-
1720
Limnol. Oceanogr., 46(7), 2001, 17201733q 2001, by the American
Society of Limnology and Oceanography, Inc.
The interaction between physical disturbance and organic
enrichment: An importantelement in structuring benthic
communities
Stephen Widdicombe1
Plymouth Marine Laboratory, Prospect Place, West Hoe, Plymouth,
PL1 3DH, United KingdomPlymouth Environmental Research Centre,
(Department of Biological Sciences), University of Plymouth, Drake
Circus,Plymouth, PL4 8AA, United Kingdom
Melanie C. AustenPlymouth Marine Laboratory, Prospect Place,
West Hoe, Plymouth, PL1 3DH, United Kingdom
Abstract
The interaction between physical disturbance and organic
enrichment, with respect to its effect on the diversityand
community structure of a macroinfaunal assemblage, has been
examined in a benthic mesocosm experiment.The experiment was
conducted at the Solbergstrand mesocosm (Norwegian Institute for
Water Research) usingsubtidal sediment collected from
Bjrnhordenbukta, a small sheltered bay in Oslofjrd. Ninety-eight
areas of ho-mogenized sediment were subjected to one of seven
levels of organic enrichment, combined with one of sevendifferent
frequencies of physical disturbance, each replicated once. This
structured matrix of physical disturbanceand organic enrichment
treatments demonstrated the combined effects of these factors to be
nonadditive. Diversitywas lower than expected when low frequencies
of physical disturbance acted in conjunction with high levels
oforganic enrichment or when high frequencies of physical
disturbance were combined with low levels of organicenrichment.
Diversity was higher than expected when both disturbance and
enrichment were either high or low.The implications of this
interaction between physical disturbance and organic enrichment for
the application of thedynamic equilibrium model (Huston 1979) to
sediment communities are discussed. Multivariate analysis
alsoshowed community structure to be significantly affected by
physical disturbance, organic enrichment, and interac-tions between
the two. It is concluded that strong interactions between physical
disturbance and organic enrichment,coupled with both small- and
large-scale variability in these factors, could promote
heterogeneity and diversity inbenthic infaunal assemblages.
However, this remains to be tested in field conditions.
Additionally, interactionsbetween physical disturbance and organic
enrichment may have important implications for matters of coastal
zonemanagement.
Marine benthic infaunal assemblages are subjected to avariety of
physical disturbance events, and their response tosuch events has
been studied extensively. These studies haveranged from the
large-scale effects of trawling (e.g., Tuck etal. 1998) and storm
events (e.g., Posey et al. 1996) to thesmall-scale disturbances
caused by mobile bioturbating or-ganisms, both epifaunal (e.g.,
Hall et al. 1991; Thrush et al.1991) and infaunal (e.g., Flach
1992; Widdicombe and Aus-ten 1998). The importance of different
scales of physicaldisturbance was discussed by Zajac et al. (1998)
who, along
1 To whom correspondence should be addressed. Present
address:Plymouth Marine Laboratory, Prospect Place, West Hoe,
Plymouth,PL1 3DH, U.K. ([email protected]).
AcknowledgementsThis work was enabled by a grant from the EC
Large Scales
Facility Programme, was funded in part by the UK Ministry
ofAgriculture, Fisheries, and Food (project No AE1113), and is
acontribution to the PML Coastal Biodiversity research project.
Wethank Torgeir Bakke, John Arthur Berge, Liv Berge, and
JoannaMaloney of NIVA, Oslo, for helping to facilitate this work;
HakonOen, Einar Johannesen, and Oddbjrn Pettersen for their
technicalsupport at the Solbergstrand Marine Station; and the crew
of R.V.Trygve Braarud. We are also very grateful to Mike Kendall,
Mal-colm Jones, and Richard Warwick for constructive comments
anddiscussions during the preparation of this manuscript.
Additionally,we thank Bob Clarke for much needed statistical advice
and guidance.
with others (e.g., Levin and Paine 1974; Hall 1994; Levin1994),
highlighted the importance of cycles of disturbanceand recovery in
maintaining heterogeneity in soft sedimentenvironments and thereby
setting community structure. Theimportance of these cycles was
first suggested by Johnson(1970, 1973), who concluded that the
continual occurrenceof small-scale disturbances can account for
part of the spatialand temporal variations of diversity within
benthic com-munities.
Much of the extensive literature concerned with how ma-rine
benthic communities respond to organic enrichment hasconcentrated
on the effects of anthropogenic inputs associ-ated with freshwater
runoff (e.g., Beukema 1991), aquacul-ture (e.g., Ritz et al. 1989),
and sewage disposal (e.g., Hallet al. 1997). These studies of
anthropogenic eutrophication,together with those examining
naturally occurring organicenrichment events (e.g., Oug et al.
1991), have served tostrengthen the validity of the community
response model ofPearson and Rosenberg (1978). Based on surveys of
macro-benthic communities along gradients of organic
enrichment,this model predicts a decline of suspension feeders and
anincrease in deposit feeders as organic input to the
sedimentincreases, irrespective of the type of organic material
re-sponsible for the enrichment. Other models (e.g., Rhoads etal.
1978) predicted a community response to physical dis-turbance
similar to that described for organic enrichment.
-
1721Diversity, disturbance, and enrichment
Such models suggest that communities along a physical
dis-turbance or organic enrichment gradient display serialchanges
in structure as the intensity or level of the pertur-bation
changes. While these investigations into the effectsof either
physical disturbance or organic enrichment are ex-tensive, they
have not addressed the possibility of interac-tions between these
two factors when acting on an assem-blage simultaneously. Such
interactions may have substantialimplications for predictive
models.
Organic enrichment has been shown to have a significanteffect on
species diversity, a particular aspect of communitystructure. The
relationship between productivity, or energyin the system (Wright
1983), and diversity was describedby Grime (1973a,b), who presented
the intermediate pro-ductivity hypothesis (IPH). This hypothesis
predicted max-imum species diversity occurred at some intermediate
levelof productivity, at which competition for food is reduced
andthe coexistence of potentially competing species was pro-moted.
Grimes (1973a,b) definition of productivity was acombination of
light availability and the level of soil nutri-ents supplied to the
herbaceous plants communities he stud-ied. Using such a definition
it is not unreasonable to viewthe supply of organic material to
marine benthic communi-ties as a resource that could influence
species diversity in amanner predicted by the IPH. A recent field
study assessingthe impacts of terrestrial runoff on macrobenthic
communi-ties supported the IPH as it demonstrated that maximum
di-versity (species richness) corresponded to intermediate levelsof
enrichment (Frouin 2000).
