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
------- |