.THE EFFECTS UPON THE MACROFAUNAL COMMUNITY OF A DOMINANT
BURROWING DEPOSIT FEEDER, CALLIANASSA CALIFORNIENSIS,
AND THE ROLE OF PREDATION IN DETERMINING ITS
INTERTIDAL DISTRIBUTION~
by
MARTIN HAROLD POSEY
A DISSERTATION
Presented to the Department of Biology
and the Graduate School of the University of Oregon
in partial fulfillment of the requirements
for the degree of
Doctor of Philosophy
June 1985
,.
I hereby approve of this dissertation by
Martin Harold Posey
r rr
__LA,Cr-~~-
Peter Frank
ii
iii
An Abstract of the Dissertation of
Martin Harold Posey for the degree of Doctor of Philosophy
in the Department of Biology to be taken June 1985
Title: THE EFFECTS UPON THE MACROFAUNAL COMMUNITY OF A DOMINANT
BURROWING DEPOSIT FEEDER, CALLIANASSA CALIFORNIENSIS, AND
THE ROLE OF PREDATION IN DETERMINING ITS INTERTIDAL
DISTRIBUTION
Approved:~~Peter Frank
The burrowing, deposit feeding ghost shrimp, Callianassa
californiensis (Dana), forms dense beds in the mid-intertidal
zone of sandflats along the west coast of North America. In most
Oregon and Washington embayments, this shrimp displays a distinct
zonation with densities declining by an order of magnitude in the
low intertidal and subtidal. This investigation focused on the
interactions between dense shrimp beds and other infauna of the
tideflat, and on the role of predation in limiting the distribution
of these beds.
The species composition within a dense shrimp bed and of an
adjacent area with low shrimp density were observed from benthic
iv
cores taken in the Coos Bay, Oregon, estuary. Sedentary polychaetes
and amphipods had significantly lower abundances within than outside
a dense Callianassa bed. The numbers of most free-burrowing fauna
were uncorrelated or positively correlated with increased shrimp
density. Sediment resuspension and destabilization have been
suggested to explain the effect of deposit feeders on other
soft-sediment organisms. My results suggest that sediment
destabilization is the primary mode of interaction between
Callianassa and intertidal fauna.
Exclusion cages were used to study the effect of predation in
restricting dense beds to the mid-intertidal zone. The distribution
and diets of several potential predators were estimated from
laboratory experiments and beach seine collections. The lower edge
of the shrimp beds during summer was strongly affected by predation.
The staghorn sculpin, Leptocottus armatus, was identified as an
important predator on adult ghost shrimp during the summer months.
Decreased activity of the shrimps contributed to the maintenance of
zonation during winter when predation was less intense. These results
show that the distribution of a deep-burrowing animal can be
significantly affected by predation; previous research on other
species had put this into question. Additionally, limitation of a
dominant organism, such as Callianassa, by predation may affect the
overall species composition of the soft-sediment community.
VITA
NAME OF AUTHOR: Martin Harold Posey
PLACE OF BIRTH: Nanjemoy, Maryland
DATE OF BIRTH: December 28, 1959
GRADUATE AND UNDERGRADUATE SCHOOLS ATTENDED:
University of Oregon
University of North Carolina
Charles County Community College
DEGREES AWARDED:
Doctor of Philosophy, 1985, University of Oregon
Bachelor of Arts, 1980, University of North Carolina
Associate of Arts, 1978, Charles County Community College
AREAS OF SPECIAL INTEREST:
Population and Community Ecology
Invertebrate Zoology
PROFESSIONAL EXPERIENCE:
Teaching Assistant, Department of Biology, University of Oregon,
Eugene, 1980-85
v
vi
ACKNOWLEDGMENTS
It is difficult to know how to thank all of the people who have
helped with this research. It has benefited from numerous conversations
with committee members and fellow graduate students, especially Dr. Paul
Rudy, Marshall Pregnall, Robert Graham, and Jeffrey Goddard. Special
thanks go to my major advisor, Dr. Peter Frank, who was always willing
to discuss problems and whose comments on earlier drafts of this
manuscript were invaluable in making it understandable.
Part of this research would not have been possible without the help
of many students from the Oregon Institute of Marine Biology. Their
willingness to pull beach seines in near-freezing weather, spend long
hours observing shrimp behavior, and repeatedly slog through the mud is
greatly appreciated.
Finally, I wish to thank my father who patiently introduced me to
the wetlands, rivers, and forests of southern Maryland. His continual
support has been important throughout this study.
This research was supported in part by funds from Sigma Xi.
Chapter
vii
TABLE OF CONTENTS
Page
I.
II.
GENERAL INTRODUCTION•••••••••••••••••••••••••••••••••••••
FAUNAL ABUNDANCES CORRELATED WITH A DENSE
CALLIANASSA CALIFORNIENSIS BED•••••••••••••••••••••••••••
1
6
III.
Introduction........................................... 6
Materials and Methods.................................. 10
Results..................................... .. 15
Discussion••••••••••••••••••••••••••••••••••••••••••••• 31
EXAMINATION OF ZONATION AND BEHAVIORAL PATTERNS
IN CALLIANASSA CALIFORNIENSIS •••••••••••••••••••••••••••• 37
Introduction ••••••••••••••••••••••••••••••••••••••••••• 37
Methods................................................ 38
Results.......................................... . ..... 41
Discussion••••••••••••••••••••••••••••••••••••••••••••• 53
IV. PREDATORS OF CALLIANASSA CALIFORNIENSIS
IN COOS BAY, OREGON •••••••••••••••••••••••••••••••••••••• 57
Introduction••••••••••••••••••••••••••••••••••••••••••• 57
Methods................................................ 58Results................................................ 62Discussion 71
V. EFFECT OF PREDATION IN LIMITING THE LOWER
DISTRIBUTION OF CALLIANASSA CALIFORNIENSIS BEDS •••••••••• 74
Introduction........................................... 74
Methods................................................ 75
Results................................................ 84
Discussion.................•......•...•................ 91
VI. CONCLUSION ••••••••••••••••••••••••••••••••••••••••••••••• 99
APPENDIX: OCCURRENCE OF ANIMALS TAKEN IN BENTHIC CORES ••••••••••• 105
BI BLIOGRAPHY • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • •• 108
viii
LIST OF TABLES
Table Page
II-I. Number of Taxa Collected near a Dense
Callianassa californiensis Bed ••••••••••••••••••••••••••• 16
11-2. Mean Numbers Per .01 m2 (and Standard
Error of Mean) for the Ten Most Abundant
Taxa Collected from Spring 1982 to Winter 1984••••••••••• 18
11-3. F-statistics for the Effect of Shrimp Density,
Season and Shrimp*Season Interactions on the
Abundance of the Ten Most Common Taxa Sampled
at Valino Island ••••••••••••••••••••••••••••••••••••••••• 24
III-I. Statistics for Regression of Shrimp Density on
Hole Density•••••••••••••••••••• ~ •••••••••••••••••••••••• 42
111-2. Duration of Behaviors Observed in Artificial Burrows ••••••• 50
111-3. Tideflat and Surface Water Temperatures at
Valino Island •••••••••••••••••••••••••••••••••••••••••••• 55
IV-I. Number of Staghorn Sculpins Caught in Fish Traps During
a Single High-Tide Period on Each of Nine Dates •••••••••• 61
V-I. Results of Migration Experiments ••••••••••••••••••••••••••• 79
V-2. Mean and Standard Deviation (Sorting) of Particle
Size in Sediment Samples Taken in Exclosure (E)
and Unmanipulated (0) Treatments ••••••••••••••••••••••••• 89
V-3. Change in the Number of Holes Within Transplant Plots •••••• 91
V-4. Comparisons of Low-tide Salinities (%0) from
Three Locations at Valino Island, Coos Bay••••••••••••••• 94
ix
11-3. Seasonal Abundance of Tube Building Crustaceans ••••••••••• 2S
11-2. Seasonal Abundance of Spionid Polychaetes ••••••••••••••••• 21
Page
LIST OF FIGURES
II-I. Study Sites Within Coos Bay ••••••••••••••••••••••••••••••• 11
Figure
11-4. Seasonal Abundance of Mobile Organisms •••••••••••••••••••• 28
III-I. Relationship Between Hole Density and Callianassa
Abundance ••••••••••••••••••••••••••••••••••••••••••••••• 43
111-2. Monthly Fluctuations of Hole Densities in a
Callianassa Bed ••••••••••••••••••••••••••••••••••••••••• 45
111-3. Hole Densities from Transects Taken Across the
Lower Edge of Three Callianassa Beds •••••••••••••••••••• 46
111-4. Relationship Between Salinity and the Per Cent of Time
Spent Actively Moving Within Artificial Burrows ••••••••• SI
IV-I. Placement of Fish Traps Relative to a Dense Ghost
Shrimp Bed ••••••••.••••••••••••••••••••••••••••••••••••• 60
IV-2. Per Cent of Staghorn Sculpins Having Callianassa
As Part of Their Diet ••••••••••••••••••••••••••••••••••• 63
IV-3. Seasonal Abundance of Staghorn Sculpins in
Joe Ney Slough, Coos Bay •••••••••••••••••••••••••••••••• 67
IV-4. Seasonal Abundance of Staghorn Sculpins at Empire,
Coos Bay •••••••••••••••••••••••••••••••••••••••••••••••••• 69
V-I. Exclosure Cage and Controls Used in Migration Studies ••••• 76
V-2. Location of Migration and Transplant Experiments •••••••••• 80
V-3. Comparison of the Relationship Between Hole Density
and Shrimp Abundance Within Exclosure Cages and
Outside of Cages During the Summer of 1982 •••••••••••••• 87
V-4. Hole Densities Within the Low Intertidal
Settlement Area ••••••••••••••••••••••••••••••••••••••••• 92
1CHAPTER I
GENERAL INTRODUCTION
A fundamental objective of ecological research is to elucidate the
factors affecting the distribution and abundance of organisms (Krebs,
1978). A particular pattern of distribution that has received
considerable attention in marine environments is intertidal vertical
zonation. Many intertidal organisms have long been recognized to occur
in zones with distinct boundaries. Some of these species are largely
restricted to this habitat while others are predominately terrestrial or
subtidal. Since the vertical profile in the intertidal presents a
gradient of physical stresses and biological pressures, studies
determining the cause of zonation have provided important information on
the role of these factors in restricting the distribution of organisms.
Lower limits to the distribution of rocky intertidal organisms have
often been attributed to biological interactions (for reviews see
Dayton, 1984; Paine, 1984). Inferior competitors may be restricted to
higher tidal levels (Connell, 1961; Dayton, 1971) and predation or
herbivory may effectively exclude many organisms from the low intertidal
or subtidal (Paine, 1974; Lubchenco and Menge, 1978). Paine (1966,
1974) suggested that predation on a dominant competitor could change the
species composition of the community. He found that sea star predation
restricted dense beds of the competitively dominant mussel, Mytilus
californianus, to the mid intertidal. When sea stars were removed, the
2primary space was slowly invaded by a monoculture of Mytilus. Where
this mussel was absent, the low intertidal was dominated by diverse
utilizers of primary space, but lacked many of the animals associated
with the mussel matrix (Suchanek, 1979).
In soft sediments, by contrast, physical factors generally have
been considered more important in determining the lower distribution of
organisms. Several studies have noted a change in sediment
characteristics (usually particle size) associated with an animal's
lower tidal distribution and have suggested that these substrate
changes, or the correlated changes in current patterns, may be
responsible for zonation (Wieser, 1959; Green and Hobson, 1970; Johnson,
1970; Longbottom, 1970; Morgan, 1970; Vassallo, 1971; Featherstone and
Risk, 1977; Grant, 1981; Willason, 1981). Eckman (1983) suggested that
local current regimes and passive settlement of larvae are important in
establishing zones and Swinbanks (1982) proposed that some lower limits,
particularly of certain plant species, are related to zones of rapid
change in submergence time (critical tide levels).
There is little evidence that directly implicates competition as a
cause of zonation in soft sediments. Levin (1981) showed that the
polychaete Pseudopolydora could aggressively exclude a second
polychaete, Streblospio, under laboratory conditions. She observed that
the densest aggregations of Streblospio existed immediately above those
of Pseudopolydora in the field. Fenchel (1975, 1977) has described a
possible case of competition among intertidal snails. Exploitive
competition (Levinton, 1977) and interference competition (Woodin, 1974;
Peterson and Andre, 1980) have been noted subtidally, but do not appear
3common. Peterson (1979) suggested that competitive exclusion is not
important in soft sediments because of the lack of a mechanism for
overgrowth or crushing, as seen on hard substrates, and the resistance
of most invertebrates to exclusion through exploitive competition.
Distinct assemblages of soft-sediment organisms may result from
interactions between functional groups (Rhoads and Young, 1970; Woodin,
1974; Brenchley, 1981, 1982a). Organisms with similar modes of life,
such as mobile deposit feeders or tube builders, may alter the
environment so that other groups are excluded (see Woodin and Jackson,
1979, for a review). However, while functional group interactions may
be responsible for the disjunct distributions of certain species
assemblages, they have not been shown to act along a tidal gradient to
produce zonation.
Predator exclusion experiments have indicated that predation may
reduce the abundance of many epifaunal and shallow infaunal species
(Naqui, 1968; Reise, 1978; Woodin, 1978; Virnstein, 1977, 1978; Lee II,
1978; Holland et a1., 1980; Wiltse, 1980; Kneib and Stiven, 1982).
Higher faunal abundances in seagrass beds have also been attributed to
lower predation there than in areas lacking seagrasses (Young et a1.,
1976; Young and Young, 1977; Heck and Thoman, 1981). Secondary
predators may affect abundances of shallow infauna in sub-tropical
seagrass beds (Nelson, 1981). Although deeper burrowing fauna may be
subject to partial predation, such as siphon snipping in clams (Peterson
and Quammen, 1982), their numbers have generally not been observed to
change with reduced predator abundances.
Evidence directly implicating predation as a cause of intertidal
4zonation has also mainly involved epifaunal and shallow infaunal
species. Densities of a sand dollar, Dendraster, are reduced subtidally
by sea star predation (Birkeland and Chia, 1971). The snail Cerithidia
occurs most commonly in salt marsh pans beyond the tolerance of its
predator, Ilyanassa (Race, 1982), and Ilyanassa larvae appear to be
susceptible to Littorina predation (Brenchley, 1982b). Dense
aggregations of the amphipod Corophium have been reported from the upper
intertidal above the regions of heaviest predation (Reise, 1978) and
other amphipods have been observed to occupy marsh habitats when
predators are abundant (Van Dolah, 1978). However, predation has not
been reported to cause zonation in deeper burrowing infauna and there is
no evidence that predation on a competitive dominant may prevent
competitive exclusion, as observed in the rocky intertidal community
(Peterson, 1979).
Thus most evidence to date suggests that zonation of soft-sediment
organisms is largely in response to physical factors, with predation
determining the lower distribution of certain near-surface animals.
Competitive exclusion appears unimportant in producing zonation, and
keystone predators, such as sea star preying on mussels (Paine 1974)
have not been observed.
The ghost shrimp, Callianassa californiensis, forms dense beds in
the mid intertidal of many Oregon and Washington sandflats, but is
usually present in low numbers in the low intertidal and subtidally
(Thompson and Pritchard, 1969; McCrow, 1971; Swinbanks and Murray, 1981;
personal observation). These mid-intertidal beds occur over a wide
range of sediment types. Because of the wide geographical distribution
5of Callianassa (MacGinitie, 1934) and its presence in a variety of
substrate types, predation was suspected as a possible cause of zonation
for this shrimp in Oregon estuaries. The following studies have two
purposes: 1) to examine the distribution of other tideflat organisms
with respect to a dense Callianassa bed and thus to look for possible
competitive interactions, and 2) to estimate the role of predation in
limiting these beds primarily to the mid intertidal.
6CHAPTER II
FAUNAL ABUNDANCES CORRELATED WITH A DENSE
CALLIANASSA CALIFORNIENSIS BED
Introduction
Biological interactions as potential organizing forces in silt and
sand bottom communities have generated considerable interest during the
past several decades. Early work on the distribution of animals in soft
substrates concentrated on the effects of sediments (especially sediment
grain size) and currents in determining the species composition of these
habitats (Allen, 1899; Peterson, 1918; MacGinitie, 1935; Shelford et
al., 1935; Jones, 1950; Coe, 1953; Pratt, 1953; Weiser, 1956, 1959).
Sanders (1958; Sanders et al., 1962) has suggested that sediment grain
size may determine the dominant feeding mode in an area, and numerous
recent studies continue to support the contention that sediment
characteristics are correlated with large scale ()100 m) patterns in
species abundance and feeding type (Pfitzenmeyer and Drobeck, 1967;
Longbottom, 1970; Nichols, 1970; Bloom et al., 1972; Boesch, 1972, 1973;
Meadows and Campbell, 1972; Driscoll and Brandon, 1973; Featherstone and
Risk, 1977; Holland and Dean, 1977; Lawrence and Murdoch, 1977;
Levinton, 1977; Mountford et aI, 1977; Whitlatch, 1977, 1981; Oliver et
al., 1980; Chester et al., 1983). However, it has become increasingly
apparent that indirect interactions among soft-sediment flora and fauna,
often involving modifications of the immediate environment, may playa
7dominant role in determining the species composition of the benthic
community on a local scale.
Rhoads and Young (1970), with their trophic group amensalism
hypothesis, were among the first to examine the effects on the benthic
community of physical changes resulting from the activities of soft-
sediment organisms. In this theory, they attempted to explain the near
absence of suspension feeders (organisms filtering food from the water
column) in fine deposits, and their dominance in sandy environments by
suggesting that deposit feeders excluded suspension feeders from silty
substrates through resuspension of sediment while feeding. The
increased amount of resuspended material associated with disturbance
from deposit feeding would clog the filtering apparatus of suspension
feeding animals and inhibit their filtering capabilities. Moreover,
deposit feeders would be uncommon in sands because of a deficiency of
detrital material in these areas. Several parts of this trophic group
amensalism hypothesis have been supported. A dominance of deposit
feeders in finer sediments and suspension feeders in sands has been
reported from several areas (Bloom et al., 1972; Driscoll and Brandon,
1973; Myers, 1977; Whitlatch, 1977). Some burrowing deposit feeders
appear to increase the amount of resuspended sediment (Brenchley, 1978;
Grant et al., 1982; Nowell et al., 1981; Murphy, 1983) and an increase
in the amount of suspended sediment affects the filtration rate of at
least certain suspension feeders (Stone and Palmer, 1975; Gerrodette and
Fleshig, 1979; Murphy, 1983).