A relationship similar to that for diversity and
organicenrichment has been observed between diversity and phys-ical
disturbance, leading Connell (1978) to propose the in-termediate
disturbance hypothesis (IDH). This hypothesisidentifies the
severity and frequency of disturbance as keyelements in setting
community diversity. Similar to the IPH,the IDH predicts maximum
species diversity occurs at someintermediate level of disturbance,
at which competitive ex-clusion is reduced. This hypothesis was
originally erectedfrom studies of terrestrial communities, and
subsequent ev-idence for it being applicable in the marine
environment hascome, traditionally, from rocky intertidal (e.g.,
Paine 1966)and coral reefs studies (e.g., Connell and Keough 1985).
Forsoft sediment communities, direct evidence for
competitiveexclusion within communities of soft sediments has, to
date,been scarce, which leads to a reluctance in accepting theIDH
as a mechanism for diversity maintenance in this hab-itat. One
study that did claim to provide such evidence wasthat of Kukert and
Smith (1992), which used artificialmounds to explore macrofaunal
community response to buri-al disturbance in the Santa Catalina
Basin. These authorsconcluded that the changes in diversity they
observed werea consequence of reduced pressure from competitive
dom-inants occurring during community succession followingburial
disturbance. In addition to this evidence, previousfield
observations of epibenthic agglutinating formainiferans,also in the
Santa Catalina Basin (Levin et al. 1991) andrecent laboratory-based
experiments on the effects of bio-turbation on meiofaunal and
macrofaunal communities (Aus-ten et al. 1998; Widdicombe and Austen
1999) have elicitedcommunity responses that were in agreement with
the IDH.
Such studies would suggest that there is justification in
con-sidering the IDH as a potential explanation of
diversitymaintenance within soft sediment communities in additionto
the traditionally accepted habitats of coral reefs and therocky
intertidal.
It was evident to Huston (1979) that both the IPH (Grime1973a,b)
and the IDH (Connell 1978) relied on competitivedisplacement. By
combining these two hypotheses, Huston(1979) proposed the dynamic
equilibrium model. This modelassumed diversity represented a
balance between growthrates (productivity/organic enrichment) and
disturbance, withmaximum diversity being observed when an
assemblage re-ceived intermediate levels of both productivity and
distur-bance. Empirical support for the dynamic equilibrium modelis
hard to come by since the necessary, multifactorial ex-periments
are intrinsically more difficult to conduct than ex-periments that
manipulate only a single factor. However,much circumstantial
evidence does exist from field surveysof interacting gradients of
disturbance and rate of displace-ment (see Huston 1994).
Additionally, several models ofplant competition and succession
(e.g., Botkin et al. 1972;Caswell and Cohen 1991) have been used to
conduct sim-ulation experiments, the results of which closely
agreed withthe predictions of the dynamic equilibrium model.
While investigations into the effects of either physical
dis-turbance or organic enrichment are extensive, they have
notaddressed the possibility of interactions between disturbanceand
organic enrichment acting simultaneously on an assem-blage. Such
interactions may have substantial implicationsfor predictive
models, such as the dynamic equilibrium mod-el (Huston 1979). Using
an experimental approach, this pa-per explores the response of a
benthic macrofaunal com-munity to the combined influence of both
physicaldisturbance and organic enrichment. We have attempted
totest the generality of the IDH, IPH, and dynamic
equilibriummodels, and their application to the marine environment
hasbeen discussed. We have examined the hypothesis that theeffects
of organic enrichment and physical disturbance onthe structure and
diversity of a macrobenthic community areadditive.
Methods
Experimental designThe experiment was carried out inthe mesocosm
facility of the NIVA marine research stationSolbergstrand,
Oslofjrd, Norway. The mesocosm was de-scribed in detail by Berge et
al. (1986). On 10 May 1996,muddy sand was collected by Day grab
from Bjrhodenbukta,a sheltered bay in the inner part of Oslofjrd.
On the sameday, the sediment was placed in large (1 m2)
containerswhere it was homogenized and used to fill 98 plastic
buckets(26-cm diameter) to a depth of 20 cm. The buckets of
sed-iment were then placed in a 5 m 3 7 m indoor, epoxy resincoated
concrete basin, at a constant water depth of 100 cm.The water depth
was maintained using an open circulationseawater supply drawn from
60-m depth from the fjord andallowing it to run to waste. A
consequence of this continuoussupply was that a small degree of
larval supply was possible.The sediment in the buckets was allowed
to consolidate for
-
1722 Widdicombe and Austen
Table 1. Results from ANOVA (three factor mixed model) analysis
of abundance data (number of individuals/treatment) for the
16numerically dominant taxa (disturbance and enrichment are fixed
factors, the position of replicates in either block 1 or 2 is a
random factor).Bold values indicate significant differences, p ,
0.05.
Source DF SS MS F P
Heteromastus filiformisDisturbanceEnrichmentDisturbance 3
enrichment
66
36
523058231668739951
871763861120554
1.771.030.63
0.2520.4840.915
Chaetozone setosaDisturbanceEnrichmentDisturbance 3
enrichment
66
36
123532704324988
20594507
694
9.898.332.00
0.0070.0100.020
Paraonis fulgensDisturbanceEnrichmentDisturbance 3
enrichment
66
36
543946524952
906775138
8.536.550.93
0.0100.0190.589
Nuculoma tenuisDisturbanceEnrichmentDisturbance 3 enrichment
66
36
65312771673
109213
46
7.843.041.31
0.0120.1010.213
Cossura longcirrataDisturbanceEnrichmentDisturbance 3
enrichment
66
36
360133395218
600556145
16.007.790.96
0.0020.0120.547
Pseudopolydora pauchibranchiataDisturbanceEnrichmentDisturbance
3 enrichment
66
36
1865676
3077
311113
85
11.751.120.91
0.0040.4480.609
Goniada maculataDisturbanceEnrichmentDisturbance 3
enrichment
66
36
23500366
48310
0.248.201.66
0.9460.0110.066
Nemertea indet.DisturbanceEnrichmentDisturbance 3 enrichment
66
36
118158386
202611
9.255.421.08
0.0080.0290.404
9 weeks before any experimental manipulation began. Thisenabled
the reestablishment of oxygen and nutrient gradientsand provided a
settlement period when infaunal organismscould regain their spatial
positions within the sediment be-fore the experimental
manipulations began.
