoe.
     1979.  Food web relationships of Northern Puget Sound  and the Strait of
     Juan de Fuca.   DOC/EPA Interagency Energy/Environment  R&D Program
     Report.  EPA-600/7-79-259, U.S.  Environmental Protection Agency,
     Washington, D.C., 335 pp.

Smith, G.F. and H.  Webber.  1978.  A biological  sampling program of
     intertidal habitats of Northern Puget Sound.  Washington State Depart-
     ment of Ecology Baseline Study Program.  Olympia, Washington, 311 pp.

Vanderhorst, J.R.  and P. Wilkinson.  1977.  Evaluation of marine invertebrate
     species diversity as an oil toxicity indicator from laboratory studies.
     In;  The Proceedings of the 4th Annual Toxicity Workshop, November 8-11,
     1977, Vancouver, B.C.

Vanderhorst, J.R.,  J.W. Anderson, P.  Wilkinson,  and D.L. Woodruff.  1978.
     Estimation of effects from oil on intertidal populations:  experimental
     perturbations versus natural variation.  Proc. Conf. on Assessment of
     Ecol. Impacts of Oil Spills.  (AIBS)  Keystone, Colo.,  June 14-17, 1978,
     pp. 807-820.

Vanderhorst, J.R.,  J.W. Blaylock, and P. Wilkinson.  1979.   Research to
     investigate effects from Prudhoe Bay crude oil on intertidal infauna
     of the Strait of Juan de Fuca.  NOAA Tech.  Memo. ERL MESA-45.  Marine
     Ecosystems Analysis Program, Environmental  Research Laboratories,
     Boulder, Colo., 38 pp.

Vanderhorst, J.R.  and P. Wilkinson.  1979.  The littleneck clam, Protothaca
     staminea. as a tool for potential oil pollution assessment:  Part 1-
     Density of stock.  Mar. Environ. Res.. 2:223-237.
                                     128

-------
Vanderhorst, J.R., J.W. Blaylock, P. Wilkinson, M. Wilkinson, and G.
     Fellingham.  1980.  Recovery of Strait of Juan de Fuca intertidal
     habitat following experimental contamination with oil.  DOC/EPA Inter-
     agency Energy/Environment R&D Program Report EPA-600/7-80-140, U.S.
     Environmental Protection Agency, Washington, D.C., 73 pp.

Warner, J.S.  1976.  Determination of aliphatic and aromatic hydrocarbons
     in marine organisms.  Analyt. Chem., 48:578-583.

Webber, H.H.  1979.   The intertidal and shallow subtidal benthos of the
     west coast of Whidbey Island, spring 1977 to winter 1978.  NOAA Tech.
     Memo. ERL MESA-37.  Marine Ecosystems Analysis Program, Environmental
     Research Laboratories, Boulder, Colo., 108 pp.

Westlake, D.W.S. and F.D. Cook.  1980.  Petroleum biodegration potential of
     Northern Puget Sound and Strait of Juan de Fuca environments.  DOC/EPA
     Interagency Energy/Environment R&D Program Report EPA-600/7-80-133,
     U.S. Environmental Protection Agency, Washington, D.C., 133 pp.
                                      129

                                                                  * GPO 798 - 672 1981

-------
United States
Environmental Protection
Agency
Center for Environmental Research
Information
Cincinnati OH 45268
Official Business
Penalty for Private Use, $300
Postage and
Fees Paid
Environmental
Protection
Agency
EPA-335
                                                                                                    Special Fourth-Class Rate
                                                                                                              Book

-------