However, several studies have described discrete assemblages of
soft-sediment organisms that are not easily explained on the basis of a
8trophic group model alone. These involve functional groups such as
tube-building animals, burrowing deposit feeders, or suspension feeding
bivalves. Mobile deposit feeders have been found to affect sedentary
deposit feeders adversely (Ronan, 1975; Brenchley, 1981; Wilson, 1981;
Bird, 1982). Tube-dwelling deposit feeders and tube-dwelling suspension
feeders may occur together in high densities (Young and Rhoads, 1971),
and deposit feeders or mobile predators may be excluded by sedentary
organisms (Woodin, 1974; Brenchley, 1982a; Peterson, 1982).
Brenchley (1978, 1981, 1982a) has proposed that mobile and
sedentary organisms represent distinct functional groups which can
exclude each other regardless of feeding method. Mobile animals disturb
the sediment, thereby preventing the establishment of sedentary forms,
while sedentary organisms will tend to bind the sediment and will
therefore inhibit burrowing. Since most suspension feeders are
sedentary and most infaunal mobile organisms are deposit feeders,
Brenchley's mobility mode hypothesis may, in certain cases, lead to the
same predictions concerning species associations as Rhoads and Young's
trophic group amensalism hypothesis.
A third functional group hypothesis involving sediment modification
has been proposed by Woodin (1976). She has suggested that discrete
assemblages of burrowing deposit feeders, tube-dwelling polychaetes and
crustaceans, and suspension feeding bivalves may occur through adult-
larval interactions between these groups. Deposit feeders may prevent
the establishment of tube builders or suspension feeding bivalves by
continually disrupting the surface sediments. Alternatively, tube
builders may occupy most of the available surface space or defecate over
9the remaining area, thus preventing successful larval colonization.
Dense assemblages of suspension feeding bivalves were considered capable
of filtering most larvae from the water before they could settle.
Although the mechanism of interaction differs among these three
hypotheses, they all seek to explain the occurrence of discrete
assemblages of species in soft sediments, and to determine how
interactions between functionally different organisms affect the species
composition of soft-sediment benthos. The present study constitutes one
example where such functional group interactions are possibly a major
factor determining the distribution and abundance of tideflat fauna, and
it will identify the major functional groups involved and their mode of
interaction.
The ghost shrimp, Callianassa californiensis, is a burrowing
deposit feeder, often living in burrows up to 0.5 m in depth, on
sandflats in embayments of the North American west coast from Alaska to
Baja California (MacGinitie, 1934). Dense beds of these animals may
have over 500 individuals/m2 (Bird, 1982) and their activity results in
a continual reworking of near-surface sediments (MacGinitie, 1934; Bird,
1982). This bioturbation often gives the sediment in a dense bed a
soft, quicksand quality. The intensity of shrimp activity varies
seasonally, with higher activity during summer and a decrease in the
winter months (see Chapter 3). In Oregon and Washington estuaries,
ghost shrimp beds are generally restricted to the high and mid-
intertidal zones (Hedgepeth, 1952; Thompson and Pritchard, 1969;
Swinbanks and Murray, 1981), the lower limit sometimes being quite
abrupt. Density may decrease by an order of magnitude over a horizontal
10
distance of only 1-2 m and a vertical profile of only a few centimeters
(personal observation).
Given the proposed importance of burrowing deposit feeders in soft
sediments (Rhoads and Young, 1970, Woodin, 1976; Brenchley, 1978), one
might expect dense beds of Callianassa californiensis to have a strong
effect on the distribution of the benthic fauna. In the present study,
benthic cores from within and outside a ghost shrimp bed were examined
to determine if dense Callianassa beds are correlated with changes in
faunal species composition and abundance over short distances. If such
correlations exist, one would expect seasonal changes in faunal
distribution or abundance to be associated with changes in shrimp
activity. Further, one should be able to discriminate whether
distribution patterns are more easily explained by functional group
interactions involving trophic groups or by a mechanism involving
physical disruption of the sediment (mobility mode or adult-larval
interactions).
Materials and Methods
Study Area
Sampling was conducted intertidally on the south side of Valino
Island within the South Slough Estuarine Sanctuary, Coos Bay, Oregon
(Figure 11-1). This area experiences mild temperature fluctuations, air
temperatures during the study usually remaining between SoC and 30°C and
surface water temperatures ranging from 10°C in winter to 20°C in summer
(personal observation). Rainfall varies seasonally, with most
11
FIGURE II-I. Study sites within Coos Bay. A: Long Island Point; B:
Valino Island; C: Battle Flats; D: Joe Ney; E: Oregon
Institute of Marine Biology; F: Empire; G: Causeway.
Numbered areas represent sites of hole density transects
of examine zonation of dense shrimp beds.
12
N
5 Km
So.t~
Slo.g~
Oregon
13
falling during November through May. Low-tide surface water salinities
ranged from a low of 4 0/00 during winter freshets to 31 0/00 during
July and August. Interstitial salinities ranged from 16 0/00 to 31
0/00, suggesting that the occurrence of extreme low salinities during
winter freshets does not last for extended periods.
Benthic samples were taken around a dense shrimp bed adjacent to a
small Salicornia marsh. The substratum consisted primarily of fine
sands (.125-.250 mm) with slightly coarser sediments within the ghost
shrimp bed. The lower boundary of the bed marks an abrupt change in
shrimp numbers with densities declining from 100/m2 to 2-3/m2 over a
horizontal distance of less than 2 m. Although this lower edge of the
shrimp bed became somewhat less abrupt during winter, the position of
the bed remained relatively constant throughout the study.
Sampling
Benthic cores were collected in three areas: within the ghost
shrimp bed; along the transition zone from high to low Callianassa
density (intermediate density); and in an area of low shrimp abundance
approximately 3 m from the ghost shrimp bed. All three areas were
approximately 1 m above mean lower low water (MLLW). The distance
between areas of high and low density was 5-6 m. Samples within each of
the three areas were chosen haphazardly with the toss of a shell. Three
replicates were taken in each area at roughly bimonthly intervals
(March, June, July, September, October, January) from March 1982 to
January 1983 and five samples were taken from each area at quarterly
intervals from February 1983 to February 1984. In order to examine
14
seasonal variations in faunal abundance and the effect of disturbance by
shrimp, the 1982 data were combined into seasonal categories for
analysis (Spring: March-May; Summer: June-August; Fall: September-
November; Winter: December-February). These intervals roughly
correspond to the rain and upwelling cycles prevalent along the Oregon
coast (such as maximum rainfall between December and February and strong
upwelling in June and July).
Cores were 12.5 em in diameter and 15 em deep. They were
immediately preserved in 4% formaldehyde with rose bengal dye added,
seived after 4-5 days on a 0.5 mm screen, and transferred to 70% ethanol
for storage (Gonor and Kemp, 1978). Counting was done under the lOx
power of a dissecting microscope, recording only the number of whole
organisms or heads present. Identifications were made using Fauchald
(1977), Hartman (1968a, 1968b), and Smith and Carlton (1975). The
identifications were kindly checked by Lynn Rudy from the Oregon
Institute of Marine Biology and by Howard Jones from Oregon State
University.
Statistical Analysis
A two-way analysis of variance (ANOVA), blocking for differences
between years, was used to test the effect of ghost shrimp density,
season, and interactions between season and shrimp density on species
abundances. Blocking for years was considered necessary because of the
unusual sea conditions associated with the 1983 EI Nino/Southern
Oscillation event. Since a significant interactio~ effect may be
expected if shrimp activity varies seasonally, it should be noted that
15
in a model I ANOVA, such as used here, the presence of significant
interactions does not invalidate an examination of main effects (Sokal
and Rohlf, 1981).
F-max tests (Sokal and Rohlf, 1981) for the ten most abundant
species indicated that variances were heterogeneous between sampling
dates and a logI0(x+l) transformation was applied to achieve
homogeneity. The absence of some species from a sampling location on a
particular date (resulting in a zero variance for that cell) was not
considered to be a major problem since recalculation of the ANOVA
substituting small artificial variances in these cases did not yield
noticeably different results. All ANOVA's were calculated with the
General Linear Models Procedure of the Statistical Analysis System (Ray,
1982).
Results
Forty-nine taxa were identified from the South Slough study site,
37 to the generic or species level (see Appendix). A single species,
the tanaid Leptochelia dubia, constituted 44-75% of the animals during
each sampling period. Over 95% of the individuals collected were
represented by only seven taxa.
To examine the relationship between the number of species and ghost
shrimp density, the number of species in the three sampling areas (Table
II-I) was compared using a G-test, blocking for seasonal variations
(Sokal and Rohlf, 1981). Differences were not significant (G=3.58; 16
d.f.). Since higher taxa may contain several species with independent
distributions, their inclusion may bias the results of this statistic.
TABLE II-I. Number of taxa collected near a dense Callianassa californiensis bed.
Spring Summer Fall Winter Spring Summer Fall Winter 2 year
1982 1982 1982 1983 1983 1983 1983 1984 total
number of samples per area 3 6 6 8 5 5 5 5 43
(5 below bed) (42 below bed)
number of taxa taken per area:
within bed 20 25 23 19 15 13 18 18 39
transition zone 18 21 24 20 17 20 17 18 35
below bed 15 25 26 22 20 22 22 25 39
......
0'
17
Repetition of the analysis using only organisms identified to the
generic or species level also indicates no relationship between shrimp
density and species number (G=2.41; 16 d.f.). There is likewise no
difference in the total number of taxa from areas of high and low shrimp
density pooled for the two year sampling period (Table II-I).
Even if high Callianassa density does not affect the number of
resident species, it may produce changes in the abundance of some. In
particular, one might expect negative effects on suspension feeding
(Rhoads and Young, 1970) or sedentary organisms (Woodin, 1974;
Brenchley, 1981). The effects of Callianassa abundance in South Slough
were examined for the ten most abundant taxa (Table 11-2), representing
over 98% of the animals collected during this study.
Three of these ten were polychaete worms belonging to the family
Spionidae: Pseudopo1ydora kempi japonica, Pygospio e1egans, and
Streb10spio benedicti. All three are sedentary, tube-dwelling deposit
feeders that collect detrital material from the tideflat surface using
ciliated palps (Faucha1d and Jumars, 1979); although Pygospio and
Pseudopo1ydora may also be facultative suspension feeders under certain
current regimes (Taghon et al., 1980). Pygospio e1egans and Streblospio
benedicti are common constituents of soft-bottom communities along the
Atlantic coast of North America and have often been considered as
opportunistic colonizers (Boesch, 1973; Grass1e and Grassle, 1974;
McCall, 1977; Whitlatch, 1981; Kneib and Stevens, 1982; Dauer, 1984).
Densities of Pygospio, Streblospio, and Pseudopolydora in this study
were negatively correlated with that of Callianassa californiensis
(Table 11-3; Figure 11-2). Only with Pseudopolydora was there a
TABLE II-2. 2Mean numbers per .01 m (and standard error of mean) for the ten most abundant taxa
collected from spring 1982 to winter 1984. Sampling locations are abbreviated as H
(within bed), T (transition zone), and L (below bed).
Taxon, mobility, Spring Summer Fall Winter Spring Summer Fall Winter
and feeding % of Sampling 1982 1982 1982 1983 1983 1983 1983 1984
mode fauna location n=3 n=6(L=5) n=6 n=8 n=5 n=5 n=5 n=5
Crustacea
Tanaidacea
Leptochelia
dubia 63.5 H 674.3 223.3 946.5 831.8 169.0 450.5 950.6 453.6
m-d.f.* (32.9) (96.8) (243.5) (53.1) (20.4) (52.8) (65.6) (71.6)
T 659.0 627.8 952.2 1048.3 249.0 645.6 448.8 906.4
(35.8) (279.1) (106.8) (53.7) (30.3) (81.7) 09.9) (51.9)
L 895.3 560.4 655.3 686.5 105.4 280.2 391.6 654.4
(9.0) (97.6) (203.6) (131.9) (9.1) (24.6) (30.4) (65.1)
Ostracoda 7.1 H 44.7 21.0 30.5 41.1 74.4 34.2 27.4 4.6
m-? (17.7) (5.8) ( 6.1) (6.6) (10.4) (5.0) 0.4 ) (0.9)
T 16.7 110.8 177.7 101.3 6.8 166.2 48.4 10.6
(8.4) (46.6) (46.7) (12.4) (1. 5) (24.3) (9.4) (1. 7)
L 39.3 255.4 182.8 136.6 71.6 16.4 1.8 0.8
(28.2) (44.4) 03.8) (20.5) (8.4) (2.2) (0.6) (0.6)
Amphipoda
Eobrolgus
spinosus 6.6 H 100.7 24.0 34.3 99.9 112.0 55.0 76.6 44.4
m-s (26.3) (10.9) (5.8) (6.8) (21.4) (9.7) (8.0) (3.8)
T 102.7 66.7 129.5 110.4 82.4 112.6 50.6 41.6
(9.7) (31.6) (63.4) (7.4 ) (6.2) (23.0) (5.6) (2.9)
L 135.3 70.4 27.5 12.8 1.2 0.4 1.4 1.8 .....
(10.7) (27.3) (6.6) (3.8) (0.4) (0.2) (0.7) (0.7) 00
TABLE 11-2. Continued.
Taxon, mobility, Spring Summer Fall Winter Spring Summer Fall Winter
and feeding % of Sampling 1982 1982 1982 1983 1983 1983 1983 1984
mode fauna location n=3 n=6(L=5) n=6 n=8 n=5 n=5 n=5 n=5
Corophium 2.8 H 2.3 4.0 73.2 16.5 4.6 4.2 6.6 1.8
spp. (1.4 ) (1.2) (12.4) (3.0) (2.4) (1. 6) (1.5) (0.7)
s-? T 3.7 64.8 50.2 31.9 5.4 12.8 19.2 3.0
(1.8) (25.5) (7.1 ) (3.5) (0.8) (4.7) (3.0) ( 1.0)
L 7.3 197.8 63.8 5.1 0.6 3.2 57.2 0.8
(1. 8) (48.0) (12.7) (1.4) (0.4) (1.0) (15.7) (0.4)
Grandidierella 1.2 H 0 0 0.2 3.3 0.6 10.8 9.4 3.0
japonica - - (0.2) (0.8) (0.4) (2.4) (0.9) ( 1.0)
s-d.f. T 8.0 0.2 1.0 6.4 6.4 26.4 52.2 18.4
(3.5) (0.2) (0.8) (1.4) ( 1.4) (6.0) (9.9) (5.7)
L 0 1.2 0.3 9.3 1.2 27.6 71.0 7.6
(1. 2) (0.3) (2.1) (0.6) (5.5) (3.7) (2.6)
Cumacea 2.1 H 3.7 3.0 1.8 38.9 0.8 1.0 1.2 0.6
m-d.f. (1.7) (0.9) (1.2) (6.3) (0.6) (0.3) (0.6) (0.2)
T 43.3 28.3 7.2 43.1 33.4 3.8 2.0 2.8
(18.2) (12.6) (2.0) (8.0) (17.5) (1.2) (0.4) (0.5)
L 94.0 67.4 18.8 26.1 24.2 21.6 5.4 0.6
( 9.1) (6.3) (1.8) (4.3) (5.0) (7.6) (1.8) (0.4)
Oligochaeta 11.1 H 108.0 197.7 77.7 80.1 37.8 24.2 57.6 119.8
m-d.f. (13.1) (61.1) (12.9) (11.3) (5.5) (1.5) (12.3) (13.1)
T 98.5 245.2 78.0 129.1 48.4 119.2 114.6 225.0
(12.5) (73.9) (19.7) (14.6) (23.6) (47.2) (29.4) (63.9)
L 218.7 85.8 75.3 98.5 58.4 45.6 76.6 98.4 ......
(81.7) (27.5) (27.4) (9.9) (5.5) (3.9) (4.5) (8.9) \0
TABLE 11-2. Continued.
Taxon, mobility, Spring Summer Fall Winter Spring Summer Fall Winter
and feeding % of Sampling 1982 1982 1982 1983 1983 1983 1983 1984
mode fauna location n=3 n=6(L=5) n=6 n=8 n=5 n=5 n=5 n=5
Polychaeta
Spionidae
Pseudopolydora 1.5 H 9.7 54.4 56.3 11.0 13.0 26.8 17.2 5.8
kempi (6.2) (5.1) (4.5) (l.8) (2.8) (4.7) (3.5) (2.0)
s-d.f. T 2.0 30.2 24.0 3.4 9.8 11.4 29.8 7.8
(0) (l4.4) (6.4) (0.8) (2.3) (3.0) (7.7) (2.6)
L 9.7 54.4 56.3 11.0 13.0 26.8 17.2 5.8
(6.2) ( 5.l) (4.5) (l.8) (2.8) (4.7) (3.5) (2.0)
Pygospio 1.5 H 14.7 7.2 10.7 3.0 0 0.6 0.4 0.2
elegans (7.7) (0.9) (1. 9) (0.9) - (0.2) (0.2) (0.2)
s-d.f. T 11. 7 24.7 7.8 11.4 1.6 0.4 0.8 0.8
(2.2) (7.1) (l.1 ) (3.3) (0.4) (0.2) (0.4) (0.4)
L 29.0 40.6 84.3 46.0 8.6 1.6 14.0 9.0
(7.6) (l8.9) (38.8) (9.8) ( 1.8) (0.9) (2.8) (2.8)
Streblospio 0.8 H 2.3 0.2 0.5 0 0 0 0 0.2
benedicti (0.7) (0.2) (0.3) - - - - (0.2)
s-d.f. T 0 0.5 1.2 0 0 0.2 0 0.4
(0.2) (0.7) - - (0.2) - (0.4)
L 0 0.4 9.0 8.1 20.6 52.0 60.8 29.2
(0.2) (3.6) (l.8) (5.1) (7.0) (10.6) (3.6)
*m=mobile; s=sedentary; d.f.=deposit feeder; s.f.=suspension feeder; ?=feeding mode not known. N
o
21
FIGURE II-2. Seasonal abundance of spionid polychaetes.
for means and standard errors are given in
Circles: within Callianassa bed; squares:
zone; triangles: below Callianassa bed.
Exact values
Table 1I-2.
transition
22
~
80 I \
.
I \ PYGOSPIO STREBLOSP1O
N~
_6 I 60 6~ \ cf/\
..... .
>- \~ I . \lit 6 Iz .6, \ .1M40 40 \a / //z
c C; \ / 61M~20 .\ 66. .I
~. / "6
.6-·CJ
/
\ ~,
.
/ /
.
0 SP SU FA
1982
PSEUDOPOLYDORA
60
6
..... 40
>-~
-lit
Z
...
a
Z
c
...
~
C1--t:
. \I .
. \I .