Buckets were held firmly within a wooden frame and ar-ranged in
two 7 by 7 blocks of 49 buckets. From 12 July1996, for a 12-week
period, each bucket within each blockwas subjected to one of seven
levels of organic enrichment,combined with one of seven different
frequencies of physicaldisturbance. Thus, there was a structured
matrix of 49 treat-ment combinations that was duplicated between
blocks. Gre-co-latin squares were used in the experimental design
so thateach row within each block contained one of each
distur-bance intensity and one of each organic enrichment
level.Additionally, the arrangement of treatments within each ofthe
two blocks was different.
Organic enrichment was administered at the start of
theexperiment by a single application of powdered, dried
As-cophyllum nodosum (L.) Le Jolis (product A120 from AlgeaProducts
A/S; maximum particle diameter 120 mm). Thepowdered A. nodosum
contained 31.5% carbon and 0.9%nitrogen. After the water level in
the mesocosm basin hadbeen lowered to below the edge of the
buckets, A. nodosum
was spread evenly across the sediment surface at seven
treat-ment levels (P0 to P6) equivalent to 0, 12.5, 25, 50,
100,200, and 400 g carbon m22, respectively. In inshore waters,the
rate of deposition of organic matter is generally in theregion of
2575 g cm22 yr21 (Gee et al. 1985 and referencestherein).
Therefore, it was assumed the quantities of organicenrichment used
in the current experiment represented arange of values from very
low to gross enrichment. Thebrown alga Ascophyllum nodosum was
chosen as it is a nat-urally occurring marine product and has been
used success-fully as a source of labile carbon for organic
enrichmentexperiments by previous authors (e.g., Gee et al.
1985;Schratzberger and Warwick 1998).
Physical disturbance of a consistent duration and intensitywas
administered with a mechanical stirrer that raked thesediment
surface to a depth of 2 cm. A plastic, circular disc(250-mm
diameter), covered with 95 stainless steel nails (4-mm diameter),
each protruding approximately 20 mm fromthe disc surface, was
lowered onto the sediment surface androtated at a constant speed
(68 rpm). The nails on the discwere orientated at different angles,
which prevented the dis-turbance being concentrated in certain
areas and ensured auniform disturbance across the entire sediment
surface. Dis-turbances lasted for exactly 24 s, with a constant
number of
-
1723Diversity, disturbance, and enrichment
Table 1. Continued.
Source DF SS MS F P
Diplocirrus glaucusDisturbanceEnrichmentDisturbance 3
enrichment
66
36
93162322
1627
9
4.4413.55
3.00
0.0460.0030.001
c. f. Syllis cornutaDisturbanceEnrichmentDisturbance 3
enrichment
66
36
4524
116
843
2.341.321.10
0.1620.3730.391
Pholoe minutaDisturbanceEnrichmentDisturbance 3 enrichment
66
36
8253
110
1493
9.105.561.36
0.0080.0280.178
Schistomeringos caecusDisturbanceEnrichmentDisturbance 3
enrichment
66
36
5145
139
874
3.811.520.95
0.0640.3130.565
Lumbrineris teturaDisturbanceEnrichmentDisturbance 3
enrichment
66
36
183092
353
1.841.571.18
0.2390.2980.310
Anobothrus gracilisDisturbanceEnrichmentDisturbance 3
enrichment
66
36
31152
212
5296
28.821.921.47
0.0010.2240.127
Lumbrineris fragilisDisturbanceEnrichmentDisturbance 3
enrichment
Eteone flavaDisturbance
66
36
6
247376
4
412
2
1
1.6312.03
1.69
0.47
0.2840.0040.060
0.808EnrichmentDisturbance 3 enrichment
636
840
11
1.081.46
0.1980.132
revolutions for each disturbance event. Suspension of
themechanical stirrer from a fixed height-moveable gantryabove the
mesocosm basin ensured that a constant pressureand depth of
disturbance was applied to all treatments. Priorto the disturbance
being administered, the water level in thebasin was lowered to
below the edge of the buckets. Thisprevented loss of any fine
material during periods of distur-bance. After allowing
approximately 1 h for any resuspendedmaterial within the disturbed
buckets to settle out, the waterlevel in the basin was raised. The
physical disturbance fre-quencies were no disturbance (D0), once
every 4 weeks(D1), once every 2 weeks (D2), once a week (D3), twice
aweek (D4), three times a week (D5), and every day (D6).Although
more rapidly and regularly administered than innatural situations,
this disturbance of the upper sediment lay-ers may be considered as
analogous to the sediment turnovercaused by the movement and
feeding behavior of large in-faunal, deposit feeding species (e.g.,
the heart urchin Bris-sopsis lyrifera).
At the end of the 12-week experimental period, the sedi-ment in
each bucket was sieved over a 500-mm mesh. Theresidue was fixed in
a 10% formaldehyde solution and wassorted under a binocular
microscope. All animals were ex-tracted and identified to the
lowest possible taxonomic level.Environmental conditions remained
constant throughout the
experiment: temperature was 7 6 18C and salinity was 34.56 0.5
psu.
Data analysisMeasures of a diversity were calculated;number of
species, Margalefs species richness, Pielousevenness, and
Shannon-Wiener (loge). Global treatment ef-fects and pairwise
interaction values for these measures wereidentified using a
three-factor mixed model ANOVA com-puted using Systat version 7.0.
Visual interpretation of re-sidual plots was used to confirm that
the data complied withthe assumptions of the ANOVA model, e.g.,
homogeneityof variance. SURFER version 5.02 was used to generate
con-tour plots for number of species using standard krigging anda
low level of smoothing.
Multivariate data analyses followed the methods describedby
Clarke (1993) using the PRIMER version 4.0. softwarepackage (Clarke
and Warwick 1994). Analysis was carriedout on both untransformed
and transformed data usingthe Bray-Curtis similarity measure to
determine the effectsof treatments on different components of the
community.Analysis of untransformed data is sensitive to changes in
theabundance of the dominant species, while analysis of transformed
data detects effects on community structure gen-erally, including
changes in abundance of the lower abun-dance and rare species,
without being unduly influenced by
-
1724 Widdicombe and Austen
Fig. 1. Contour plots demonstrating the abundance (number of
individuals/treatment) of 11 numerically dominant species at
differentcombinations of disturbance intensities and organic
enrichment levels.
dominant, high-abundance species. Two-way crossed AN-OSIM
(analysis of similarities) was carried out to test fortreatment
effects. The ANOSIM test for multivariate data isequivalent to an
ANOVA test for univariate data, except thatANOSIM does not allow
testing for interaction effects. Mul-tidimensional scaling (MDS)
was used to visualize patternsof community change (1) in response
to physical disturbanceat the seven different levels of organic
enrichment and (2)in response to organic enrichment at the seven
different lev-els of disturbance. The significance of differences
betweenthese patterns of response was tested using RELATE (Clarkeet
al. 1993) by calculation of the Spearman rank correlations(r)
between two matrices and in this case between each pairof faunal
similarity matrices. RELATE was then used to as-sess how closely
patterns of response correlated to perfectseriation, by calculating
the strength of correlation betweenthe observed rank
dissimilarities for the experimental bioticdata with an
artificially constructed rank distance matrix sim-ulating perfect
seriation (rank 0 between replicates, rank 1between adjacent
treatments etc., up to rank 6 between ex-treme treatments). The r
values from this analysis (Spear-man rank correlations between the
biotic dissimilarity andthe perfect seriation distance matrices)
enabled the directcomparison of seriation strength between patterns
producedat different disturbance frequencies or organic
enrichmentlevels (Clarke, pers. comm.).
Results
Patterns in species abundanceA total of 81 taxa wererepresented
within the mesocosm community, with the 16numerically dominant taxa
being Heteromastus filiformis,Chaetozone setosa, Paraonis fulgens,
Nuculoma tenuis, Cos-sura longocirrata, Pseudopolydora
pauchibranchiata, Gon-iada maculata, Nemertea indet., Diplocirrus
glaucus, c.f.Syllis cornuta, Pholoe minuta, Schistomeringos
caecus,Lumbrineris tetura, Anobothrus gracilis, Lumbrineris
fra-gilis, and Eteone flava (listed in order of decreasing
abun-dance). Of these, the abundance of six taxa was shown tobe
significantly affected by both physical disturbance andorganic
enrichment (Table 1). In ANOVAs that investigatemultiple
comparisons there is a possibility that, if enoughtests are
performed, significant values will be encounteredeven if the null
hypothesis of no effects was true throughout.To counteract this
possibility it is customary to apply a cor-rection that reduces the
p value at which a test is consideredsignificant in line with the
number of tests performed. Un-fortunately, in tests such as that
presented in this paper,which rely on a large number of multiple
tests, in this case48, such corrections reduce the p value to such
an extent asto make the tests worthless. However, in a total of 48
tests,assuming a model of 5% of error, the number of
falselysignificant results is still quite small, with only three
testspotentially appearing significant even if the null
hypothesis
of no effects was true throughout. In Table 1 there were 19such
significant values and the probability of seeing thismany
significant values by chance, calculated on a singlebinomial model,
is approximately zero. Consequently, a pvalue of 0.05 was
maintained as to indicate situations wheresignificant effects were
likely to be occurring, although thereader should be mindful of the
possibility that a small num-ber of these significant results may
have occurred by chance.For Nemertea indet. and the polychaetes
Paraonis fulgens,Phole minuta, and Cossura longocirrata there was
no inter-action between the effects of disturbance and enrichment
onthe abundance of these taxa (Table 1). A significant inter-action
term indicated the effects of disturbance and enrich-ment on the
abundance of the polychaetes Diplocirrus glau-cus and Chaetozone
setosa were not independent. Threespecies, the bivalve Nuculoma
tenuis and the tube buildingpolychaetes Pseudopolydora
pauchibranchiata and Ano-bothrus gracilis, were significantly
affected by disturbancealone, whereas a further two species, the
large bodied, mo-bile polychaetes Goniada maculata and Lumbrineris
fragilis,were affected by enrichment but not by disturbance.
Graph-ical representation of the effects of physical disturbance
andorganic enrichment on the abundance of these 11 taxa isgiven in
Fig. 1. Where effects were significant there was ageneral trend for
abundance to be lowest where physicaldisturbance was most frequent.
However, changes in abun-dance in response to different levels of
organic enrichmentwere less clear. The abundance of most species
was lowestat the highest levels of enrichment, but, the lowest
levels ofenrichment did not always correspond to the highest
levelsof abundance. For example, the abundance of Nemertea
washighest in treatments receiving intermediate levels of
enrich-ment. No significant effect of either disturbance or
enrich-ment was demonstrated for the remaining five
polychaetespecies; Heteromastus filiformis, c.f. Syllis cornuta,
Schis-tomeringos caecus, Lumbrineris tetura, and Eteone flava.
At the end of the experiment, the structure and diversityof the
mesocosm community from treatments that had re-ceived no enrichment
and no physical disturbance was com-parable to the naturally
occurring community present at thecollection site, Bjrnhordenbukta
(see Valderhaug and Gray1984). Abundance in treatments ranged from
seven to 1194individuals, and the number of taxa per treatment
rangedfrom one to 36.
Patterns in diversityAll diversity measures were signif-icantly
affected by organic enrichment, but only number ofspecies were
significantly affected by physical disturbance(Table 2).
Significant interaction effects between physicaldisturbance and
organic enrichment for three diversity mea-sures, number of
species, species richness, and Shannon-Wiener index, were
demonstrated by the three-factor mixedmodel ANOVA (Table 2). This
indicated that the effects ofphysical disturbance and organic
enrichment, on these three
-
1725Diversity, disturbance, and enrichment
-
1726 Widdicombe and Austen
Table 2. Results from ANOVA (three factor mixed model) analysis
of community structure measures (disturbance and enrichment
arefixed factors, the position of replicates in either block 1 or 2
is a random factor). Bold values indicate significant differences,
p , 0.05.
Source DF SS MS F P
Number of speciesDisturbanceEnrichmentDisturbance 3
enrichment
66
36
415.272701.411313.16
69.21450.2336.48
5.4535.44
2.87
0.0010.0010.001
Number of individualsDisturbanceEnrichmentDisturbance 3
enrichment
66
36
1137535803951
1113402
189589133992
30928
3.272.340.69
0.0870.1620.868
Species richness (Margalef)DisturbanceEnrichmentDisturbance 3
enrichment
66
36
4.244461.157832.2220
0.707410.1930
0.8951
1.9834.37
3.23
0.2130.0010.001
Shannon-Wiener diversityDisturbanceEnrichmentDisturbance 3
enrichment
66
36
1.016737.602626.39821
0.169461.267100.17773
1.2322.19
2.55
0.4050.0010.003
Pielous evenessDisturbanceEnrichmentDisturbance 3 enrichment
66
36
0.212730.236600.56878
0.034560.039430.01580
1.434.461.10
0.3390.0460.392
diversity measures, were nonadditive. The contour plots inFig. 2
demonstrate the response of the four diversity indices(number of
species, species richness, Shannon-Wiener, andPielous evenness) to
increasing levels of organic enrichmentand physical disturbance. It
would appear that, as the rich-ness (number of species) becomes
less important and thedistribution of individuals within each
species (evenness) be-comes more important in the calculation of
the indices, theeffect of increasing the frequency of physical
disturbanceappears to lessen compared with the effect of increasing
theamount of organic enrichment (Fig. 2, Table 2).