.~, \ 6 q
I/·'q· i\~\
./ \ \ j' 1'6'd/ \ . ',\ A._'C::; d/ \'-J 0--
I /' ~c"""--
significant seasonal effect on abundance, and this species showed a
slightly significant interaction between season and shrimp effects
(Table 11-3). The abundance patterns of the other two spionid species
indicate strong yearly fluctuations that may have obscured seasonal
variations.
Three tube-building crustaceans were also commonly observed in
this study, the amphipods Corophium and Grandidierella japonica and
the tanaid Leptochelia dubia. Several of the Corophium were
identified as Corophium brevis; however, the small number of
individuals and lack of female specimens on certain dates made it
difficult to ascertain whether this was the only species of Corophium
present. Since members of the genus Corophium use filtering
mouthparts to collect detritus from the tideflat surface (Rudy and
Rudy, 1983; Miller, 1984), they do not easily fit into the usual
trophic categories and have been variously classified as both surface
deposit feeders and suspension feeders (Brenchley, 1978, 1981; Bird,
1982). Grandidierella and Leptochelia deposit feed or scavenge (Smith
and Carlton, 1975; Mendoza, 1982).
Corophium and Grandidierella exhibited a spatial pattern similar
to that of the tube-dwelling polychaetes, with lower abundances within
the ghost shrimp bed and higher numbers in areas of low Callianassa
density (Table 11-3; Figure 11-3). The numbers of both amphipods
varied seasonally. Strong seasonal variation in the effect of shrimp
density on abundance was observed for Corophium. The interaction
between shrimp density and seasonality was only slightly significant
for Grandidierella.
23
24
TABLE 11-3. F-statistics for the effect of shrimp density, season and
shrimp*season interactions on the abundance of the ten most
common taxa sampled at Va1ino Island. Degrees of freedom
are 2 for shrimp density, 3 for season and 6 for
interaction.
Shrimp Season*Shrimp Density
Taxon Density Season Interaction
Leptoche1ia dubia .99 11.52 1.38
N.S. p<.OOOl N.S.
Ostracoda 9.68 2.50 .61
p<.OOOl N.S. N.S.
Eobro1gus spinosus 32.77 3.29 .99
p<. 000 1 p<'025 N. S.
Corophium spp. 6.79 32.50 6.56
p<.005 p<.OOOl p<.OOOl
Grandidiere11a japonica 11. 52 10.61 1.68
p<.OOOl p<.OOOl N.S.
Cumacea 37.02 10.79 8.18
p<.OOOl p<.OOOl p<'OOOl
01igochaeta 1.04 2.79 2.40
N.S. p<.05 p<.05
Pseudopo1ydora kempi 42.10 20.68 2.20
p<.OOOl p<.OOOl p<.05
Pygospio e1egans 39.03 .86 2.54
p<. 000 1 N.S. p<.025
Streb1ospio benedicti 96.50 2.27 1.34
p<.OOOl N.S. N.S.
25
FIGURE 11-3. Seasonal abundance of tube building crustaceans. Exact
values for means and standard errors are given in Table
11-2. Circles: within Callianassa bed; squares:
transition zone; triangles: below Callianassa bed.
26
UPTOCHELIA1000
......
~
...
-1ft
Z
...
c
Z
~
...
S 25
....~ 7
o
.
COROPHIUM GRANDIDIERELLA
200 0 80I~
. \
I .
.
\
....S 150 I 60
.
• \- I~ .
..... • \~ I
...
-
.
1ft 100 I \ 40Z
... 0 .c I \
Z . 0,~ J 6... I " \S50 . / o'
J I \
. I . .I I / \
• I
'_0, " Q.II .....
"
o SP SU A WISP 5 I
1982 1983 1984
27
In contrast to the tubicolous amphipods, the tanaid Leptochelia
dubia did not have lower numbers within a dense ghost shrimp bed than
outside the bed. Highest abundances occurred in the zone of
transitional (intermediate) ghost shrimp density during seven of the
eight seasons sampled.
The four remaining taxa, cumaceans, ostracods, oligochaetes, and
the amphipod Eobrolgus spinosus, are generally considered mobile forms
(Smith and Carlton, 1975; Brenchley, 1978; Barnes, 1980). The dominant
cumacean species were Cumella vulgaris and Hemileucon comes, of which at
least Cumella is a deposit feeder (Weiser, 1956). The two are not dealt
with separately because of problems in distinguishing them in early
samples, but they showed similar distributions with respect to shrimp
density during the second year of sampling. Total cumacean abundance
showed a decreasing trend over the two years, while varying with both
season and Callianassa density (Table 11-3; Figure 11-4). Highest
numbers during summer, fall, and the first spring were observed in areas
of low ghost shrimp abundance; however, the densities within the ghost
shrimp bed and along the transition zone increased dramatically in
winter, suggesting seasonal variation in the effect of ghost shrimp
disturbance.
Ostracods proved difficult to identify and classification is only
to the class level. As a group, ostracods are generally mobile, but
display a diverse array of feeding types, including carnivory,
herbivory, scavenging, filter feeding, and deposit feeding (Devine,
1980). Ostracod numbers were initially high in areas of low Callianassa
28
Figure 11-4. Seasonal abundance of mobile organisms. Exact values for
means and standard errors are given in Table 11-2.
Circles: within Callianassa bed; squares: transition
zone; triangles: below Callianassa bed.
29
100
o 5P SU FA WI
1982
1982
o 5P 5U FA WI
OLiGOCHAETA
o 5P 5U FA WISP 5U FA WI
1982 1983 1984
150
6
j\\ OSTRACODAI .
. ~~ ~
I / \\ 1\
. \6I / \\ I \
/. ~ ~\ / \• I \ \ I \
I I \ . I \
. I q
~J"'\\
\
160
200
......
>~
-1ft
Z;80
Z
C
III
~40
30
abundance but declined steadily after the summer of 1982 (Figure 11-4).
During the second year of sampling, highest ostracod abundances were
along the lower edge of the shrimp bed. Density did not vary
significantly with season.
The gammarid amphipod, Eobrolgus spinosus (=Paraphoxus, see Barnard
and Barnard, 1981), is an active burrower (Smith and Carlton, 1975;
Brenchley, 1978, Bird, 1982) classified by Brenchley (1978, 1981) as a
suspension feeder. Eobrolgus abundance in areas of low Callianassa
density showed a pattern similar to that of the ostracods, with
initially high abundances declining steadily throughout the two year
period (Figure 11-4). During most sampling periods, highest Eobrolgus
density was observed from within the ghost shrimp bed or from the
transition zone. Although numbers of this amphipod varied seasonally,
interactions between season and shrimp density did not appear important.
Dr. R. O. Brinkhurst (Institute of Ocean Sciences, British
Columbia, Canada) identified oligochaetes from several of the South
Slough samples as Limnodriloides victoriensis Brinkhurst and Baker
(written communication). It is not possible to determine if other
species of oligochaetes also occurred commonly in the study area.
Macrofaunal marine oligochaetes are generally considered to be mobile
deposit feeders ( Smith and Carlton, 1975; Brenchley, 1978). No
significant pattern of abundance was observed for these worms with
respect to Callianassa density (Table 11-3; Figure 11-4).
To summarize the results, distributional patterns for eight of the
ten most abundant organisms observed in this study were correlated with
Callianassa density. Seven of these showed highest abundance with low
31
or intermediate shrimp numbers. They included four sedentary, deposit
feeding species, one mobile, deposit feeding taxon, and one sedentary
and one mobile taxon of uncertain feeding mode. One mobile, suspension
feeding species achieved highest densities within the ghost shrimp bed.
The two most abundant taxa, oligochaetes and a tanaid, did not seem
significantly affected by ghost shrimp activity. Interactive effects
between season and ghost shrimp density were strong for two species.
Discussion
Although marine animals have often been thought to have little
effect on their environment, Callianassa activity appears a potentially
important structuring agent in certain habitats. Bird (1982) suggested
that C. californiensis reduces sediment organic content and affects
sediment grain size and sorting. A similar effect has been noted for a
Caribbean species (Suchanek, 1983). Increased resuspension rates
associated with C. californiensis have been reported from laboratory
experiments (Brenchley, 1978; Murphy, 1983) and from field studies on a
tropical species of Callianassa (Aller and Dodge, 1974). Records of
Callianassa bioturbation also appear in geologic studies of near-shore
sediments (Picket et al., 1971; Dewindt, 1974; Moreno and Curros, 1979).
Several studies have discussed the effects of Callianassa activity
on the infaunal community. Brenchley (1978) and Bird (1982) noted a
negative correlation between the abundance of several sedentary
organisms and dense C. californiensis beds, including spionid
polychaetes, the amphipod Corophium, and the tanaid Leptochelia dubia.
Highsmith (1982) also provided evidence that Leptochelia was susceptible
32
to bioturbation. Peterson (1979), Murphy (1983), and Stevens (1928)
found that C. californiensis may exclude certain bivalve species and
Ronan (1975) showed that the ghost shrimp may inhibit a tube-dwelling
phoronid. Aller and Dodge (1974) observed a negative effect of a
tropical species of Callianassa on suspension feeding organisms.
Turt1egrass growth was reduced in the presence of a Caribbean species
(Suchanek, 1983). Spionid polychaete distributions were negatively
correlated with that of C. ceramica in Australia (Dorsey and Synnot,
1980). Bird (1982) also observed significantly fewer species within a
C. californiensis bed compared to beds of the sedentary burrowing
shrimp, Upogebia puggettensis, but no data were provided from habitats
in which neither species was common.
This study of C. californiensis at South Slough differs from others
in observing community changes on the scale of only a few meters. Yet
even along the lower edge of the shrimp bed examined here, abundances of
several species were strongly correlated with ghost shrimp density.
However, the number of species did not change along this gradient, as
might have been expected from Bird's (1982) results. Since Bird
compared samples from within a Callianassa bed to those taken within a
Upogebia bed, the difference in number of species may be partially a
result of positive effects associated with Upogebia, such as increased
area for surface contact and increased sediment stability (Thompson,
1972). Also, Bird used sampling locations separated by more than 100
meters along an estuarine gradient. The greater distance between
samples could imply greater habitat heterogeneity that may have
contributed to differences in species number.
33
The results from the benthic samples presented thus far have been
correlative, involving comparisons of samples taken in different areas.
In order to demonstrate a causal relationship between shrimp density and
faunal abundances, it would be desirable to manipulate shrimp numbers
and observe the response patterns of other members of the community.
Bird (1982) placed containers of defaunated sediment within a northern
Oregon ghost shrimp bed and, after six months, observed an almost five-
fold increase in spionid and capitellid polychaete densities as compared
to the surrounding bed. Corophium densities increased to three or four
times that of a dense bed. After 24 to 36 months, Callianassa had
successfully invaded the exclosures and the abundances of polychaetes
and amphipods declined to those of the surrounding area. Bird's
results, along with laboratory observations of the effect of ghost
shrimp disturbance on selected organisms (Brenchley, 1978; Murphy,
1983), suggest that the distribution patterns seen in this study are at
least partially a result of Callianassa activity.
The effect of a decrease in Callianassa activity during winter is
unclear. Assuming some migration of benthic fauna, one might expect a
reduction in the differences between areas of low and high ghost shrimp
density with a decrease in their bioturbating activities. Such a
pattern occurred among the cumaceans and the amphipod Corophium
(although for Corophium this coincided with a period of generally low
numbers). However, for other species interactive effects were either
nonsignificant or slight in comparison with the main effects of season
and shrimp density. One possible explanation for this result may be
that the application of a transformation masked interactive effects.
34
Since one effect of a transformation is to produce additivity in the
data, nonadditive interactions may not be apparent after its
application. However, examination of the abundances of t~e ten
organisms over time (Figures 11-2, 11-3, 11-4) does not suggest
interactions between season and shrimp density which were not detected
by the statistical analysis. A second reason for this lack of an
interactive effect may be an error in assuming that adult migration
occurs over several meters. Among spionid polychaetes, both Streblospio
benedicti and Pygospio elegans have been reported to emigrate in
response to stress (Levin, 1982). If there is no such emigration under
low stress conditions in the field, if it involves only short distances,
or is limited in the winter months, little reduction in spionid zonation
would occur. Also, winter and spring represented a time of low
abundance in all three areas for several species, suggesting low
settlement or survival at this time.
The presence of both mobile and sedentary deposit feeders in the
South Slough community allows an examination of the extent to which
functional group models based on mobility mode or trophic group
interactions explain the influence of Callianassa on species abundance.
Five deposit feeding organisms reached highest numbers outside a shrimp
bed, whereas numbers of one suspension feeder were positively correlated
with Callianassa abundance. Two other deposit feeders did not show a
significant difference in distribution with respect to the shrimp bed.
The trophic group amensalism hypothesis (Rhoads and Young, 1970) would
not predict a negative effect of a deposit feeding shrimp on other
deposit feeders, but would predict a negative effect upon suspension
35
feeders. A mobility mode hypothesis predicts a negative effect by ghost
shrimp on sedentary organisms and a neutral or positive effect on mobile
animals. Five of the six sedentary organisms examined in this study
were negatively correlated with Callianassa density, as were two of the
four mobile taxa. One mobile species was positively associated with the
shrimp bed, and one mobile and one sedentary form were not significantly
affected. Most of these relationships appear consistent with
predictions of a mobility mode hypothesis; however, the exceptions
underscore the need for a more thorough understanding of species' habits
and other potential interactions on community composition. Levin (1981,
1982) has shown that aggressive competition among spionid polychaetes
(Pseudopolydora and Streblospio) is important to the spatial
distribution of these worms. Similar interactions may provide an
explanation for the lack of an increase of Leptochelia in areas with low
Callianassa density, as was observed in this study. This relatively
small crustacean may be particularly susceptible to aggression from
larger polychaetes or amphipods. Levinton (1972, 1977) proposed that
deposit feeding bivalve communities may be regulated by exploitive
competition for food. Although exploitive competition would not explain
the rapid response of many species to Callianassa under laboratory
conditions (Brenchley, 1978), it could be important in specific cases,
in this study possibly for cumaceans.
While the distributions of most infauna examined here tend to
support a mobility mode of interaction, these conclusions should not be
generalized beyond this habitat. Rhoads and Young (1970) suggested that
trophic group amensalism may occur only in subtidal areas, although
36
several researchers have noted distinct assemblages of deposit and
suspension feeders in intertidal regions as well (Bloom et al., 1972;
Myers, 1977; Whitlatch, 1977). As discussed previously, several
studies, mostly from subtidal areas, have provided support for the
trophic group amensalism hypothesis while Aller and Dodge (1974) and
Murphy (1983) have found evidence that Callianassa spp. can affect flora
and fauna subtidally through resuspension of particles rather than
strictly via sediment disruption. It is becoming increasingly apparent
that several modes of interaction among functional groups may be
important, depending on the physical environment or species involved.
37
CHAPTER III
EXAMINATION OF ZONATION AND BEHAVIORAL PATTERNS
IN CALLIANASSA CALIFORNIENSIS
Introduction
Because ghost shrimp live within relatively deep burrows, their
density is difficult to measure directly without the excavation of large
quantities of sediment. Aside from being time consuming and labor
intensive, excavation prevents the monitoring of an area over time.
Several researchers have used counts of burrow holes or mounds to
estimate ghost shrimp density (MacGinitie, 1934; Aller and Dodge, 1974;
Ott et al., 1976; Swinbanks and Murray, 1981; Suchanek, 1983). While
some of these studies have provided estimates for the number of holes
per shrimp, ranging from 1 to 4 among the various species of Callianassa
(Pohl, 1946; Biffar, 1971; Ott et al., 1976; Swinbanks and Murray, 1981;
Suchanek, 1983), they have not confirmed a statistically significant
relationship between the density of burrow holes and ghost shrimp
density. Devine (1966) and Frankenberg et ale (1974) have suggested
that the number of holes/burrow may be highly variable, and Thompson and
Pritchard (1969) and Ott et ale (1976) considered holes to be relatively
impermanent structures that collapse during each tidal cycle.
Behavioral patterns of ghost shrimp are likewise difficult to
observe outside of narrow glass tanks (limnoria) or artificial tubes.
However, several authors have provided field observations of Callianassa
38
behavior that suggest that some members of this genus are susceptible to
predation by epibenthic animals. Torres et ale (1977) observed C.
californiensis activity near the burrow entrance immediately after
submergence and McCrow (1971) found that this ghost shrimp moved towards
the surface when the tide covered the burrow. Carson (1955) noted that
C. major can be enticed out of its burrow with small pieces of meat and
Pohl (1946) suggested that C. major may leave its burrow for short
periods. C. longiventris remained near the entrance of their burrows
until disturbed (Biffar, 1971) and C. fi1ho1i feeds on the surface layer
of sand (Devine, 1966).
This chapter will examine the relation between hole and shrimp
density in nature as well as observations of Callianassa behavior within
laboratory aquaria. Hole densities will be used to examine zonation
patterns in four estuaries. A specific aim of the behavioral
observations was to determine whether C. californiensis spends
appreciable time near the burrow entrance. Potential seasonal
variations in activity will also be examined in relation to temperature
fluctuations and through laboratory tests of the effect of reduced
salinity on activity.
Methods
Hole/Shrimp Correlations and Zonation Patterns
The relationship between the density of burrow holes and the number
of ghost shrimp was examined by counting the number of holes in a
2haphazardly placed 0.5 m quadrat and then taking three 50 cm deep by
39
12.5 cm diameter cores within the same area. The contents of the cores
were combined and sifted by hand to determine ghost shrimp density. A
few samples were reseived through a 1 mm mesh screen to ensure the
accuracy of the hand seiving method. Samples were taken at low tide
during four time periods: 1) July 9 to August 10, 1982; 2) November 10,
1982 to March 4, 1983; 3) June 21 to July 10, 1984; and 4) July 29 to
August 4, 1984. All samples came from the Valino Island and Battle
Flats areas (Figure 11-1). Holes were counted monthly in three
2
replicate 0.25 m quadrats in a dense shrimp bed at Valino Island.
Hole density was observed qualitatively at several locations within
Coos Bay that contained dense aggregations of Callianassa (numbered
areas in Figure 11-1) between 1982 and 1984. Surveys were made only
during daylight tides lower than mean lower low water and generally
involved walking from the wrack line down to the water's edge. Tidal
levels were not measured directly, but were estimated for some areas by
repeatedly comparing observed high and low tides with those predicted
from a tide table and with a tide gauge in the Charleston boat basin.
During 1983, ghost shrimp beds were also surveyed in Winchester Bay (at
the mouth of the Umpqua River), along the south causeway of Tillamook
Bay, the eastern side of Sand Lake, and in several areas of both North
and South Bay, Humboldt Bay, California.