The rate at which diversity decreases appears to be greaterin
response to changes in organic enrichment than in re-sponse to
increased physical disturbance (Fig. 2). The patternobserved in
Fig. 2 can be seen as a subset of the patterndescribed by Huston
(1979), with the areas of Hustons mod-el corresponding to lower
physical disturbance and lowerorganic enrichment not being
represented in Fig. 2.
The ANOVA interaction values between the effects ofphysical
disturbance and organic enrichment on the numberof species, species
richness, and Shannon-Wiener (Table 3)highlight where the effects
of physical disturbance and or-ganic enrichment act on a community
either synergisticallyor antagonistically. In areas of no physical
disturbance buthigh levels of organic enrichment, the diversity was
lowerthan would be expected. However, in areas of high
experi-mental disturbance and high levels of organic
enrichment,diversity was higher than would be expected if these
twofactors were acting independently on the community.
Theinteraction values demonstrate that, when acting simulta-neously
on a community, physical disturbance and organicenrichment can act
either synergistically or antagonisticallydepending on the
magnitudes of each factor.
An ameliorating effect, on species diversity, of
physicaldisturbance in areas of high organic enrichment is
identified
in Table 3 and demonstrated in Fig. 2. At high levels oforganic
enrichment (P5 and P6) increasing the physical dis-turbance
frequency resulted in higher diversity, while in ar-eas of low
organic enrichment (P0, P1, and P2) increasingphysical disturbance
caused diversity to decrease. In areasof midlevel organic
enrichment (P3 and P4), diversity re-mained relatively constant
over a range of low to midfre-quencies of physical disturbance
(D3). In areas of low phys-ical disturbance frequency (D0 and D1)
diversity is highestin treatments with little or no organic
enrichment (P0 andP1). At midfrequencies of physical disturbance
(D2D4),maximum diversity was observed in midlevel organic
en-richment treatments (P2P4). While at high physical distur-bance
frequencies, changes in the level of organic enrich-ment had little
effect on diversity. The overall trend was forthe level of organic
enrichment at which diversity was high-est to increase as the
frequency of physical disturbance in-creased (Fig. 3). Where one
type of disturbance factor waslow (physical disturbance or organic
enrichment) diversitydecreased as the intensity of the other factor
increased. How-ever, this decrease was greatest in the combination
of lowphysical disturbance with increasing organic enrichment(Fig.
3). This suggested that the spread of treatments chosenfor organic
enrichment was broader than that of the treat-ments chosen for
physical disturbance.
Patterns in community structureGlobal tests using two-way
crossed ANOSIM showed significant treatment effectsfor both
physical disturbance (untransformed; R 5 0.174, p5 0.001:
transformed; R 5 0.266, p 5 0.000) andorganic enrichment
(untransformed; R 5 0.121, p 5 0.003: transformed; R 5 0.180, p 5
0.001).
Faunal composition of the test community varied in re-sponse to
different intensities of physical disturbance oramounts of organic
enrichment. These differences in com-
-
1727Diversity, disturbance, and enrichment
Fig. 2. Contour plots demonstrating diversity (number of
species, Margalefs species richness, Shannon-Wiener, Pielous
evenness) atdifferent combinations of disturbance intensities and
organic enrichment levels. (P0P6 5 organic enrichment; D0D6 5
physical distur-bance).
-
1728 Widdicombe and Austen
Table 3. Interaction terms from ANOVA showing
pair-wisecomparisons of disturbance and organic enrichment
treatment lev-els.
D0 D1 D2 D3 D4 D5 D6
(a) Number of speciesP0P1P2P3P4P5P6
6.61.98.1
20.20.8
29.627.6
20.95.40.61.8
23.70.4
23.6
20.52.32.02.20.7
23.223.7
20.821.522.3
3.41.42.0
22.0
25.724.4
0.30.00.54.15.1
0.021.223.523.3
0.22.84.8
1.222.525.324.020.1
3.57.0
(b) Margalefs species richnessP0P1P2P3P4P5P6
0.90.21.10.20.2
21.521.1
20.10.80.00.2
20.50.1
20.5
20.10.40.60.50.0
20.620.8
20.120.320.5
0.40.30.5
20.3
21.020.6
0.220.1
0.10.60.8
0.220.420.520.620.1
0.60.7
0.020.020.920.620.0
0.41.1
(c) Shannon-Wiener (loge)P0P1P2P3P4P5P6
0.30.10.30.30.1
20.720.4
20.00.3
20.10.00.0
20.120.1
0.00.20.40.3
20.120.320.4
20.120.120.120.2
0.10.5
20.2
20.420.1
0.220.120.0
0.10.3
0.220.320.320.220.2
0.40.4
20.10.1
20.420.2
0.00.10.5
Fig. 3. Nonmetric multidimensional scaling (MDS) ordinationsof
macrofaunal abundance on transformed data comparing (a)
the effect of organic enrichment on community response
betweenareas subjected to different physical disturbance
frequencies and (b)the effect of physical disturbance on community
response betweenareas subjected to different organic enrichment
levels. (Not all en-richment data are included in MDS for D1 and
D2see text fordetails.) Stress values in italics. (0 5 low physical
disturbance/or-ganic enrichment; 6 5 high physical
disturbance/organic enrichment)
munity structure constituted a community response pat-tern to
either of the two experimental variables and thesepatterns are
represented two-dimensionally as MDS ordina-tions (Fig. 3). As an
approach to identifying interaction be-tween the effects of
physical disturbance and organic en-richment on multivariate
aspects of community structure,RELATE was used to test whether the
community responsepatterns to organic enrichment were influenced by
chang-ing the intensity of physical disturbance (Table 4), or
wheth-er the community response patterns to physical distur-bance
were influenced by changing the level of organicenrichment.
When using transformed data in the RELATE anal-ysis, there were
a high number of significant correlationsbetween the community
response patterns of the differentphysically disturbed treatments,
each subjected to the samerange of organic enrichment (Table 4).
This showed that, forlower abundance and rarer species, the
response to organicenrichment was generally consistent regardless
of the phys-ical disturbance frequency. However, when using
untrans-formed data in the RELATE analysis, fewer significant
cor-
-
1729Diversity, disturbance, and enrichment
Table 4. Pairwise comparisons of the mutivariate community
structure generated by a gradient of organic enrichment in single
distur-bance regimes using RELATE on (a) untransformed and (b)
transformed data and Bray-Curtis similarities. Bold values
indicatesignificant correlations in the community structure in the
treatments compared, p , 0.05. Results represented as r-values with
p-values fora test of no relationship in parentheses.
D0 D1 D2 D3 D4 D5
(a) Untransformed dataD1D2D3D4D5D6
0.33 (0.06)0.62 (0.00)0.70 (0.00)0.35 (0.02)0.15 (0.22)
20.14 (0.82)
0.36 (0.05)0.32 (0.05)
20.04 (0.57)20.03 (0.49)20.05 (0.56)
0.64 (0.00)0.24 (0.09)0.24 (0.13)
20.33 (1.00)
0.23 (0.08)0.22 (0.14)
20.24 (0.97)20.06 (0.60)20.14 (0.79) 20.23 (0.94)
(b) transformed dataD1D2D3D4D5D6
0.64 (0.00)0.85 (0.00)0.60 (0.01)0.44 (0.02)0.50 (0.01)0.21
(0.15)
0.65 (0.00)0.52 (0.01)0.39 (0.03)0.09 (0.29)0.46 (0.02)
0.61 (0.00)0.45 (0.02)0.44 (0.02)0.23 (0.12)
0.16 (0.20)0.36 (0.04)
20.09 (0.62)0.02 (0.43)0.36 (0.04) 20.17 (0.83)
Table 5. R-values from RELATE test of seriation. Bold values
indicate significant correlations between community responses
observedin actual data and a similarity matrix observing perfect
seriation, p , 0.05.
(a) Community response to changes in organic enrichment level at
different frequencies of physical disturbanceDisturbance
frequencyUntransformed data transformed data
D00.3770.475
D10.2570.467
D20.2150.400
D30.4030.387
D420.024
0.148
D50.0060.108
D60.0250.045
(b) Community response to changes in physical disturbance
frequency at different levels of organic enrichmentOrganic
enrichment levelUntransformed data transformed data
P00.3780.152
P10.1190.190
P20.3840.364
P30.3290.449
P40.1020.014
P50.3410.100
P620.018
0.230
relations were observed (Table 4). This suggested thatchanges in
the relative abundance of numerically dominantspecies, in response
to organic enrichment, were affected bydifferent physical
disturbance frequencies.
When using both untransformed and transformeddata, very few
significant correlations were observed be-tween the community
response patterns of different or-ganically enriched treatments,
each subjected to the samerange of physical disturbance
frequencies. Pairwise compar-isons using RELATE showed only one
significant correlationwhen using untransformed data (P0 vs. P5)
and two signif-icant correlations when using transformed data (P2
vs.P3 and P1 vs. P4). This suggests that the response to phys-ical
disturbance of both the numerically dominant speciesand of the low
abundance and rare species was affected bythe amount of organic
material received.
To examine the pattern of community response to eitherorganic
enrichment or physical disturbance in more detail,the ordinations
of community structure in Fig. 3 were com-pared with a pattern that
observed perfect seriation. RE-LATE was used to test for
significant correlations (Table 5).At low physical disturbance
frequencies (D0D3), the com-munity response to increasing levels of
organic enrichmentwas serial, with communities being most similar
to each oth-er when the differences between their respective levels
oforganic enrichment are small. At high physical
disturbancefrequencies (D4D6), seriation breaks down (Table 5a).
A
serial response to physical disturbance was shown mostly
inmidorganic enrichment treatments (P2 and P3) (Table
5b).Additional serial response was observed in P0 and P5
foruntransformed data and in P6 for transformed data.
In two of the ordinations presented in Fig. 3 (D1 and D2)some of
the highest organic enrichment levels were omittedfrom the
ordinations. The community structure in these treat-ments was very
different to that in the other treatments com-pared; consequently,
the latter were tightly clustered in theMDS. By omitting the high
enrichment data it was possibleto visualize the patterns contained
within the remaining datamore clearly. For calculation of seriation
values all data wereincluded.
Discussion
The results of this experiment were consistent with
thosepredicted by the dynamic equilibrium model (Huston 1979),and,
therefore, this model may serve as a basis of an expla-nation for
the relationship between diversity, physical dis-turbance, and
organic enrichment.
The discrepancy between the predicted surface plot of thedynamic
equilibrium model (Huston 1979) and the plotsgenerated from actual
data (Fig. 2a,b) may be due to a num-ber of factors, acting
singularly or in combination. First, anysediment collected from the
field that contains a viable faunawill have a preexisting level of
organic material. Conse-
-
1730 Widdicombe and Austen
quently, it is not possible to produce treatments with no
or-ganic material, and this will, therefore, truncate the
responsesurface of Hustons model. Similarly, the removal of the
sed-iment from the field and its collection in the treatment
buck-ets will result in some physical disturbance of the
sedimentand resident fauna. Therefore, the reestablished
assemblagewas a nonequilibrium abstraction of the natural one and
wasin a state of recovery from this disturbance at the start ofthe
experiment, 9 weeks later. Physical disturbance wouldalso be
introduced by the bioturbation generated by organ-isms already
present within the sediment. This bioturbation,together with
disturbance due to collection, meant it was notpossible to produce
treatments with no physical disturbance,and this will truncate the
response surface of Hustons mod-el. The response surface observed
for number of species inFig. 2 may, therefore, be seen as a subset
of the predictedresponse surface presented by Hustons model, with
the areasof his model corresponding to zero/low disturbance and
or-ganic enrichment absent from the response surface in Fig. 2.
An additional discrepancy between Hustons model andthe response
surface in Fig. 2a may have resulted from thepragmatic way in which
the treatment levels were chosen.Axes units were not assigned to
Hustons conceptual modeland it is, therefore, difficult to
determine the appropriatescale and spread for each of the two
factors. The spread oftreatments chosen for both organic enrichment
and physicaldisturbance represented a subset of a larger
disturbance orenrichment gradient. Although their relative
positions weredetermined, the exact position of experimental
treatments onthe larger gradients was unknown. Therefore, when
describ-ing the frequency of physical disturbance, terms such
ashigh or low are relative to other treatments rather thanto field
disturbances or the disturbance effects of organicenrichment. The
levels used for organic enrichment werejustified in the methods and
were expected to correspond tonaturally occurring high and low
levels of enrichment. Ad-ditionally, diversity appeared to decrease
more rapidly inresponse to increased organic enrichment than to
increasedphysical disturbance. This may also have resulted from
thepragmatic way in which treatment levels were selected. Priorto
this experiment, it was impossible to know what frequen-cy of
disturbance equated to any specified amount of organicenrichment in
terms of impact on the macrobenthic com-munity. The relative rates
at which changes in disturbanceand enrichment altered diversity
were also unknown. There-fore, the severity of the gradient between
low (P0) and high(P6) organic enrichment treatments may have been
greaterthan that of the gradient between low (D0) and high
(D6)physical disturbance treatments.