Behavioral Observations
General observations of Callianassa californiensis behavior were
made in a 190 1 aquarium. The aquarium had a flow-through water system
and was filled to a depth of 0.4 m with sifted sand from a moderately
40
dense shrimp bed. The substrate surface remained oxygenated throughout
the study. Daylight observations were in natural light and a red light
was used at night.
Because of difficulties in observing and timing behaviors of shrimp
burrowing in sediment, ghost shrimp behavior was quantitatively observed
in an artificial mudflat aquarium that simulated burrow conditions
(Hoffman, 1981). The aquarium was divided into two compartments, each
with three suspended U-shaped clear plastic tubes (2 cm inner diameter).
One compartment received seawater (30-33 0/00 salt during the duration
of the study) pumped from the mouth of Coos Bay. The salinity of the
water entering the other compartment could be varied from that of the
obay down to 0 /00. The entire apparatus was located out-of-doors and
the temperature was allowed to vary. Since ghost shrimp normally
receive light only from the surface, a black plastic skirt was used to
reduce light exposure to the plastic tubes.
Ghost shrimp were exhumed from Battle Flats and were transferred
directly from the field to the observation aquarium. One shrimp was
added to each of the plastic tubes. They were left undisturbed for 1-2
days and then were observed for a single 20 minute period. Because of
occasional rapid changes between behaviors, timing of a given pattern
was not begun until it had lasted at least 10 seconds. General
observations of behavior patterns in this chapter are presented only for
salinities of 30 0/00.
In examining the effect of salinity on activity, salinity in the
experimental compartment was changed gradually and the shrimp were
allowed to acclimate for 3-4 days before observations were begun
41
(Thompson and Pritchard, 1969). Observation periods were 20 minutes as
described above. Linear regression, blocking for time of observation
(morning, midday, or evening) was used to examine the effect of
decreasing salinity on the time shrimp spent actively moving within the
tubes. The data were recorded as percentages of the total observation
time and were arc sine transformed before analysis (Sokal and Rohlf,
1981). The regression was calculated using the General Linear Models
Procedure of the Statistical Analysis System (Ray, 1982).
Results
Hole-to-Shrimp Ratios
For both July/August periods, hole density and number of ghost
shrimp recovered in the cores were relatively well correlated (Table
III-I, Figure III-I, r~.88 for both periods). The relationship between
holes and shrimp density was also highly significant during the longer
1982/1983 winter period, although the correlation is lower (r=.79). The
slope of the line for the winter data is higher than that for the 1982
summer (F=8.22; p<.OI), suggesting that there are fewer holes per
individual between November and March. Monthly monitoring of hole
density at Valino Island also suggests a seasonal variation (Figure 111-
2). Hole densities were lowest during mid-winter and increased to a
peak during mid-summer. As discussed in Chapter 5, peak recruitment of
Callianassa generally happened during late summer.
In contrast to the other three periods, the relationship between
hole and shrimp density was not significant during late June and early
42
TABLE III-I. Statistics for regression of shrimp density on hole
density.
Source of Sum of
Period Variation Squares d.f F P r
July 9-Aug. 10, 1982 Hole Density 30.24 1 39.46 .0001 .88
Error 9.43 12
Nov. 10-March 4, 1983 Hole Density 49.01 1 16.76 .005 .79
Error 29.24 10
June 21-July 10, 1984 Hole Density 26.13 1 2.35 .2 .45
Error 100.05 9
July 29-Aug. 4, 1984 Hole Density 71.67 1 77 .53 .0001 .94
Error 9.24 10
July, 1984. However, higher shrimp numbers were still associated with
higher hole density and most of the points lay between the July/August
and winter values.
Zonation
Using holes as an indicator of relative density, there was a distinct
zonation of dense Callianassa beds in 18 of the 19 sites examined in Coos
Bay (Figure II-I). The beds occurred predominantly in the mid intertidal
with a lower limit in many areas between +0.3 and +0.9 m above mean lower
low water (MLLW). The vertical position of the lower edge often varied by
as much as 0.3 m within a site. Short transects taken across the lower
boundary indicate that the transition from high to low density is
relatively abrupt in some areas (Figure 111-3). However, this abruptness
appears to vary seasonally and between areas.
43
Figure III-I. Relationship between hole density and Callianassa
abundance. Lines represent best fit regression lines.
F-statistics and correlation coefficients are given in
Table III-I.
••12
10
44
•
6 • •C")~
"0.
11& 4 0
1M
..
..
~
- 2lie
:z:
tit
0
O:JUL- AUG 1982
.:NOY-MAR 198211983
0 60 100 140 180 220
~ 8 • 0
"0.
11&
1M 6 e
..
0
..
~
- 4 011&
:z:
tit
0 • •
2 •
e:JUN-JOL 1984
O:JUL-AUG 1984
0 2 60 100 140 2 180 2 0
HOLES PER .5 M
100
.J A SON D .J F
1982
M A M J .J A SON D
1983
F M A M .J .J
1984
Figure 111-2. Monthly fluctuations of hole densities in a Callianassa bed. Small bars represent 1
standard error (n=3 for all dates).
.t:-
Vl
46
Figure 111-3. Hole densities from transects taken across the lower edge
of three Callianassa beds. Small bars indicate 1
standard error (n=4 for all means) and 0 m represent an
arbitrary starting point in the lower portion of the bed.
A-I: Battle Flats, north bed, August 1984; A-2: Battle
Flats, north bed, February 1984; B: Valino Island,
January 1984; C: Battle Flats, south bed, August 1984.
47
1 2 3
DiSTANCE M
o 1 2 3
DISTANCE M
80
70 70
N~ 60 • 60 C
1ft SO SO
~
40 40
•~ 30 30
d 20 20
Z
10 10
0 0
0 1 2 3 .4 0 2 .4 6 8 10
DISTANCE M DISTANCE M
7
N
~ SO
1ft
C"!40
=30Go
d 20
Z 10
48
The one site that did not show a pronounced intertidal zonation was
near a causeway at the mouth of North Slough Coos Bay (designated
"causeway" in Figure II-I). A Callianassa bed of moderate density at
this site extended below 0 m MLLW. The area appeared to receive seepage
from a nearby paper mill and the sediment had a high concentration of
wood fragments.
A single site examined in Winchester Bay had a dense bed that
extended to just above MLLW. In Tillamook Bay, dense beds were limited
primarily to the mid intertidal and numbers declined in the low
intertidal and subtidally. In the site examined at Sand Lake, however,
dense Callianassa beds extended subtidally. Within Humboldt Bay,
California, C. californiensis were generally less common with no beds
approaching the densities seen in Oregon.
Behavioral Observations
Callianassa showed a varied repertoire of behaviors within the
sediment-filled tank. For the purposes of this study, these behaviors
were grouped into five general categories: burrowing; cleaning; pleopod
movement (ventilation); aggressive interactions between shrimp; and
sedentary behavior. Burrowing involved the excavation of sediment,
packing of burrow walls using the walking legs, and transport of
particles (often to the substrate surface). Feeding, which often
involved sifting sediment with the maxillipeds, could not be readily
distinguished from nonfeeding activities. The most common method of
transporting sand out of the burrow was to carry the sediment to within
1-3 cm of the substrate surface and then to expel it forcibly with a
49
burst of pleopod activity. Plumes of sediment coming from ghost shrimp
burrows could also be seen while diving over a dense bed at high tide.
This activity resulted in cones similar to those reported from other
Callianassa beds (Aller and Dodge, 1974; Ott et al., 1976; Suchanek,
1983). If the burrow was in the initial stages of construction, the
shrimp often pushed the sediment out of the burrow rather than using its
pleopods to blow the sediment out. This technique was used even after
shrimps had constructed a turn-around chamber and usually resulted in an
individual being partially exposed.
Cleaning was most commonly directed at the gills, antennae, feeding
appendages (maxillipeds), and the egg masses of females. Pleopod
movement often occurred simultaneously with cleaning and burrowing.
Aggressive interactions happened when two burrow systems met and usually
ended where one of the shrimp walled off the opening into its burrow. A
lack of detectable movement was classified as sedentary behavior.
Most of the patterns seen within the sediment-filled tank were also
observed within the artificial burrows, including pleopod movement,
cleaning, and sedentary behaviors (Table 111-2). Burrowing was not
possible within the plastic tubes; however, shrimp spent a considerable
amount of time moving within these tubes and often carried sediment that
had fallen into them. Since ghost shrimp often tried to carry sediment
or "pack" the tube walls while moving, it is likely that this behavior
is related to burrowing in sediment. Active movement was the most
frequently observed behavior.
As mentioned earlier, several field observers have noted ghost
shrimp near the entrance of their burrow. Ghost shrimp within the
50
TABLE 111-2. Duration of behaviors observed in artificial burrows.
Number of Mean Min. Per
Observation 20 Min.
Behavior Periods Period S.E. % of time
Active Movement 119 9.3 .7 46.6
Sedentary 116 5.5 1.1 27.5
PIeopod Movement 115 3.7 1.2 18.3
Cleaning 55 6.0 .8 29.7
Near Burrow Entrance 120 5.4 2.0 27.0
plastic tubes spent over 25% of the observation time within 2 cm of the
burrow entrance. Although they were seldom seen outside the burrows,
shrimp occasionally moved from one tube to another. This required a
shrimp to expose itself on the surface. When hanging near the burrow
entrance, a shrimp often exposed part of its chela above the substrate
surface.
When Callianassa in the artificial burrows were exposed to lowered
salinity, 75%-100% died between 8 and 9 0/00 salt, a value that agrees
closely with the results of Thompson and Pritchard (1969). These
shrimp, which apparently died of osmotic stress, all showed signs of
swelling in the gill areas. No salinity related mortality occurred at
10 0/00 salt. Only observations at higher concentrations were used to
examine the effects of salinity on behavior.
An apparent trend towards decreasing activity with decreasing
salinity (Figure 111-4) is not significant at the .05 level (p<.066).
51
Figure 111-4. Relationship between salinity and the per cent of time
spent actively moving within artificial burrows. Bars
represent + 1 standard deviation, and sample sizes are
given above the bars.
100
3
S2
80
6
CD
:II:
-
6
>0
=- 7
1M 9
=-
17
-...
6
...
0 37
ae 40
t58 I9 5
20
5
t
3
0
10 15 20 25 30 35
SALINITY poo)
53
However, much of the deviation is due to the high activity seen at 15
o /00. Unlike most of the other salinity levels, these three values all
stem from a single mid afternoon observation period. Since the tank is
located outdoors, unusual conditions, such as warmer temperatures, could
have contributed to the higher activity. Dixon's test for extreme
observations (Snedecor and Cochran, 1980) indicates that this mean is an
outlier when compared to the other data (ratio=.593; p<.05).
Recalculation of the regression excluding these three values suggests a
significant decline in activity with reduced salinity (F=6.15; p<.015;
r=.35).
Discussion
Holes appeared to be a relatively good estimator of shrimp density
within a given time period. The one sampling period that did not show a
significant relationship, June 21 to July 10, 1984, did show a
concordant trend, although the points appear to fit a~ exponential
distribution better than a linear one. In previous studies of the
number of holes per shrimp in C. californiensis beds, Swinbanks and
Murray (1981) reported a mean of 2.5 holes per shrimp and MacGinitie
(1934) "two to several openings to a burrow." In this study, the number
of holes per shrimp ranged from a high of 4.2 during the summer of 1982
to a low of 3.8 during the 1982/1983 winter. Since ghost shrimp may
burrow below the 0.5 m sampled by the corer (MacGinitie, 1934) and some
individuals may be able to evade a corer, these values probably
overestimate the number of holes per individual.
Although holes provide a good estimate of relative shrimp density
54
within the same time interval, the number of holes per shrimp varies
seasonally. This is shown both by the lower hole-to-shrimp ratios and
lower hole densities in the field during the winter months. The spring
and early summer increase in hole densities cannot be attributed to
recruitment, since peak settlement of C. californiensis in Oregon
estuaries appears to occur in late August (Bird, 1982; Johnson and
Gonor, 1982; personal observation).
Since holes are produced in association with burrowing and may
collapse if not maintained (MacGinitie, 1934; Thompson and Pritchard,
1969), the winter decline in hole densities and hole-to-shrimp ratios
may indicate reduced burrowing activity at this time. Ghost shrimp in
Yaquina Bay move deeper into their burrow during winter (McCrow, 1971)
and C. californiensis produces fewer eggs between December and March
(Bird, 1982). Hailstone and Stephenson (1961) and Devine (1966)
similarly found that several species of Callianassa in Australia and New
Zealand have reduced reproduction and lowered growth rates during the
Southern Hemisphere winter (June-through August). During winter in
Oregon, lower air and water temperatures (Table 111-3) can be expected
to lead to lower activity in the ectothermic ghost shrimp. Although
salinity has only a small effect on movement, as indicated by the low
correlation, lower winter salinities would also tend to contribute to
reduced activity.
Surveys of Callianassa californiensis burrows in Coos Bay and two
other estuaries indicate that a mid-intertidal peak in ghost shrimp
abundance is common in Oregon estuaries. However, the two examples of
low-intertidal and subtidal beds suggest that C. californiensis is not
55
TABLE 111-3. Tideflat and surface water temperatures at Valino Island.
All temperatures were measured during low tide.
Date Water Tideflat
Temperature (OC) Temperature (OC)
1983
~bruary 19 10.0 10.5
March 3 12.0 12.0
March 13 12.5 11.5
April 6 15.5 15.5
May 4 16.0
May 14 12.5
May 23 20.0 18.5
June 19 17.0
June 26 18.0
July 2 17.0
July 12 17.5
July 16 20.0
August 13 18.0
August 25 18.0
September 24 16.0
October 10 12.0
October 23 13.5
November 28 12.5 12.5
December 13 12.5
1984
January 11 11.0 11.5
February 25 10.0 10.0
March 10 13.5
April 22 13.5
May 5 12.0 13.0
June 13 15.0
June 24 21.0
July 3 18.0
July 12 17.0 14.5
physiologically prevented from inhabiting these areas. In the Boundary
Bay tide flats of British Columbia, densest aggregations of ghost shrimp
occurred between +2.2 and +1 m above MLLW, with low densities below this
level (Swinbanks and Murray, 1981). McCrow (1971) noted that densest
56
beds occurred between +0.6 and +1.2 m above MLLW in Yaquina Bay.
Thompson and Pritchard (1969) observed them in a zone between 0 and +0.3
m in the same estuary, although their reference to this as "higher
intertidal" makes it unclear whether the heights are relative to mean
sea level or MLLW. MacGinitie (1934) reported densest beds between 0
and +0.3 m in Elkhorn Slough, California and Ronan (1975) observed that
Callianassa occurred in the upper intertidal of Bodega Bay. However,
Millicent Quammen (University of California, Santa Barbara, personal
communication) has indicated that dense aggregations of Callianassa
californiensis may extend subtidally in Mugu Lagoon. References to a
mid-intertidal distribution of C. californiensis also exist in Bybee
(1969) and Smith and Carlton (1975).
One problem associated with a hypothesis that predation limits the
lower distribution of these beds is how a burrowing shrimp may become
vulnerable to surface predators. The presence of ghost shrimp near the
entrance of their tubes in the artificial mudflat aquarium suggests a
possible mechanism. One can often collect adult Callianassa during high
tide by scraping the sediment by hand, indicating that they are
relatively close to the surface, and on two occasions shrimp were seen
at the entrance of their burrow in the field. During observations at
high tide of a dense bed in June, a few shrimp were also observed out of
their burrows on the substrate surface. Such behavior in the field, as
well as reports by others, suggest that the behavior observed in the
plastic tubes also represents field conditions. The explanation for
this pattern is less clear, but feeding (Bird, 1982) and mating
activities may be involved.
57
CHAPTER IV
PREDATORS OF CALLIANASSA CALIFORNIENSIS IN
COOS BAY, OREGON
Introduction
Although Callianassa californiensis is a popular bait item of
sports fishermen in Oregon and Califprnia, little is known about the
natural predators of adult shrimp. Russo (1975) reported ~
californiensis and Upogebia from the stomachs of leopard sharks, Triokis
semifasciata, while Stevens et al. (1975) found that C. californiensis
constituted the major portion of the diets of Dungeness crabs in Gray's
Harbor on at least one date. Biffar (1971) and Darnell (1958) have
noted cases where Caribbean species of ghost shrimp have been eaten by
bottom-feeding fish. The presence of Callianassa in Western Gull fecal
pellets from Coos Bay (personal observation) also suggests that
inter tidally foraging birds may occasionally take this shrimp.
For predation to restrict dense ghost shrimp beds primarily to
higher tidal levels, a situation analogous to sea star limitation of
mussel beds (Paine, 1974), an aquatic predator would have to be capable
of consuming significant numbers of this burrowing shrimp. Within Coos
Bay, and particularly at the Valino Island study site in South Slough,
bottom-feeding fish and crabs are the predominant large epibenthic
predators. Benthic cores (Appendix) and random samples taken with a
shovel indicate that most infauna immediately below the Valino Island
58
shrimp beds are relatively small. The only common infaunal predator
large enough to prey upon even juvenile ghost shrimp (8-12 mm length;
Bird, 1982) is the nemertean, Cerebratulus californiensis. In this
chapter, the potential importance of fish, the crab, Cancer magister,
the epibenthic feeding shrimp, Crangon, and the infaunal nemertean,
Cerebratulus, as predators on Callianassa californiensis will be
examined through a series of laboratory and field studies.
Methods
From 1982 to 1984, fish were collected with beach seines near the
Empire boat landing and in Joe Ney Slough, Coos Bay (Figure II-I), in
conjunction with Estuarine Biology and Vertebrate Ecology classes taught
by Daniel Varoujean at the Oregon Institute of Marine Biology. Dense
mid-intertidal ghost shrimp beds, with patchy Zostera marina beds
subtidally, were present at both sites. The net was 40 m long by 2 m
high with 1.5 cm mesh and a centrally located catch bag. Seines were
placed parallel and approximately 50 m from shore using a motorboat and
were pulled in by hand using ropes attached to the ends. In order to
sample less common fish for stomach content analysis, two beach seine
hauls were made at each site during a sampling period; however, these
consecutive seines were not considered replicates because of disruption
to the fish fauna after the first seine and microhabitat differences
(such as vegetated versus unvegetated areas) that may occur if samples
are taken far enough apart to prevent disturbance effects. Therefore,
only the first seine at a site was used to estimate relative densities
of fish. Most samples were taken during neap low tides.
59
Subsamples of the fish caught in beach seine hauls during 1982 and
1983 were immediately preserved in 10% formalin for later examination of
stomach contents. In addition to the samples from Coos Bay, stomachs
were also analyzed from fish collected during a 1981 study of the Umpqua
River estuary, Oregon.