In addition to supporting the predictions of the
dynamicequilibrium model, this paper has also shown that the
effectsof physical disturbance and organic enrichment do not acton
diversity independently. Diversity was lower than ex-pected
assuming an additive model when low frequencies ofphysical
disturbance acted in conjunction with high levelsof organic
enrichment or when high frequencies of physicaldisturbance were
combined with low levels of organic en-richment. Diversity was
higher than expected when both dis-turbance and enrichment were
either high or low.
The interaction between physical disturbance and organic
enrichment may have several causes. Physical disturbancemay
increase the depth of oxygen penetration in enrichedsediments,
supplying the increased oxygen demand of mi-crobial decomposers as
the additional organic material isprocessed. This will reduce the
impact of oxygen depletionon species of macrofauna sensitive to low
oxygen levels,while also stimulating the activity of aerobic
microbial de-composers and accelerating carbon processing. Hulthe
et al.(1998) observed that fresh material degrades at the same
ratein oxic and anoxic conditions, but old buried material
de-grades 3.6 times faster in oxic conditions than in
anoxicconditions. Physical disturbance may bury fresh material
foranoxic degradation and expose old buried material to
oxicconditions, thus enhancing organic carbon oxidation in ma-rine
sediments. Alternate exposure of material to the activ-ities of
both oxic and anoxic microorganisms through phys-ical disturbance
will result in greater carbon degradation thanwhen material is
subjected to a constant oxic or anoxic re-gime (Aller 1994). The
idea that physical disturbance pre-vents oxygen depletion may be
part of the reason why thepresence of Beggiatoa sp. mats were
limited to treatmentswith high organic enrichment and little or no
disturbance.The presence of this bacterial mat has been associated
witheutrophic conditions and severe oxygen depletion (Sampouand
Oviatt 1991), and its presence in grossly enriched treat-ments was
expected. Consequently, the absence of Beggia-toa sp. from the high
organic treatments that had receivedsome physical disturbance may
have been as a result of in-creased sediment oxygenation. What is
more likely, however,is that a combination of increased sediment
oxygenation andthe physical disruption of these mats as a result of
the dis-turbance prevented their formation in physically
disturbedareas. In field conditions, it has been shown that
interactionsbetween physical disturbance and organic enrichment
occurwhen physical disturbance of the sediment surface causesthe
resuspension of the organic material and results in theremoval of
that material via lateral water movements (e.g.,Guidi-Guilvard and
Buscail 1995). In the current study, thisfinal mechanism would have
had minimal effect because,after disturbance had been administered,
disturbed sedimentwas allowed to settle before the water level was
raised andthe treatments exposed to lateral water movement.
Conse-quently, the results from the current study would suggest
thatlateral removal of organic material is not the only mecha-nism
by which physical disturbance ameliorates the effectsof organic
enrichment. Processes such as increased sedimentoxygenation are
also important. Previous studies using fieldexperiments and
observations have prompted authors to ad-vocate the use of both
direct sediment disturbance (e.g.,ploughing) and the addition of
bioturbating species as meth-ods for reconditioning organically
polluted sediments(e.g., Chareonpanich et al. 1994). By showing
that, in a lab-oratory experiment, sediment disturbance can offset
thedamaging effects of high levels of organic material depositedon
the benthos, this paper supports the conclusions of
theseauthors.
Examining the combined effects of physical disturbanceand
organic enrichment on the abundance of the numericallydominant
species revealed differences in the way speciesrespond to physical
disturbance and organic enrichment. As
-
1731Diversity, disturbance, and enrichment
was predicted by Brenchley (1981), tube building speciessuch as
Anobothrus gracilis and Pseudopolydora pauchi-branchiata were shown
to be extremely sensitive to increas-es in the frequency of
physical disturbance. This sensitivityis assumed to be a result of
either damage to individuals orthe failure of organism to
regain/maintain their positionwithin the sediment during or after
disturbance. Lumbrinerisfragilis and Goniada maculata are large,
mobile species andshowed no such intolerance to physical
disturbance. How-ever, their abundance was significantly reduced in
areas re-ceiving the highest levels of organic enrichment. These
re-sults concurred with the conclusions of Pearson andRosenberg
(1978) in that A. gracilis and P. pauchibranchia-ta were typical of
transitory or second order progres-sive species. It is likely that
the low levels of oxygen, char-acteristic of extremely enriched
environments, preventedlarger species, with relatively small body
surface to volumeratios and no specialized respiratory apparatus,
from per-sisting in these areas. A second species of Lumbrineris
wasrecorded in the current study. Lumbrineris tetura was gen-erally
smaller than Lumbrineris fragilis, and this size differ-ence may
explain why the abundance of the former was notsignificantly
reduced by increased organic enrichment. Ingeneral, it seems that
an organisms tolerance to physicaldisturbance is influenced by its
level of mobility, while tol-erance to the deoxygenation associated
with organic enrich-ment is influenced by an organisms body surface
to volumeratio. It may be assumed, therefore, that species of
limitedmobility and with no specialized branchial structures
wouldbe significantly affected by increases in both physical
dis-turbance and organic enrichment. In the current study,
theabundance of three taxa fitting these criteria was shown todo
just that, with the abundance of Nemertea, Pholoe minuta,and
Cossura longocirrata decreased in response to an in-crease in both
disturbance and enrichment. Two species thatalso demonstrated
significant changes in abundance in re-sponse to both disturbance
and enrichment were the poly-chaetes Chaetozone setosa and
Diplocirrus glaucus; how-ever, statistical analysis demonstrated a
significantinteraction between the two factors. In treatments
combininghigh levels of both enrichment and disturbance, their
abun-dance was higher than would have been expected if bothfactors
had been acting independently. Chaetozone setosapossess many long,
thin branchial filaments, while Diplocir-rus glaucus has eight
stout branchial filaments in addition toa papilated body surface.
It is possible that these speciesused their breathing
apparatus/adaptations to capitalize onany oxygenation resulting
from increased physical distur-bance. However, no such interaction
between disturbanceand enrichment was observed for the abundance of
Paraonisfulgens, despite its possessing 22 pairs of small digitate
gills.It may have been that P. fulgens was more susceptible
todamage from physical disturbance than either C. setosa orD.
glaucus. Alternatively, as the gills of P. fulgens have asmaller
surface area to volume ratio than those of C. setosaor D. glaucus,
the branchial apparatus of P. fulgens mayhave been insufficient to
benefit fully from the possible in-crease in oxygen due to
increased physical disturbance.