Because beach seines sample a relatively large area, fish traps
were used to observe local fish distributions associated with dense
shrimp beds at Battle Flats and the Causeway (Figure II-I). All traps
were of 1.2 cm galvanized wire mesh and were not baited. They were set
out only after all areas were covered by at least 0.3 m of water and
were retrieved before exposure. This helped to ensure that differences
in the number of fish caught in the various zones were not due to
differences in fishing time. Shrimp beds at Battle Flats showed a
typical pattern with highest densities in the mid intertidal while the
causeway was the only site examined in Coos Bay where a dense bed
extended throughout the intertidal (see Chapter III). The distribution
of fish in three zones at Battle Flats was studied by placing traps over
a dense bed, approximately 1 m above the lower edge, along the lower
edge of the bed, and approximately 1 m below the shrimp bed (Figure IV-
1). Traps were placed during 9 time periods, and equal numbers were put
into each zone (Table IV-I). A second study involving traps was begun
to determine whether the lower extension of the shrimp bed at the
causeway site was due to lower predator abundance. Four traps were
placed at both the causeway and Battle Flats at a tidal level
corresponding to approximately 1 m below the Battle Flats bed.
Comparisons were made for two consecutive days during July 1984.
60
61
TABLE IV-I. Number of staghorn sculpins caught in fish traps during a
single high-tide period on each of nine dates. B=over
dense bed; T=over transition zone; L=below dense bed.
Date n B T L
1983
23 July 1 2 9 6
7 August 1 0 0 2
22 August 1 0 1 0
1984
19 ~y 2 0 0 7
13 June 3 2 5 11
21 June 2 0 0 9
12 July 2 0 2 2
28 July 2 0 1 4
29 July 2 7 4 2
Laboratory tanks were used to observe Dungeness crab (Cancer
magister) predation because of difficulty in catching large numbers of
adults near dense shrimp beds and problems in identifying stomach
contents (many of the items seen in Cancer stomachs were mascerated and
unrecognizable). Adult and juvenile C. magister were fed live ghost
shrimp of various sizes in clean 380 and 57 liter tanks, respectively.
Predation by adult Dungeness crabs was also observed under simulated
tideflat conditions in a second 380 liter aquarium filled to a depth of
0.4 m with sifted sand from a ghost shrimp bed. The tank had a flow-
through water system that allowed percolation of water through the
62
sediment and was divided into two halves with 1.2 cm screening. Equal
numbers of ghost shrimp were added to each side at a density of 30-
250/m. The shrimp were allowed to burrow for 3-4 days before a "single
crab was added to one side. The other side was periodically disturbed
by hand as a control for disruption due to the crab's movements. Crabs
were observed from a darkened blind which covered one side of the tank.
Daytime observations were under natural light and a red light was used
at night.
Predation by Crangon spp. and the nemertean, Cerebratu1us
californiensis, was observed in a 57 liter tank without sediment.
Crangon were observed during summer using ghost shrimp 30-40 mm long
while Cerebratu1us were observed in early fall using newly settled
shrimp (10-15 mm length).
Results
Predator Diets
Only two species of fish commonly ate Callianassa californiensis:
the cutthroat trout, Sa1mo c1arkii, and the staghorn sculpin,
Leptocottus armatus. Since fewer than 20 cutthroat trout were caught
during the 3 years of sampling, this fish is probably unimportant to the
distribution of C. californiensis.
Ghost shrimp formed a significant portion of the diet of staghorn
scu1pins from both Coos Bay sites as well as from the Umpqua River
(Figure IV-2). Over the first two years of sampling in Coos Bay, an
average of 15.9% of the staghorn sculpin had eaten ghost shrimp.
63
Figure IV-2. Percent of staghorn sculpins having Callianassa as part of
their diet. Numbers above bars indicate sample sizes.
COOS BAY
40 13
lli:J EMPIRE
35
.JOE NEY 18
30
25 29
..
Z 30
... 20
14 •
U ··ae •
...
.. 15
14 10
10
· 2~
·28
·
·
5,· ·•
·
• •
• • • ·• ·
0 :~ ~s •M J J A S 0 .. N D oJ F M A M oJ oJ A S o N
1982 1983
o ----aPR~M--Ay~-ioJ UN oJU L AUG
1981
15-
64
11
SEP
10
UMPQUA RIVER
15
20-
35-
30-
10-
5-
.. 25-
Z
...
u
-=...
..
65
Predation on C. californiensis appears to increase as summer progresses
(Figure IV-2). At the Joe Ney and Empire stations in 1983, relatively
few fish had ghost shrimp in their diets during April. The percentage
had increased by a factor of four at the two Coos Bay sites in July
1983. Within the Umpqua, however, highest consumption was observed in
April 1981, decreasing in June, and then increasing again in September.
Consumption of ghost shrimp by staghorn scu1pins also depended upon
the size of the fish. No scu1pins less than 70 mm standard length
contained C. californiensis, even though this size class contributed 58
of the 276 stomachs examined from Coos Bay. The average standard length
of Leptocottus that had eaten Callianassa in the Coos Bay samples was
HO.S mm (S.E.=3.8 mm; n=44).
Predation on C. californiensis by the three invertebrate predators
was less evident. Within tanks that lacked sediment, adult Dungeness
crabs rapidly moved their antennae and maxi11ipeds and began searching
even when a shrimp was placed where it was not visible. Similarly,
juvenile Cancer magister attacked both juvenile and adult ghost shrimp,
although they were usually unsuccessful when attacking adults and
appeared to have difficulty piercing the exoskeleton of these larger
shrimp. The juvenile crabs also tended to retreat from frontal
confrontations with the adults. Occasionally, juvenile Cancer were
successful in stripping the eggs from berried females. Despite the
readiness of adult crabs to attack ghost shrimp in clean tanks, no C.
magister tried to excavate Callianassa in the sediment-filled tanks,
none of the adult crabs in this tank had shrimp in their stomachs at the
end of the experiment, and there was no significant difference in number
66
of holes or in numbers of shrimp recovered as compared to the control
side.
Neither Crangon shrimp nor nemerteans were observed to attack ghost
shrimp even after several days of exposure. All ghost shrimp appeared
active at the end of observations.
Sculpin Distribution
Since peak numbers of juvenile staghorn sculpins occurred several
months earlier than adults and juveniles are not effective predators,
the examination of seasonal and spatial distributions of Leptocottus
will be limited to fish longer than 70 mm standard length. Within both
Coos Bay sites, this size class was most plentiful during summer. At
Joe Ney, highest catches of sculpins were usually between May and July,
although July 1983 and May 1984 are exceptions (Figure IV-3). The four
samples taken between November and April contained the fewest
Leptocottus. The Empire site showed a similar pattern with sculpin
abundances rising from a winter low to a peak during summer (Figure IV-
4). In general, during the summer months the staghorn sculpin was one
of the most common fish in areas of moderate to low wave action,
comprising up to 30% of the fish caught in a sampling period.
Catches from the fish traps placed at Battle Flats (Table IV-I)
indicate that sculpins forage in the areas occupied by dense shrimp
beds. Aside from Leptocottus, the only fish taken in the traps were two
shiner surfperch (Cymatogaster aggregata). However, sculpins appeared
to forage preferentially below the shrimp bed (G=20.96; 2 d.f.; p<.OOS),
with almost four times as many fish caught during the same time interval
67
Figure IV-3. Seasonal abundance of staghorn sculpins in Joe Ney Slough,
Coos Bay. Bars represent catches from first beach seine
haul taken on that date.
...
...
~
1 000-
e
~
...
...
0
z
-
0
-e
...
•
M
"'00
0-I ....
e
I
...
...
0
• Z
0
til
e
...
N
"'00
0-I ....
e
I
...
...
0
Z
0
til ....
00
e 0-
...
...
•
i i i i i i
0 0 0 0 0 0 0
('II O' 00 -0 ~ ('II
.l)Nin~il:J
68
69
Figure IV-4. Seasonal abundance of staghorn sculpins at Empire, Coos
Bay. Bars represent catches from first beach seine haul
taken on that date.
~~
~
cex)
0-
~
...
~
c:II
Z
-
0
III
C
• ~
.,
M
~~
-
C
~
...
~
c:II
Z
0
III
C
~
.,N
ex)
• ~~
C
~
...
,
o
,..",
M
•o
II'l
M
o
ex)
i
o
-0
6
~
-
o
N
6
o
-
,
c..
-0
,
o
N
I
o
70
71
in the lower area as over a dense bed. Intermediate numbers of fish
were taken along the transition zone from high to low shrimp density.
As discussed earlier, the Battle Flats site had a dense shrimp bed
restricted primarily to the mid intertidal, as seen in most areas within
Coos Bay, whereas the causeway bed extended throughout the intertidal.
A total of 44 fish were caught in traps at Battle Flats on the first day
while only 7 fish were taken from the causeway during the same period.
On the following day, catches from Battle Flats and the causeway were 23
and 20 sculpins respectively. Although the total catches for the two
days indicate a significantly higher number of sculpins at Battle Flats
(G=1S.0S; 3 d.f.; p<.OOS), the minimal difference in catches on the
second day leads to uncertainty about whether a decrease in predator
abundances can explain the lower extension of a dense bed at the
causeway.
Discussion
Stomach analysis of Leptocottus armatus suggests that this fish is
an important predator on Callianassa in Coos Bay. Since most studies of
fish diets do not identify stomach contents to species, information on
the importance of staghorn sculpin predation to Callianassa in other
embayments is limited. Tasto (197S) reported 4.4% of staghorn sculpins
less than 70 mm standard length and 21.S% of those more than 70 mm had
Callianassa -in their diets. Jones (1962) observed Callianassa
longimanus in the diets of a small number of sculpins from San Francisco
Bay. Millicent Quammen (University of California, Santa Barbara;
written communication) found that 10-20% of the staghorn sculpins she
72
examined in Mugu Lagoon, California, had eaten ghost shrimp.
Within the vicinity of the shrimp beds, the abundance of adult
Leptocottus varied seasonally over the two years of sampling. In 1979,
the Oregon Department of Fish and Wildlife used beach seines to survey
fish distributions at the Charleston boat basin, approximately .5 km
north (downstream) of the Joe Ney site (R. Bender, unpublished data).
During January, only two adult sculpins were caught at this site and
only six adults were taken during a May 2 seine. However, samples taken
on May 29, June 15, and August 13 had 250, 200, and 197 adult sculpins
respectively. A somewhat lower total abundance of staghorn sculpins was
also observed during winter in Yaquina Bay (Bayer, 1981), but no
distinction was made between juveniles and adults. Horn (1980) noted
primarily small juveniles during winter and larger sculpins during
summer in Morro Bay, California. Bell (1978) found that Leptocottus
left an Alaska estuary during winter, perhaps in response to freezing
temperatures. Migration of larger staghorn sculpins out of Oregon
estuaries during winter may be related to the requirement for breeding
fish to remain in areas of higher salinity (Jones, 1962).
Although Leptocottus armatus was the only major predator on
Callianassa implicated in this study, several other predators may be
important in Coos Bay or other embayments. As mentioned, bottom feeding
rays and sharks may be expected to prey upon Callianassa in some areas
(Russo, 1975). Even though Dungeness crabs, Cancer magister, were not
seen feeding on Callianassa in sediment-filled tanks, these crabs
apparently do eat ghost shrimp under certain conditions (Stevens et al.,
1982). Juvenile Cancer have been observed to attack exposed ghost
73
shrimp at Valino Island and large numbers of these smaller crabs foraged
in the low intertidal of Coos Bay sandflats during spring (personal
observation). Additionally, several species of fish present in Coos
Bay, such as the starry flounder, Platichthys stellatus, were not
sampled with the beach seines but may prey on the ghost shrimp
subtidally.
74
CHAPTER V
EFFECT OF PREDATION IN LIMITING THE LOWER DISTRIBUTION
OF CALLIANASSA CALIFORNIENSIS BEDS
Introduction
Although the analysis of predator diets shows that infauna are
eaten, the effect of predation on infaunal distributions is difficult to
determine directly. The distribution of infauna must often be inferred
indirectly from burrow openings or tubes (Aller and Dodge, 1974; Ronan,
1975; Swinbanks and Murray, 1981; Suchanek, 1983) or observed through
destructive, and usually time consuming, core samples. Tests of
predation effects have usually involved field or laboratory
manipulations or monitoring of natural experiments, all three methods
posing inherent drawbacks (Virnstein, 1978; Peterson, 1979). Yet,
despite these problems, manipulations of predator densities, and
especially field exclusion studies, have provided important information
on the role of predation in soft-sediment communities (Virnstein, 1977;
Woodin, 1978; Peterson, 1979; Quammen, 1984).
Two field predator exclusion experiments and one natural experiment
examined the effect of predation in limiting the lower distribution of
C. californiensis in Coos Bay, Oregon. The predator exclusion studies
involve two distinct but related aspects of the zonation pattern: 1)
. does the ghost shrimp bed extend downwards when predator abundance is
reduced; and 2) do aggregations of adult Callianassa placed below a
75
dense bed survive significantly better with reduced predation. The
natural experiment involves monitoring a cohort of juvenile ~
californiensis that settled in the low-intertidal zone during early fall
1983.
Methods
Migration Studies
The relatively abrupt lower edge common to many Callianassa
californiensis beds allowed an experimental approach analogous to
Paine's (1974) study of sea star predation on mussels in the rocky
intertidal. Predator exclosures and controls were placed immediately
below several dense ghost shrimp beds at Valino Island and Long Island
Point, Coos Bay. The upper edge of these treatments bordered the lower
edge of the bed (Figure V-I), and change in number of shrimp burrow
holes within these treatments was assumed to be caused by migration of
adult shrimp from the adjacent dense bed. A direct test of this
assumption is discussed later. If pressure for downward expansion of
the bed exists and if predation is important in limiting this expansion,
then a higher density of shrimp (and shrimp holes) would be expected
within the exclosures relative to the other plots.
There were three control treatments: 1) a no-top cage with sides
but only a partial roof; 2) a sideless cage; and, 3) an unmanipulated
control plot lacking any cage structure. The no-top treatment was
designed to control for potential effects of current disruption by a
cage while the sideless treatment simulated some of the shading
76
Figure V-I. Exclosure cage and controls used in migration studies.
Inset shows placement of cage immediately below the lower
edge of a dense shrimp bed.
77
78
properties. The three cage structures were made of 12.5 mm mesh
galvanized hardware cloth placed over a light wooden frame. This mesh
size was small enough to exclude most predators capable of preying on
adult ghost shrimp (Chapter IV) yet large enough to allow the passage of
most Callianassa individuals. Behavior and survivorship of ghost shrimp
were not noticeably affected after several weeks of exposure to cage
materials in laboratory aquaria. Exclosure and no-top treatments
extended approximately 12.5 cm above the tideflat surface, with 5 cm
buried into the sediment. The sideless cage control was 21-24 cm high.
The increased height of this control was needed to give predators access
to the underlying substrate. All four treatments, full cage and three
controls, were 1 m on each side.
Replicate sets of these four treatments (Table V-I) were placed
seasonally at Long Island Point and Valino Island (Figure V-2).
Physical conditions at these sites varied from strong tidal currents,
direct wave action, and sand ripples near the channel to low currents
and fine sediments near the salt marsh. Experiments were begun during
the summer of 1982 and repeated each season through the summer of 1984,
although sideless controls were not added until spring 1983. Exact
location of a treatment within a site was chosen haphazardly with the
toss of a shell. Because of potential seasonal variations in ghost
shrimp activity and predator abundance and the possibility of
significant cage artifacts with increasing duration of caging
experiments, experimental sets were left in place for 40 to 60 days and
then removed. Monitoring during the course of the experiments suggested
that density responses, if present, became noticeable within 14-28 days.
79
TABLE V-I. Results of migration experiments. Values for a row represent
a single set of treatments. Different rows represent
different areas or years. Standard errors are calculated by
taking the square root of the ANOVA error mean square.
Change in Number of Holes
Season Exclosure Unmanipulated Sideless No-top
Fall 3 7 5
-16 -16 -23
1 -11 -3
-10 -12 -6
53 16 -34
13 47 7 20
-2 15 -1 8
-20 -12 -42
S.E.=16.86 F=I.59 p).2
Winter 19 0 2
29 19 6
26 12 13 26
10 -1 12 8
-22 1 21 -44
-8 37 20 -1
S.E.=15.14 F=I.41 p).3
Spring -3 -2 2 3
8 4 18 18
31 -10 18 8
6 1 -22
26 11 12 -6
72 4 -12 6
12 25 -12 20
-1 -7 -9 -11
S.E.=16.35 F=2.13 p).1
Summer 46 25
69 8 10
10 -14 -25
84 30 -1
29 -5 11 11
3 -13 0 32
21 2 11 10
n*(A) 21 21 -1
9*(B) 7 18 8
71*(C) 31 15 10
S.E.-17.83t F=5.57 p).OI
*Sets used for sediment grain analysis.
tFrom untransformed data.
80
Figure V-2. Location of migration (triangles) and transplant (squares)
experiments. Dashed lines show approximate level of mean
lower low water and dotted lines indicate position of shrimp
beds during the summer of 1984.
81
I
H
SM
VALINO ISLAND
.
.
.
.
.
•
...... - - .
...... ..- ..
..-..- .(C::..SH I .
,- ~ '.
....... ' .... ., .
.
..
I
J
I
/
I
(
I
I
~
\
\
\
\
\
\
\
"\. \
.. .J
'\
\
'\
'\
\
,
",
\.
,
,
,
·
.
I'
/
,/
, ,
.
.
•
·
-
--
.. ..
-
••••
. ..
•.
.
....
••••
-
.
.
•....
...
.. .
.. .
.. .. ..".
".
... .... . ..
--- -..... -
CHANNEL
)
,/
.. /
·, , , .
• / /' - - -- - -
) , // "
I
,
. / \ I
,
, 1/ \ LO'G
I 1'-" \ ISLCND,
•
PO NT
I , I
. I
.
'f ~. I
,
I
I
. /
• I,
82
Treatment cages were cleaned of drift algae at least biweekly.
The number of C. californiensis holes was recorded from each of the
replicates at the beginning and end of an experiment, counts for all sets
being made on the same day. Some counts were independently repeated by
students from the Oregon Institute of Marine Biology; the relative number
of holes among the four treatments recorded by these students was
consistent with that observed here.
Sediment characteristics within exclosure cages and in the
surrounding areas were determined from samples taken with a 50 cm long by
12.5 diameter corer. The core was halved lengthwise, and grain size and
sorting properties of one half were analyzed separately for the top and
bottom 25 cm, using dry seiving techniques (Ingram, 1971).
Settlement Studies
During fall of 1983, dense aggregations of ghost shrimp burrows were
noticed at several locations in the low intertidal of Coos Bay sandflats,
often equalling or exceeding the densities observed in well established
beds. Excavation of some of these areas indicated that they were
inhabited almost entirely by newly settled (11-13 rom length) C.
californiensis. At Long Island Point, a 25 m2 area containing one such
bed of newly settled shrimp, approximately at MLLW, was monitored several
times by counting the number of burrow holes in haphazardly placed 0.25
2
m quadrats.
Transplant Studies
During the summer of 1984, four pairs of transplant plots on Valino
83
Island were placed at approximately 0.6 m above MLLW (Figure V-2). The
lower edge of the main shrimp bed was about 10 m away and 0.9 m above
MLLW. Each pair of transplants included a 1 m2 full cage (exclosure) and
2
aIm uncaged treatment. Fifty shrimp were transplanted into each
treatment of three experimental pairs and 40 ghost shrimp were placed
into the two plots of the fourth. Callianassa were collected from a
sandflat near the Oregon Institute of Marine Biology and transplanted
into the experimental plots within two days of collection. Reburrowing
was aided by putting the shrimp into 8 cm deep, water-fil1ed starter
holes. Most individuals immediately began enlarging these initial
burrows. Since large shrimp had difficulty in excavating new burrows,
only individuals between 10 and 70 mm length were used.
Both exclosure and control treatments were covered with a cage for
3-5 days after transplanting to allow shrimp to begin constructing their
burrow systems. (In the laboratory the initial stages of burrow
construction, formation of a Y-shaped burrow with two openings and a
turn-around chamber, usually occurred within a few days.) After 3-5
days, the cage was removed from the control plot. Holes were counted at
this time and again after 35-50 days. Both treatments within an
experimental pair were counted on the same day. Partial cage structures
were not used to control for cage effects since the exclusion cages were
identical to those in the migration experiments, for which partial cages
were included, and the work involved in transplanting an additional 100
shrimp per set was prohibitive.
84
Statistical Analysis
In both migration and transplant studies, changes between the
initial and final number of holes were used as an indicator of treatment
effects on shrimp density. Treatments for the migration experiment were
compared separately for each season using a one-way analysis of variance
(ANOVA), blocking for differences between experimental sets (Sokal and
Rohlf, 1981). A square root transformation (~/lchange + .51) was
applied where an F-max test indicated heteroscedasticity (this occurred
only for the summer experiments; application of the transform resulted in
a slight decline in significance levels). Similar analysis also tested
treatment effects in the transplant studies.
Sediment grain size and sorting coefficients were calculated using
moment statistics as described in McBride (1971).
Results
Migration Studies
The effect of excluding predators on Callianassa californiensis
abundance immediately below a bed varied seasonally (Table V-I).
Experiments during the three summers demonstrated a highly significant
effect. No significant difference between treatments was observed during
the fall, winter and spring experiments. The results for spring,
however, appear intermediate to those of winter and summer.
Although the presence of unequal sample sizes prevented an analysis
of single degrees of freedom, much of the difference between treatments
during summers can be attributed to the great increase in the number of
85
holes within the exclosure cages. Their average was over three times
that in the controls, with exclosure cages exhibiting the greatest
increases in eight of the ten experimental sets (Table V-I). Excavation
of a limited number of treatment sets indicated that the great majority
of individuals within both exclosure and control plots were more than 1
year old (40 mm or longer), few individuals coming from that year's size
class.
An interesting result of these experiments was the variability in
response to a particular treatment between different experimental sets,
especially during fall and winter. This is demonstrated by the fall no-
top treatment, which varied from the lowest decline in hole numbers to an
anomalously large decline (Table V-I). There was also fluctuation within
a treatment over time. Fluctuations in hole number were especially
noticeable between weekly summer visits. Such variations, both between
areas and within a treatment, suggest that the edge of the bed is not
static but rather represents a dynamic front with contraction and
expansion.
Cage Artifacts
A major problem in the interpretation of caging experiments in soft
sediments is the potential for cage artifacts. Cage structures may
disrupt currents possibly leading to increased larval settlement
(Virnstein, 1978; Eckman, 1983; Hannan, 1984) or siltation (McCall, 1977;
Hulberg and Oliver, 1980), shade the substrate surface affecting growth
of microalgae (Peterson, 1979), become fouled increasing local food
concentrations for some species and decreasing food for others (Peterson,
86
1979), or provide refuges for secondary predators or juveniles of larger
predators (Arntz, 1977; Young et al., 1976; Lee II, 1978). Even straws
partially inserted into the sediment may lead to significant local
effects on species composition and abundance (Eckman, 1983).
The relatively short duration of my caging experiments should
minimize cage artifacts, as suggested by the lack of significant cage
effects during spring, winter, and fall periods. However, in order to
determine if the summer increases within the exclosure cages may be due
to physical effects associated with the presence of a cage, two types of
potential cage artifacts will be examined further: 1) that the hole
numbers may increase in response to changing hole-to-shrimp ratios rather
than to increases in shrimp number; and, 2) that shrimp density may
increase in response to physical changes associated with the cage
structure rather than to predator exclusion.
Because a cage structure may disrupt currents, holes may persist
longer within an exclosure treatment, leading to an increase in hole
numbers without a corresponding increase in shrimp density (the reverse
hypothesis, that increased turbulence may lead to decreased persistence
of holes, is also possible, but would not contribute to the significantly
higher number of holes seen in summer exclosures). The relationship
between shrimp numbers and shrimp holes was measured within summer
exclosure cages using the procedure described in Chapter III. There is a
strong correlation between holes and number of shrimp within these
treatments (Figure V-3; r=.93; p<.001); however, the slope of the least
squares regression line is significantly higher than that measured
outside the cages during the same time period (F=31.96; p<.0001). This
87
•
•
o
2
0\'u-~----L-.-,.6"""0---1~00---14-0~'-'-~1~8~0-~~2+-0~-+-­
HOLES PER .5 M2
14
12
10 0
~
"0 8.
.:
III
Go
Go 6 0 0
~
- 0
.:
%
eft
4 •
Figure V-3. Comparison of the relationship between hole density and
shrimp abundance within exclosure cages (open circles) and
outside of cages (filled circles) during the summer 1982. r
for cages = .92; r for outside = .88. p for both >.001.
88
suggests that holes provide a conservative estimate of shrimp density in
exclosures as compared to uncaged areas, possibly even underestimating
the density of ghost shrimp in caged plots. The decline of hole-to-
shrimp ratios within exclosures is not readily explained. The pattern is
similar to the difference between summer and winter measurements, which
may be attributable to reduced winter activity. A similar mechanism may
operate here. Because of the initially low numbers of Callianassa, the
exclosure areas may represent a richer food source (Bird, 1982) and more
difficult burrowing conditions (Brenchley, 1978) than a dense bed. Both
the greater accessibility of detrital food and increased effort needed to
burrow may contribute to reduced burrowing activity.
Whether the increase in shrimp numbers within exclosures is due to a
cage artifact or to treatment effects represents a second problem with
interpreting the results of these experiments. If cage artifacts are
important, cage controls that allow access to predators while mimicking
some of the physical disturbances associated with a complete cage should
show intermediate results. There is little difference between the two
cage controls and the unmanipulated plot during summer experiments.
Sediment grain size analysis from cores taken within exclosure and
unmanipulated plots also suggests that exclosure cages do not cause major
changes in the underlying substrate (Table V-2). Comparisons between
cores taken in cages and their unmanipulated controls within the same
experimental set showed similar mean particle sizes and sorting
coefficients between these treatments. Sediment characteristics were
similar even when treatments were several meters from each other and
hence subject to local variability in current patterns. Another
89
TABLE V-2. Mean and standard deviation (sorting) of particle size in
sediment samples taken in exclosure (E) and unmanipulated (0)
treatments. Letter identifications of experimental sets
refer to notations in Table V-I.
Experimental Depth Mean Grain Standard
Set Treatment Range (cm) Size (~) Deviation
A E 0-25 2.4 .783
0 0-25 2.4 .738
E 25-50 2.4 .798
0 25-50 2.3 .748
B E 0-25 2.4 .777
0 0-25 2.4 .838
E 25-50 2.4 .757
0 25-50 2.4 .829
C E 0-25
0 0-25 2.6 .759
E 25-50 2.3 .853
0 25-50 2.5 .756
potential artifact may arise from increased settlement around a cage. As
mentioned previously, excavation of exclosures showed that the majority
of the shrimp in these plots were adults. Callianassa settlement in Coos
Bay was highest during late August (Johnson and Gonor, 1979; personal
observation) while increases in hole number within summer exclosures
usually were evident by mid July. Also, in summer 1982 the cages
contained only 4.5% newly settled ghost shrimp by late August while an
adjacent bed at the same time contained 8.8% juveniles. The lack of
increases in number of holes within summer cage controls, no differences
in sediment grain size and sorting between caged and uncaged areas, and
90
no increase of shrimp settlement within complete cages all suggest that
cage artifact effects are small relative to treatment effects and that
the increases in number of holes are real and resulted from reduced
predation.
Transplant Studies
All four exclosure cages had significantly higher numbers of holes
relative to initial counts than their paired control plots (Table V-3;
F=40.09; p<.008). Since the exclosure cages for the transplant studies
were identical to those in the migration experiments, since the
transplant plots were not near any dense aggregations of ghost shrimp,
and because these experiments were conducted before most settlement
occurred, most of the differences between caged and uncaged plots are
explained by differential mortality or emigration. At the start of these
experiments the density of Callianassa holes immediately surrounding the
transplants was low, I-21m2; it did not increase significantly during the
course of the summer. The lack of an increase in hole numbers in the
area around the transplant plots and the negligible cage artifacts
observed among the cage controls of the migration experiments suggest
that decreased mortality (resulting from decreased predation) is the
major cause of the difference between caged and uncaged treatments.
Settlement Studies
Callianassa californiensis collected near the low-intertidal
settlement area in October were 11-13 mm in length, indicating that they
settled during the previous two to three months (Johnson and Gonor, 1979;
91
TABLE V-3. Change in the number of holes within transplant
plots. Values for a row represent a single
treatment pair. Different rows represent
different areas.
Exclosure
8
15
33
38
Unmanipu1ated
-4
-5
7
12
Bird, 1982). Collections of ghost shrimp from higher in the intertidal
indicated that Callianassa larval settlement had occurred throughout a
tidal range from at least 1.5 m above (within an established bed) to
-0.2 m below MLLW. However, hole density declined in the study area (at
approximately 0 m above MLLW) from October to July (Figure V-4).
Collecting shrimp from an adjacent area in October and again in June
also suggested a decline in Callianassa abundance over the nine month
. period. 2The single October observation of 67 ho1es/0.25 m was close to
the 57 ho1es/0.25 m2 observed in a dense bed at Va1ino Island during
2this same period, while the late July estimate of 8 holes/0.25 m
approximates the 5-6 ho1es/0.25 m2 at the same tidal height on Va1ino
Island in early August of the preceding year (before most settlement had
occurred).
Discussion
Several authors have suggested mechanisms to account for the
70
60
1
•
92
NI:
50
1ft 3 3
" j•1M 40A.
eft
1M
.. 300
z
20 3t 4
10 f
0
OCT- DEC FEB APR JUN AUG
1983 1984
FIGURE V-4. Hole densities within low intertidal settlement area. Bars
indicate + 1 standard deviation and numbers above bars give
the sample size.
93
intertidal zonation of Callianassa californiensis. However, these
explanations have seldom been tested and their appropriateness to the
Valino Island community is questionable.
Thompson and Pritchard (1969) suggested. that occupation of higher
tidal levels may provide Callianassa with protection "from lower
interstitial salinities at times of overflowing brackish water." In
river dominated systems, one might expect channel salinities to drop at
low tide and salinities to increase from the influx of ocean water on
flood tides. While low-tide surface water salinities at Valino Island
(Table V-4) may drop below the 8 0/00 salinity tolerance of C.
californiensis (Thompson and Pritchard, 1969) for short periods, such a
hypothesis does not appear adequate to explain the zonation pattern in
this area. First, interstitial salinities, measured from pooled water
using a refractometer, did not differ greatly between a dense bed and an
area 3 m away that had few ghost shrimp (Table V-4). Secondly, the
presence of ghost shrimp in the lower intertidal and subtidally,
although at lower densities, indicates that they can survive in these
areas. Also, zonation is evident near the mouths of many Oregon Bays
where tidal variations in salinity would be reduced. Finally a salinity
hypothesis cannot explain the downward migration of adult ghost shrimp
observed in exclosure cages.
McCrow (1971) suggested that the lower limit of C. californiensis
in Oregon may be determined through competition with the burrowing
shrimp Upogebia pugettensis. Although Upogebia beds are commonly found
along the lower edge of Callianassa beds in some Oregon estuaries, this
pattern is not prevalent in Coos Bay (personal observation). At both
94
TABLE V-4. Comparisons of low-tide salinities (0/00 ) from three
locations at Valino Island, Coos Bay.
Shrimp Burrows in Interstitial 3m
Date Channel a Dense Bed Below Bed
1983
19 February 4 18 17
4 March 11 22 22
19 March 12 20
4 April 6 14
4 May 16 20 20
21 July 17 30
31 July 27 31
28 November 7 16
13 December 9 16 16
1984
11 January 14 19 18
25 February 9 17 19
5 May 4 24 24
Long Island Point and Valino Island, dense beds of these two burrowing
shrimps are widely separated.
Although not proposed specifically to explain zonation patterns,
MacGinitie (1935) suggested a cyclic pattern for the establishment of
ghost shrimp beds that may lead to the occurrence of dense aggregations
at higher tidal levels. He suggested that as sandy bottoms are built up
through deposition, benthic algae, such as Enteromorpha spp., may occupy
95
the substrate. These algae would greatly increase the sedimentation
rate, causing the sandf1at to be rapidly built up above the 10w-
intertidal zone and the original growth of algae to be buried. As the
algae begin to decay, Callianassa invade the area with its abundant
detrital food source. The activities of the ghost shrimp further build
up the substrate surface and prevent the establishment of additional
algae. When the organic matter has been depleted, the ghost shrimp
leave, the remaining clean, sifted sand is eroded away, and the level of
the substrate is lowered so that the cycle may begin anew. Although
such a cycle may occur under certain circumstances, it does not appear
to be widespread. Bird (1982) saw no evidence for a cyclic pattern in a
northern Oregon shrimp bed and Peterson (1979) implied a relatively
stable position for ghost shrimp beds in Mugu Lagoon, California.
Except for seasonal variations in hole number (Chapter III), cycles were
not observed at Va1ino Island and comparisons with surveys conducted
during the early 1970s (P. Rudy, Oregon Institute of Marine Biology,
unpublished data) suggests that the position of the bed is similar to
that of a decade ago. Additionally, the beds in South Slough are above
the area of greatest algal growth (Pregna1l and Rudy, 1985).
Sediment grain size and sorting have been suggested as important
factors to the distribution of several deposit feeders, including
Callianassa major from the Atlantic coast of North America (Pearse et
a1., 1942; Poh1, 1946). Because bioturbation by C. californiensis may
change both mean grain size and sorting (Bird, 1982), direct comparisons
between Callianassa beds and adjacent low density areas are not very
useful. However, ghost shrimp beds in northern Oregon occur under a
96
wide variety of grain sizes (1.5-2.7 mean grain size) and sorting
coefficients (0.4-1.2 standard deviations) (Bird, 1982). The below-bed
controls used in the migration studies (Table V-3) fall well within this
range.
Devine (1966) found that Callianassa filholi had a predominately
intertidal distribution and suggested this was in response to the
intertidal distribution of one of its major food items, the diatom
Chaetoceras. Feeding ecology varies among species of Callianassa and it
is not likely that a similar explanation could explain the sharp
zonation of C. californiensis. Powell (1974) mentions that the diet of
C. californiensis included both planktonic and detrital components while
MacGinitie (1934) found mainly detrital material and bacteria. Stomach
content analysis of ghost shrimp collected from Valino Island and near
the Oregon Institute of Marine Biology in Charleston revealed
predominately detrital material, with a small percentage of diatoms.
Under laboratory conditions, ghost shrimp readily feed upon commercially
sold shrimp pellets (dried shrimp sold as fish food). A sharp zonation
pattern would be difficult to explain on the basis of the availability
of detrital food, especially without a significant difference in
sediment grain characteristics and the observation that dense shrimp
beds are often lower in organic content than adjacent low density areas.
The migration and transplant studies discussed here suggest that
during the summer months predation may play an important role in
limiting the distribution of Callianassa californiensis. The migration
studies further indicate that this effect is highly seasonal, there
being virtually no difference between exclosures and controls by the end
97
of fall or winter. As discussed in Chapter IV, abundances of the major
predator of C. californiensis, adult staghorn sculpins (Leptocottus
armatus), also peak in the early summer and decline in the fall.
However, although predation may limit the distribution of Callianassa
beds during late spring and summer, another mechanism must be involved
in the maintenance of zonation during the winter months.
The use of holes to estimate the density of the fall 1983 cohort of
ghost shrimp at Long Island Point requires that care be taken in the
interpretation of these results. Monitoring of hole densities within
established beds and determination of hole-to-shrimp ratios (Chapter
III) suggests a decline in the number of holes observed per shrimp
during the winter months. Within a shrimp bed, hole densities may
decline by 60% between September and February with a corresponding
increase, gradual at first, between April and June. Thus the 40%
decline in hole densities in the low-intertidal settlement area between
October and March may be explained by natural fluctuations in hole-to-
shrimp ratios. Actual reductions in shrimp numbers are probably
reflected by the decline of hole densities between March and late April
(censuses were taken during the last week of April) and between April
and late June. These periods are well past the time of harshest
salinity and wave conditions, but do coincide with the appearance of
several predators in the shallow waters of Coos Bay. Adult Leptocottus
began increasing in numbers in April and May. An unusually large
settlement of juvenile Cancer magister also occurred from late March
1983 to April 1983 (personal observation). These crabs attacked exposed
adult Callianassa under field conditions (although usually
98
unsuccessfully) and killed and consumed juvenile ghost shrimp in the
laboratory. During April low tides, juvenile Cancer were found buried
in the sand of the low-intertidal area, where Callianassa settled, at
densities between 120-200/m2 • A third group of predators, juvenile
Leptocottus, was present from mid-winter into early spring, but the few
fish examined did not have Callianassa in their stomachs. Although it
is difficult to determine which predators are most important, the
appearance of these predators at about the same time at which juvenile
shrimp densities began to decline in the low intertidal strongly
suggests that they control the number of juveniles surviving in this
zone.
99
CHAPTER VI
CONCLUSION
The studies discussed in the preceding chapters lead to a number of
conclusions:
1. Dense aggregations of Callianassa californiensis in most Oregon
embayments are limited primarily to the mid intertidal.
2. High densities of Callianassa are correlated with changes in
the abundance of other tidef1at organisms. In particular, sedentary
polychaetes and amphipods displayed lower abundances within shrimp beds.
These effects appear to depend upon the ghost shrimps' disruption of the
sediment.
3. During summer, the lower limit of dense shrimp beds within Coos
Bay, Oregon, appears to be determined largely by predation.
4. Predation has little effect in maintaining zonation during
winter. However, the lack of a strong downward expansion in this season
seems related to lower winter activity of the shrimp.
These results suggest a general model for the zonation of
Callianassa in Coos Bay. One of the major factors preventing the
establishment of dense beds in the low intertidal and sub tidally is
seasonal predation. Although predation is reduced during winter, there
is also less movement of adult shrimp, and hence little expansion of
existing beds. Aggregations of shrimp that do occupy the low intertidal
during the winter, such as periodic settlements of juvenile shrimp,
100
appear to be cropped during the following year. It should be emphasized
that this model would explain only the present lower limit to dense
beds. The genus Callianassa is generally restricted to shallow water
(Pickett et al., 1971) and predation is not a likely explanation for its
rarity in greater depths.
As mentioned above, the two most important mechanisms for the
expansion of dense ghost shrimp beds into the low intertidal and
subtidal involve the summer migration of adult Callianassa and larval
settlement. Larval settlement is sporadic (Bird, 1982; Chapter V) and
juvenile shrimp may be susceptible to several predator species.
However, only one major predator of adult ghost shrimp, the staghorn
sculpin Leptocottus armatus, was observed in Coos Bay. Since
information on diet and abundance is available for this species, it is
possible to assess whether predation by the staghorn sculpin is
sufficient to prevent downward expansion of the bed through migration of
adult shrimp. Since fish diets and abundances, as well as shrimp
migration, vary seasonally, the following discussion will be limited to
data from June and July.
Between 6% and 38% of the staghorn sculpins caught in beach seines
during June and July had Callianassa in their stomachs (Chapter IV). In
virtually all cases, only a single shrimp was eaten. An average summer
rate of 0.2 shrimp/sculpin is consistent with other estimates from diet
studies (Tasto, 1975; M. Quammen, University of California, Santa
Barbara, written communication) and will be assumed here. Since several
hours are needed to digest an item as large as a ghost shrimp, this
estimate will be taken as an approximation of predation rates over a
101
tidal cycle.
Determining the density of Leptocottus immediately below a shrimp
bed is more difficult. Information on staghorn sculpin abundances in
Coos Bay comes from beach seines and fish traps. Seines cover a
relatively large area, not sampling specifically in the mid-intertidal
zone beneath a shrimp bed, and they often have a low catch efficiency
owing to fish avoidance of the net. Avoidance or attraction may also
occur with fish traps, but traps have the advantage of sampling a
specific area. The total catch in the 24 traps placed immediately
beneath the Callianassa bed at Battle Flats was 110 sculpins (Chapter
IV). Although the exact area fished by a trap is uncertain, each
covered a section 1 m wide and an area of 1 m2 will be used as a
conservative estimate. All traps were left in place for a single high
tide. These figures suggest a maximum density of 4.6 sculpins/m2 per
tidal cycle. This estimate assumes that sculpins forage over a limited
area and that adjacent 1 m2 quadrats have similar densities. At the
other extreme, a small number of fish could forage along the entire
length of the shrimp bed and the sculpins caught in the traps may
represent the entire population foraging beneath the Battle Flats bed.
Since the Battle Flats bed was approximately 50 m long, this suggests a
lower estimate of 0.09 sculpins/m2 per tidal cycle. Using these
densities and the 0.2 Callianassa/m2 per tidal cycle rate discussed
previously, Leptocottus may consume between 0.02 and 0.92 ghost
shrimp/m2 per tidal cycle, or 0.04 to 1.84 shrimp/m2 per day, at Battle
Flats during June and July.
Comparing summer exclosures and unmanipulated controls in the
102
migration experiments provides a rough estimate of downward migration
for adult Callianassa (Chapter V). Using an estimate of 4 holes/shrimp
(Chapter III), there was an average summer difference of 8.2 shrimp/m2
between these treatments and the maximum difference between an exclosure
and its paired control was 15 shrimp/m2• Since these experiments were
continued for approximately 60 days, this gives an average migration
rate of 0.14 shrimp m-2day-1 and a maximum rate within cages of 0.25
-2 -1
shrimp m day • Because of partial predation on burrows near the edge
of a cage, these values may underestimate the actual downward migration.
Summer Callianassa densities at both the Valino Island and Battle Flats
beds were approximately 70 shrimp/m2 (Chapter III). The final density
in the 1 m wide band beneath the shrimp bed would not be expected to
exceed this value, indicating a maximum immigration rate of 1.17 shrimp
-2 -1
m day during the 60 day period.
The figures given above represent only rough estimates of predation
rates along the lower edge of a shrimp bed. However, estimates of
-2 -1predation rates between 0.04 and 1.84 shrimp m day and migration
-2 -1
rates ranging from 0.14 to 1.17 shrimp m day suggest that Leptocottus
predation could be sufficient to prevent downward expansion of the bed
resulting from adult Callianassa movement.
Although zonation of shrimp beds is common in Coos Bay, the
sharpness of this lower boundary varies between areas and over time.
The lower edge of many beds was less defined during winter when
predation decreased (Chapter III). Observations of beds then suggest
that the lower boundary is more distinct on steeper slopes. Another
factor that may contribute to sharp boundaries around moderately dense
103
beds is intraspecific facilitation of burrowing (Brenchley, 1981). As
noted earlier, dense aggregations of Callianassa disrupt the sediment,
often giving it a soft, quicksand quality. Burrowing in such an area
would require less energy than in packed sands. However, at higher
densities, the negative intraspecific effects of food depletion and
aggressive interactions may counteract burrowing facilitation.
As discussed in Chapter I, much of the previous work on intertidal
communities in soft sediments has indicated that sediment grain size and
current characteristics are the major factors that affect zonation of
most deep-burrowing organisms. Competition is not generally important
in causing zonation of soft-sediment fauna (Peterson, 1979), and
predation appears to affect primarily epifauna and shallow infauna
(Birkeland and Chia, 1971; Reise, 1978; Van Dolah, 1978; Race, 1982).
Callianassa californiensis often forms burrow systems more than 50 cm
deep (MacGinitie, 1934) and the lower distribution of shrimp beds
appears to be determined to a large e~tent by summer predation. Adult
C. californiensis apparently become susceptible to predation because
they spend time near the burrow entrances and may leave the burrow on
some occasions. These results emphasize the need to understand
behavioral patterns of individual species in determining the relative
importance of physical and biological factors to their distribution.
Predation on Callianassa californiensis in Coos Bay may also affect
the abundance of other tideflat organisms in a manner analogous to the
effect of sea star predation on mussels in the rocky intertidal (Paine,
1974). Due to the shrimps' strong bioturbating activity, the presence
of a dense Callianassa bed is correlated with decreases in the
104
abundances of certain sedentary organisms, such as spionid po1ychaetes
and tubico10us amphipods, and increased numbers of some mobile forms
(free burrowing amphipods in this study). If dense Callianassa beds
existed throughout the intertidal, the relative abundance of many
species would be quite different from that observed. The zonation of C.
californiensis beds indicates that predation may be important in
determining the relative abundances of intertidal organisms in soft
sediments as well as on hard substrates.
Although Callianassa californiensis may represent an extreme
example, other organisms have been reported to influence the relative
abundance of soft-sediment fauna. Dense aggregations of the burrowing
polychaete Abarenicola pacifica are associated with lower numbers of
some tube-dwelling animals (Wilson, 1981) and Arenicola marina appears
to have a significant effect on meiofaunal composition (Reise, 1983).
Changes in the abundance of soft-sediment organisms have also been
reported from dense assemblages of tube-building po1ychaetes (Wilson,
1979). If predation limits the distribution of these dominant animals
under certain conditions, it could affect the composition of the
intertidal community similarly to the zonation of C. californiensis
beds.
105
APPENDIX
OCCURRENCE OF ANIMALS TAKEN IN BENTHIC CORES
A complete listing of animals taken in benthic cores (See Chapter
II) is given below. Per cent abundance is calculated based on the total
number caught during the two years of sampling. Less than 0.1% is shown
by an asterisk (*). The data come from 8 sampling periods (seasonally
from spring 1982 to fall 1984; see Chapter II)
Taxa
Per Cent
Abundance
II of Sampling
Periods Found
Crustacea
Tanaidacea
Leptochelia dubia
Amphipoda
Eobrolgus spinosus
Corophium spp.
Grandidierella japonica
Ampithoe valida
Eohaustorius sp.
Ostracoda
Isopoda
Cumacea
63.5 8
6.6 8
2.8 8
1.2 8
0.3 2
* 4
7.1 8
* 3
2.1 8
Decapoda
Callianassa californiensis
(juveniles)
Hemigrapsis sp.
Insecta
Arachnida (spiders)
*
*
*
*
3
2
7
8
Taxa
Oligochaeta
Polychaeta
Spionidae
Pygospio elegans
Pseudopolydora kempi
Polydora ligni
Boccardia sp.
Streblospio benedicti
Spio filicornis
Spiophanes bombyx
Rhynchospio arenicola
Spionid sp.
Capi te llidae
Mediomastus californiensis
Capitella sp.
Glyceridae
Hemipodus californiensis
Paraonidae
Paraonella platybranchia
Phyllodocidae
Eteone californica
Eteone dilatae
Eteone sp.
Opheliidae
Armandia brevis
Goniadidae
Glycinde armigera
Glycinde polygnatha
Glycinde sp.
Orbiniidae
Haploscoloplos elongatus
Syllidae sp.
Arenicolidae
Abarenicola pacifica
Per Cent
Abundance
11.1
1.5
1.5
'*
'*
.8
'*
'*
'*
'*
0.3
0.3
'*
'*
0.5
'*
'*
'*
'*
'*
'*
'*
'*
'*
/I of Sampling
Periods Found
8
8
8
5
1
8
2
7
1
1
8
8
1
6
7
1
1
3
7
1
1
3
1
1
106
IPer Cent II of Sampling
Taxa Abundance Periods Found
Oweniidae
Owenia fusiformis
*
1
Bivalvia
Macoma nasuta * 8
Macoma balthica
*
5
Cryptomya californica
*
7
Transenella sp.
*
3
Mytilus sp. (Juvenile)
*
1
Saxidomus sp.
*
1
Protothaca sp.
*
1
Bivalve sp. * 8
Gastropoda sp. * 2
Nemertea spp. * 8
107
108
BIBLIOGRAPHY
Allen, E. J. (1899). On the fauna and bottom-deposits near the thirty
fathom line from the Eddystone Grounds to Start Point. J. Mar. Biol.
Ass. U. K. 5: 365-540
Aller, R. C., Dodge, R. E. (1974). Animal-sediment relations in a
tropical lagoon, Discovery Bay, Jamaica. J. Mar. Res. 32: 209-232
Arntz, W. E. (1977). Results and problems of an "unsuccessful" benthos
cage predation experiment (western Baltic). In: Keegan, B. F.,
Ceidegh, P.O., Boaden, P. J. S. (eds.) Biology of benthic organisms.
Pergamon Press, New York, p. 31-44
Barnard, J. L., Barnard, C. M. (1981). The amphipod genera Eobrolgus and
Eyakia (Crustacea: Phoxocephalidae) in the Pacific Ocean. Proc. Biol.
Soc. Wash. 94: 295-313
Barnes, R. D. (1980). Invertebrate zoology. Saunders College,
Philadelphia
Bayer, R. D. (1981). Shallow-water intertidal ichthyofauna of the
Yaquina Estuary, Oregon. Northwest Science 55: 182-193
Bell, M. A. (1978). Fishes of the Santa Clara river system, southern
California. Nat. Rist. Mus. Los. Ang. Cty. Contrib. Sci. 295: 1-20
Biffar, T. A. (1971). The genus Callianassa (Crustacea, Decapoda,
Thallassinidea) in south Florida, with keys to the western Atlantic
species. Bull. Mar. Sci. 21: 637-715.
Bird, E. W. (1982). Population dynamics of the Thalassinidean shrimps
and their community effects through sediment modification. Ph.D.
Dissertation. University of Maryland, College Park
Birkeland, C., Chia, F.-S. (1971). Recruitment risk, growth, age, and
predation in two populations of sand dollars, Dendraster excentricus
(Eschscholtz). J. Exp. Mar. Biol. Ecol. 6: 265-278
Bloom, S. A., Simon, J. L., Hunter, V. D. (1972). Animal sediment
relations and community analysis of a Florida estuary. Mar. Biol. 13:
43-56
Boesch, D. F. (1972). Species diversity of marine macrobenthos in the
Virginia area. Ches. Sci. 13: 206-211
109
Boesch, D. F. (1973). Classification and community structure of
macrobenthos in the Hampton Roads area, Virginia. Mar. BioI. 21: 226-
244
Brenchley, G. A. (1978). On the regulation of marine infaunal
assemblages at the morphological level: a study of the interactions
between sediment stabilizers, destabilizers, and their sedimentary
environment. Ph.D. Dissertation. The Johns Hopkins University,
Baltimore
Brenchley, G. A. (1981). Disturbance and community structure: an
experimental study of bioturbation in marine soft-bottom
environments. J. Mar. Res. 39: 767-790
Brenchley, G. A. (1982a). Mechanisms of spatial competition in marine
soft-bottom communities. J. Exp. Mar. BioI. Ecol. 60: 17-33
Brenchley, G. A. (1982b). Predation on encapsulated larvae by adults:
effects of introduced species on the gastropod Ilyanassa obsoleta.
Mar. Ecol. Prog. Sere 9: 255-262
Bybee, J. R. (1969). Effects of hydraulic pumping operations on the
fauna of Tijuana Slough. Calif. Fish and Game 55: 213-220
Carson, R. (1955). The edge of the sea. Houghton Mifflin, Boston
Chester, A. J., Ferguson, R. L., Thayer, G. W. (1983). Environmental
gradients and benthic macroinvertebrate distributions in a shallow
North Carolina estuary. Bull. Mar. Sci. 33: 282-295
Coe, W. (1953). Resurgent populations of littoral marine invertebrates
and their dependence on ocean currents and tidal currents. Ecology
34: 225-229
Connell, J. H. (1961). The influence of interspecific competition and
other factors on the distribution of the barnacle Chthamalus
stellatus. Ecology 42: 410-723
Darnell, R. M. (1958). Food habits of fishes and larger invertebrates of
Lake Pontchartrain, Louisiana, an estuarine community. Publ. Mar.
Sci. Univ. Texas 5: 353-416
Dauer, D. M. (1984). High resilience to disturbance of an estuarine
polychaete community. Bull. Mar. Sci. 34: 170-174
Dayton, P. K. (1971). Competition, disturbance and community
organization: the provision and subsequent utilization of space in a
rocky intertidal community. Ecol. Monogr. 41: 351-389
110
Dayton, P. K. (1984). Processes structuring some marine communities: are
they general? In: Strong, D. R. Jr., Simberloff, D., Abele, L. G.,
Thistle, A. B. (eds.) Ecological communities: conceptual issues and
the evidence. Princeton University Press, Princeton, New Jersey, p.
181-200
Devine, C. E. (1966). Ecology of Callianassa filholi Milne-Edwards 1878
(Crustacea, Thalassinidea). Trans~ Roy. Soc. New Zeal. Zoology 8: 93-
110
DeWindt, J. T. (1974). Callianassid burrows as indicators of subsurface
beach trend, Mississippi River delta plain. J. Sedim. Petrol. 44:
1136-1139
Dorsey, J. H., Synnot, R. N. (1980). Marine soft-sediment benthic
community offshore from the Black Rock sewage outfall, Connewaire,
Victoria, Australia. Aust. J. Mar. Freshw. Res. 31: 155-162
Driscoll, E. G., Brandon, D. E. (1973). Mollusc-sediment relationships
in northwest Buzzards Bay, Mass. Malacologia 12: 13-46
Eckman, J. E. (1983). Hydrodynamic processes affecting benthic
recruitment. Limnol. Oceanogr. 28: 241-257
Fauchald, K. (1976). The polychaete worms: definitions and keys to the
orders, families and genera. Nat. Hist. Mus. Los. Ang. Cty. Sci. Sere
28
Fauchald, K., Jumars, P. A. (1979). The diet of worms: polychaete
feeding guilds. Oceanogr. Mar. BioI. Ann. Rev. 17: 193-284
Featherstone, R. P., Risk, M. J. (1977). Effect of tube-building
polychaetes on intertidal sediments of the Minos Basin, Bay of Fundy.
J. Sedim. Petrol. 27: 446-450
Fenchel, T. (1975). Factors determining the distribution patterns of mud
snails (Hydrobiidae). Oecologia 20: 1-7
Fenchel, T. (1977). Competition, coexistence and character displacement
in mud snails (Hydrobiidae). In: Coull, B. C. (ed.) Ecology of marine
benthos. University of South Carolina Press, Columbia, South Carolina
Frankenberg, D., Coles, S. L., Johannes, R. E. (1967). The potential
trophic significance of Callianassa major fecal pellets. Limnol.
Oceanogr. 12: 113-120
Gerrodette, J., Fleshig, A. O. (1979). Sediment-induced reduction in the
pumping rate of the tropical sponge Verongia lacunosa. Mar. BioI. 55:
103-110
III
Gonor, J. J., Kemp, P. F. (1978). Procedures for quantitative ecological
assessments in intertidal environments. EPA-600/3-78-087
Grant, J. (1981). Sediment transport and disturbance in an intertidal
sandflat: infaunal distribution and recolonization. Mar. Ecol. Prog.
Sere 6: 249-255
Grant, W. D., Boyer, L. F., Sanford, L. P. (1982). The effects of
bioturbation on the initiation of motion of intertidal sands. J. Mar.
Res. 40: 659-677
Grassle, J. F., Grassle, J. P. (1974). Opportunistic life histories and
genetic systems in marine benthic polychaetes. J. Mar. Res. 32: 253-
284
Green, R. H., Hobson, K. D. (1970). Spatial and temporal variation in a
temperate intertidal community, with special emphasis on Gemma gemma
(Pelecypoda: Mollusca). Ecology 51: 999-1011
Hailstone, T. S., Stephenson, W. (1961). The biology of Callianassa
(Trypaea) australiensis Dana 1852 (Crustacea Thalassinidea). Univ.
Queensl. Pap., Dep. Zool. 1: 259-285
Hannan, C. A. (1984). Initial settlement of marine invertebrate larvae:
the role of passive sinking in a near-bottom turbulent flow
environment. Ph.D. Dissertation. Woods Hole Oceanographic Institution
and Massachusetts Institute of Technology, Woods Hole, Massachusetts
Hartman, O. (1968a). Atlas of the errantiate polychaetous annelids from
California. University of Southern California Press, Los Angeles
Hartman, O. (1968b). Atlas of the sedentariate polychaetous annelids
from California. University of Southern California Press, Los Angeles
Heck, K. L. Jr., Thoman, T. A. (1981). Experiments on predator-prey
interactions in vegetated aquatic habitats. J. Exp. Mar. BioI. Ecol.
53: 125-134
Hedgepeth, J. W. (ed.) (1952). Between Pacific tides. Stanford
University Press, Stanford, California
Highsmith, R. C. (1982). Induced settlement and metamorphosis of sand
dollar (Dendraster excentricus) larvae in predator-free sites: adult
sand dollar beds. Ecology 63: 329-337
Hoffman, C. J. (1981). Associations between the arrow goby Clevelandia
ios (Jordan and Gilbert) and the ghost shrimp Callianassa
californiensis Dana in natural and artificial burrows. Pacific
Science 35: 211-216
I&
112
Holland, A. F., Dean, J. M. (1977). The biology of the stout razor clam
Tagelus plebeius: I. Animal-sediment relationships, feeding
mechanisms, and community biology. Ches. Sci. 18: 58-66
Horn, M. H. (1980). Diel and seasonal variations in abundance and
diversity of shallow water fish populations in Morro Bay, California,
USA. U.S. Natl. Mar. Fish. Servo Fish. Bull. 78: 759-770
Hulberg, L. W., Oliver, J. S. (1980). Caging manipulations in marine
soft-bottom communities: importance of animal interactions or
sedimentary habitat modifications. Can. J. Fish. Aquat. Sci. 37:
1130-1139
Ingram, R. L. (1971). Seive analysis. In: Carver, R. E. (ed.) Procedures
in sedimentary petrology. Wiley-Interscience, New York, p. 49-68
Johnson, G. E., Gonor, J. J. (1982). The tidal exchange of Callianassa
californiensis (Crustacea, Decapoda) larvae between the ocean and the
Salmon River estuary, Oregon. Estuar. Coast. Shelf Sci. 14: 501-516
Johnson, R. G. (1970). Variations in diversity within benthic marine
communities. Am. Nat. 104: 285-300
Jones, A. C. (1962). The biology of the euryhaline fish Leptocottus
armatus Girard (Cottidae). Univ. Calif. Publ. Zool. 67: 321-367
Jones, N. S. (1950). Marine bottom communities. BioI. Rev. 25: 283-313
Kneib, R. J., Stiven, A. E. (1982). Benthic invertebrate responses to
size and density manipulations of the common mummichog, Fundulus
heteroclitus, in an intertidal salt marsh. Ecology 63: 1518-1532
Krebs, C. J. (1978). Ecology: the experimental analysis of distribution.
Harper and Row, New York
Lawrence, J. M., Murdoch, J. (1977). The effect of particle-size
frequency distribution of the substratum on the burrowing ability of
Chiridota rigida (Semper.) (Echinodermata: Holothuroidea). Mar.
Behav. Physiol. 4: 305-311
Lee II, H. (1978). Seasonality, predation and opportunism in high
diversity soft-bottom communities in the Gulf of Panama. Ph.D.
Dissertation. Chapel Hill, North Carolina
Levin, L. A. (1981). Dispersion, feeding behavior and competition in two
spionid polychaetes. J. Mar. Res. 39: 99-117
Levin, L. A. (1982). Interference interactions among tube-dwelling
polychaetes in a dense infaunal assemblage. J. Exp. Mar. BioI. Ecol.
65: 107-119
113
Levinton, J. S. (1972). Stability and trophic structure in deposit-
feeding and suspension-feeding communities. Am. Nat. 106: 472-486
Levinton, J. S. (1977). Ecology of shallow water deposit-feeding
communities Quisset Harbor, Massachusetts. In: Coull, B. C. (ed.)
Ecology of the marine benthos. University of South Carolina Press,
Columbia, South Carolina
Longbottom, M. R. (1970). The distribution of Arenicola marina (L.) with
particular reference to the effects of particle size and organic
matter of the sediments. J. Exp. Mar. BioI. Ecol. 5: 138-157
Lubchenco, J., Menge, B. A. (1978). Community development and
persistence in a low rocky intertidal zone. Ecol. Monogr. 48: 67-94
MacGinitie, G. E. (1934). The natural history of Callianassa
californiensis Dana. Amer. MidI. Natur. 15: 155-177
MacGinitie, G. E. (1935). Ecological aspects of a California marine
estuary. Amer. MidI. Natur. 16: 630-763
McBride, E. F. (1971). Mathematical treatment of size distribution data.
In: Carver, R. E. (ed.) Procedures in sedimentary petrology. Wiley-
Interscience, New York, p. 109-128
McCall, P. L. (1977). Community patterns and adaptive strategies of the
infaunal benthos of Long Island Sound. J. Mar. Res. 35: 221-226
McCrow, L. T. (1972). The ghost shrimp, Callianassa californiensis, in
Yaquina Bay, Oregon. M. S. Thesis, Oregon State University,
Corvallis, Oregon
Meadows, P. S., Campbell, J. I. (1972). Habttat selection by aquatic
invertebrates. Adv. Mar. BioI. 10: 271-382
Mendoza, J. A. (1982). Some aspects of the autecology of Leptochelia
dubia (Kroyer, 1842) (Tanaidacea). Crustaceana 43: 225-240
Miller, D. C. (1984). Mechanical post-capture particle selection by
suspension- and deposit-feeding Corophium. J. Exp. Mor. BioI. Ecol.
82: 59-76
Moreno, E. C., Curros, P. V. (1979). The paleontology of the Cretaceous
of Los Condemios, Guadalajara Province, Spaine. Bol. R. Soc. Esp.
His. Nart. Secc. Geol. 77: 67-90
Morgan, E. (1970). The effect of environmental factors on the
distribution of the amphipod Pectenogammarus planicurus with
particular reference to grain size. J. Mar. BioI. Ass. U.K. 50: 769-
785
•114
Mountford, N. K., Holland, A. F., Mihursky, J. A. (1977). Identification
and description of macrobenthic communities in the Calvert Cliffs
region of Chesapeake Bay. Ches. Sci. 18: 360-369
Murphy, R. C. (1983). The introduced bivalve, Mercenaria mercenaria, in
a shallow coastal ecosystem: 1) factors affecting its distribution;
2) contribution to benthic community metabolism. Ph.D. Dissertation.
University of Southern California, Los Angeles
Myers, A. C. (1977). Sediment processing in a marine subtidal sandy
bottom community: II. Biological consequences. J. Mar. Res. 35: 633-
647
Naqui, S. M. Z. (1968). Effect of predation on infaunal invertebrates of
Alligator Harbor, Florida. Gulf Res. Rep. 2: 313-321
Nelson, W. G. (1981). Experimental studies of decapod and fish predation
on seagrass macrobenthos. Mar. Ecol. Prog. Sere 5: 141-149
Nichols, F. H. (1970). Benthic polychaete assemblages and their
relationship to the sediment in Port Madison, Washington. Mar. BioI.
6: 48-57
Nowell, A. R. M., Jumars, P. A., Eckman, J. E. (1981). Effects of
biological activity on the entrainment on marine sediments. Mar.
Geol. 42: 133-154
Oliver, J. S., Slattery, P. N., Hulberg, L. W., Nybakken, J. W. (1980).
Relationships between wave disturbance and zonation of benthic
invertebrates along a subtidal high-energy beach in Monterey Bay,
California. Fish. Bull. 78: 437-454
Ott, J. A., Fuchs, B., Fuchs, R., Malasek, A. (1976). Observations on
the biology of Callianassa stebbingi Borrodaile and Upogebia
littoralis Risso and their effect upon sediment. Senckenberg. Mar. 8:
61-79
Paine, R. T. (1966). Food web complexity and species diversity. Am. Nat •
100: 65-75
Paine, R. T. (1974). Intertidal community structure. Experimental
studies'on the relationship between a dominant competitor and its
principal predator. Oecologia 15: 93-120
Paine, R. T. (1984). Ecological determinism in the competition for
space. Ecology 65: 1339-1348
Pearse, A. S., Humm, H. J., Wharton, G. W. (1942). Ecology of sand
beaches at Beaufort, N. C. Ecol. Monogr. 12: 135-190
115
Peterson, C. G. J. (1918). The sea bottom and its production of
fishfood. Rep. Danish BioI. Sta. 25: 1-62
Peterson, C. H. (1979). Predation, competitive exclusion and diversity
in the soft-sediment benthic communities of estuaries and lagoons.
In: Livingston, R. J. (ed.) Ecological processes in coastal and
marine systems. Plenum Press, New York
Peterson, C. H. (1982). The importance of predation and intra- and
interspecific competition in the population biology of two infaunal
suspension-feeding bivalves, Protothaca staminea and Chione
undatella. Ecol. Monogr. 52: 437-478
Peterson, C. H., Andre, S. V. (1980). An experimental analysis of
intraspecific competition among marine filter feeders in a soft-
sediment environment. Ecology 61: 129-139
Peterson, C. H., Quammen, M. L. (1982). Siphon snipping: its importance
to small fishes and its impact on growth of the bivalve Protothaca
staminea (Conrad). J. Exp. Mar. BioI. Ecol. 63: 249-268
Pfitzenmeyer, H. T., Drobeck, K. G. (1967). Some factors influencing
reburrowing activity of the soft-shell clam, Mya arenaria. Ches. Sci.
8: 193-199
Pickett, T. E., Kraft, J. C., Smith, K. (1971). Cretaceous burrows--
<Chesapeake and Delaware Canal, Delaware. Paleontology 45: 209-211
Pohl, M. E. (1946). Ecological observations on Callianassa major Say at
Beaufort, North Carolina. Ecology 27: 71-80
Powell, R. R. (1974). The functional morphology of the fore-gut of the
thalassinid crustaceans, Callianassa californiensis and Upogebia
pugettensis. Univ. Calif. Publ. Zool. 102: 1-41
Pratt, D. M. (1953). Abundance and growth of Venus mercenaria and
Callocardia morrhuana in relation to the character of bottom
sediments. J. Mar. Res. 12: 60-74
Pregnall, A. M., Rudy, P. P. (1985). The contributions of green
macroalgal mats (Enteromorpha spp.) to seasonal production in an
estuary. Mar. Ecol. Prog. Sere In Press
Quammen, M. L. (1984). Predation by shorebirds, fish, and crabs on
invertebrates in intertidal mudflats: an experimental test. Ecology
65: 529-537
Race, M. S. (1982). Competitive displacement and predation between
introduced and native mud snails. Oecologia 54: 337-347
116
Ray, A. A. (ed.) (1982). SAS user's guide, statistics, 1982 edition. SAS
Institute, Cary, North Carolina
Reise, K. (1978). Experiments on epibenthic predation in the Wadden Sea.
Helgolander wiss. Meeresunters 31: 55-101
Reise, K. (1983). Experimental removal of lugworms from marine sands
affects small zoobenthos. Mar. BioI. 74: 327-332
Rhoads, D. C., Young, D. K. (1970). The influences of deposit-feeding
organisms on sediment stability and community trophic structure. J.
Mar. Res. 28: 150-178
Ronan, T. E. (1975). Structural and paleo-ecological aspects of a modern
marine soft-sediment community: an experimental field study. Ph.D.
Dissertation. University of California, Davis, California
Rudy, P., Rudy, L. H. (1983). Oregon estuarine invertebrates: an
illustrated guide to the common and important invertebrate animals.
Fish and Wildlife Service. FWS/OB5-83/16
Russo, R. A. (1975). Observations on the food habits of leopard sharks
(Triokis semifasciata) and brown smoothhounds (Mustelus henlei).
Calif. Fish Game 61: 95-103
Sanders, H. L. (1958). Benthic studies in Buzzards Bay. I. Animal-
sediment relationships. Limnol. Oceanogr. 3: 245-258
Sanders, H. L., Goudsmit, E. M., Mills, E. L., Hampson, G. E. (1962). A
study of the intertidal fauna of Barnstable Harbor, Mass. Limnol.
Oceanogr. 7: 63-79
Shelford, V. E., Weese, A. 0., Rice, L. A., Rasmussen, D. I., Maclean,
A. (1935). Some marine biotic communities of the Pacific Coast of
North America. Part I. General survey of the communities. Ecol.
Monogr. 5: 251-332
Smith, R. I., Carlton, J. T. (eds.) (1975). Light's manual: intertidal
invertebrates of the central California coast. University of
California Press, Berkeley, California
Snedecor, G. W., Cochran, W. G. (1980). Statistical methods. The Iowa
State University Press, Ames, Iowa
Sokal, R. R., Rohlf, E. F. (1981)~ Biometry. W. H. Freeman, San
Francisco
Stevens, B. A. (1928). Callianassidae from the west coast of North
America. Publ. Puget Sound BioI. Sta. 6: 315-369
117
Stevens, B. G., Armstrong, D. A., Cusimano, R. (1982). Feeding habits of
the Dungeness crab Cancer magister as determined by the index of
relative importance. Mar. BioI. 72: 135-145
Stone, R., Palmer, R. (1975). Effects of turbidity on the bay scallop,
Argopecten iriradiaus. Amer. Zool. 15:816 (abstract)
Suchanek, T. H. (1979). The Mytilus californianus community: studies on
the composition, structure, organization, and dynamics of a mussel
bed. Ph.D. Dissertation. University of Washington, Seattle,
Washington
Suchanek, T. H. (1983). Control of seagrass communities and sediment
distribution by Callianassa (Crustacea, Thalassinidea) bioturbation.
J. Mar. Res. 41: 281-298
Swinbanks, D. D. (1982). Intertidal exposure zones: a way to subdivide
the shore. J. Exp. Mar. BioI. Ecol. 62: 69-86
Swinbanks, D. D., Murray, J. W. (1981). Biosedimentological zonation of
Boundary Bay tidal flats, Fraser River delta, British Columbia.
Sedimentology 28: 201-237
Taghon, G. L., Nowell, A. R. M., Jumars, P. A. (1980). Induction of
suspension feeding in spionid polychaetes by high particulate fluxes.
Science 31: 562-564
Tasto, R. N. (1975). Aspects of the biology of Pacific staghorn sculpin,
Leptocottus armatus Girard, in Anaheim Bay. In: Lane, E. D., Hill, C.
W. (eds.) The marine resources of Anaheim Bay. Calif. Dept. Fish
Game, Fish. Bull. no. 165, p. 123-135
Thompson, L. D., Pritchard, A. W. (1969). Osmoregulatory capacities of
Callianassa and Upogebia (Crustacea: Thalassinidea). Biol. Bull. 136:
114-129
Thompson, R. K. (1972). Functional morphology of the hindgut of Upogebia
pugettensis (Crustacea, Thalassinidea) and its role in burrow
construction. Ph. D. Dissertation. University of California,
Berkeley, California
Torres, J. J., Gluck, D. L., Childress, J. J. (1977). Activity and
physiological significance of the pleopods in the respiration of
Callianassa californiensis (Dana) (Crustacea: Thalassinidea). BioI.
Bull. 152: 134-146
Van Dolah, R. F. (1978). Factors regulating the distribution and
population dynamics of the amphipod Gammarus palustris in an
intertidal salt marsh community. Ecol. Monogr. 48: 191-217
118
Vassalo, M. T. (1971). The ecology of Macoma inconspicua (Broderip and
Sowerby, 1829) in central San Francisco Bay. Part I. The vertical
distribution of the Macoma community. The Veliger 11: 223-234
Virnstein, R. W. (1977). The importance of predation by crabs and fishes
on benthic infauna in Chesapeake Bay. Ecology 58: 1199-1217
Virnstein, R. W. (1978). Predator caging experiments in soft sediments:
caution advised. In: Wiley, M. L. (ed.) Estuarine interactions.
Academic Press, New York
Weiser, W. (1956). Factors influencing the choice of substratum in
Cumella vulgaris Hart (Crustacea, Cumacea). Limnol. Oceanogr. 1: 274-
285
Weiser, W. (1959). The effect of grain size on the distribution of small
invertebrate$ inhabiting the beaches of Puget Sound. Limnol.
Oceanogr. 4: 181-194
Whitlach, R. B. (1977). Seasonal changes in the community structure of
the macrobenthos inhabiting the intertidal sand and mud flats of
Barnstable Harbor, Massachusetts. BioI. Bull. 152: 275-294
Whitlach, R. B. (1981). Animal-sediment relationships in intertidal
marine benthic habitats: some determinants of deposit-feeding species
diversity. J. Exp. Mar. BioI. Ecol. 53: 31-45
Willason, S. W. (1981). Factors influencing the distribution and
coexistence of Pachygrapsus crassipes and Hemigrapsus oregonensis
(Decapoda: Grapsidae) in a California salt marsh. Mar. BioI. 64: 125-
134
Wilson, W. H. Jr. (1979). Community structure and species diversity of
the sedimentary reefs constructed by Petaloproctus socialis
(Polychaeta: Maldanidae). J. Mar. Res. 37: 623-641
Wilson, W. H. Jr. (1981). Sediment-mediated interactions in a densely
populated infaunal assemblage: the effects of the polychaete
Abarenicola pacifica. J. Mar. Res. 39: 735-748
Woodin, S. A. (1974). Polychaete abundance patterns in a marine soft-
sediment environment: the importance of biological interactions.
Ecol. Monogr. 44: 171-187
Woodin, S. A. (1976). Adult-larval interactions in dense infaunal
assemblages: patterns of abundance. J. Mar. Res. 34: 25-41
Woodin, S. A. (1978). Refuges, disturbance, and community structure: a
marine soft-bottom example. Ecology 59: 274-284
119
Woodin, S. A., Jackson, J. B. C. (1979). Interphy1etic competition among
marine benthos. Amer. Zool. 19: 1029-1043
Young, D. K., Buzas, M. A., Young, M. W. (1976). Species densities of
macrobenthos associated with seagrasses: a field study of predation.
J. Mar. Res. 34: 577-592
Young, D. K., Rhoads, R. C. (1971). Animal-sediment relations in Cape
Cod Bay, Massachusetts. I. A transect study. Mar. Bio1. 11: 242-254
Young, D. K., Young, M. W. (1978). Regulation of species densities of
seagrass-associated macrobenthos: evidence from field experiments in
the Indian River estuary, Florida. J. Mar. Res. 36: 569-593