The effects of increasing levels of physical disturbance onthe
community structure of samples treated with organic ma-
terial were also complex. When organic enrichment levelswere
either high or low, increasing the frequency of distur-bance did
not have a predictable, serial effect. Serial changesin community
structure as an effect of increasing physicaldisturbance were only
observed at intermediate (25 and 50g cm22) organic enrichment
levels. In the field the supply oforganic material to coastal
sediments is highly variable, bothtemporally and spatially.
Additionally, there are many sourc-es of organic material, some
from large-scale inputs (e.g.,algal blooms, benthic primary
production, terriginous ma-terial from riverine input) and some
that operate at smaller,more localized scales (e.g., fish/mammal
carcasses, macroal-gal detritus). The results presented in this
paper illustrate theimportance of considering this natural
variability in the sup-ply of organic material when predicting or
assessing the ef-fect of physical disturbance on benthic
communities. Coastalzone management often requires monitoring of
the effects ofboth physical disturbance, e.g., demersal fishing and
dredg-ing, and organic enrichment, e.g., fish farm waste. The
re-sults presented here emphasize that these monitoring
effortsshould not address single disturbance types in isolation
butshould consider all environmental conditions that may alle-viate
or exacerbate any community response.
Localized forms of physical disturbance are common infine
sediments and occur at a range of scales and frequen-cies.
Large-scale disturbances (e.g., natural events such asstorms or
anthropogenic impacts such as trawling or dredg-ing) cover large
areas but may be relatively infrequent, al-lowing extended periods
of recovery between disturbances.On a smaller scale, bioturbation
by large macrofaunal or-ganisms have been shown to have a
considerable effect onboth the associated macrofaunal and
meiofaunal communi-ties (e.g., Austen et al. 1998; Widdicombe and
Austen 1999).In areas sheltered from large-scale hydrodynamic
distur-bances, the seasonal and spatial patchiness in the
distributionof bioturbating macrofauna may increase the
heterogeneitywithin an otherwise homogeneous area, in accordance
withthe spatial-temporal mosaic theory (Grassle and Morse-Por-teous
1987). In addition, the interactive effects observed inthis study
suggest that the frequency of physical disturbancecaused during
bioturbation will both structure the fauna andaffect the manner in
which that fauna responds to changesin organic enrichment.
Consequently, such bioturbation in-duced community heterogeneity
will be exacerbated by var-iability in the supply of organic
material. Variable nutrientinput and hydrodynamic features, such as
internal waves(Lennert-Cody and Franks 1999) may act to
concentrateplanktonic organisms, resulting in a patchy supply of
organicmaterial to the benthos. Additionally, physical
structurescaused by macrobenthic organisms, e.g., feeding pits,
tubes,expulsion mounds, can act to increase variability by
furtherconcentrating or dissipating organic material (Yager et
al.1993).
This study has demonstrated experimentally a relationshipbetween
physical disturbance, organic enrichment, and di-versity in marine
benthic communities, consistent with thedynamic equilibrium model.
It has also shown that the ef-fects of physical disturbance and
organic enrichment do notact independently on the abundance of some
species, benthiccommunity structure, and diversity. However, in
order to ful-
-
1732 Widdicombe and Austen
ly accept the generality of these conclusions, it is
imperativethat further evidence demonstrating the effects of
physicaldisturbance and organic enrichment on benthic
infaunalcommunities is obtained from naturally occurring field
sit-uations. Until such validation is available, the results
pre-sented here should be used with an awareness for the poten-tial
limitations of mesocosm experimental approaches,particularly when
applied to community processes. In par-ticular, recovery after
disturbance plays an important role insetting levels of diversity
in benthic communities (e.g., Gras-sle and Morse-Porteous 1987).
Owing to restricted immigra-tion, the importance of this process
within mesocosm sys-tems is reduced, while the influence of an
individualorganisms tolerance to a particular perturbation is
increased.The main consequence of this is to increase the scale
atwhich the experiment is relevant with small-scale distur-bances
in mesocosm systems being analogous to much largerevents in the
field (Widdicombe 2001). An additional artifactis the inability of
some taxa to persist within mesocosms.The benthic assemblages
maintained within the Solbergs-trand mesocosm have been shown to
have higher faunal den-sities and lower species diversities than
locally occurringfield communities (Widdicombe 2001) with the
majority ofthe taxa missing from the mesocosm being
crustaceans.However, community structure analysis has
demonstratedthat the polychaete assemblage was a good
approximationof that recorded in the field. With the majority of
the con-clusions drawn from the current study having been
concen-trated on the responses of polychaete fauna, it may be
as-sumed that the results presented here are robust and
haverelevance to field situations.
References
ALLER, R. C. 1994. Bioturbation and remineralization of
sedimen-tary organic matter: Effects of redox oscillation. Chem.
Geol.114: 331345.
AUSTEN, M. C., S. WIDDICOMBE, AND N. VILLANO. 1998. Effect
ofbiological disturbance on diversity and structure of a
nematodemeiobenthic community. Mar. Ecol. Prog. Ser. 174:
233246.
BERGE, J. A., M. SCHAANNING, T. BAKKE, K. SANDY, G. M. SKEIE,AND
W. G. AMBROSE, JR. 1986. A soft-bottom sublittoral me-socosm by the
Oslofjrd: Description, performance and ex-amples of application.
Ophelia 26: 3754.
BEUKEMA, J. J. 1991. Changes in composition of bottom fauna ofa
tidal-flat area during a period of eutrophication. Mar. Biol.111:
293301.
BOTKIN, D. B., J. F. JANAK, AND J. R. WALLIS. 1972. Some
eco-logical consequences of a computer model of forest growth.
J.Ecol. 60: 849872.
BRENCHLEY, G. A. 1981. Disturbance and community structure:
Anexperimental study of bioturbation in marine soft-bottom
en-vironments. J. Mar. Res. 39: 767790
CASWELL, H., AND J. H. COHEN. 1991. Disturbance,
interspecificinteractions and diversity in metapopulations. Biol.
J. Limnol.Soc. 42: 193218.
CHAREONPANICH, C., H. TSUTSUMI, AND S. MONTANI. 1994.
Effi-ciency of the decomposition of organic matter, loaded on
thesediment, as a result of the biological activity of Capitella
sp.I. Mar. Pollut. Bull. 28: 314318.
CLARKE, K. R. 1993. Non-parametric multivariate analyses
ofchanges in community structure. Aust. J. Ecol. 18: 117143.
, AND R. M. WARWICK. 1994. Changes in marine commu-nities: An
approach to statistical analysis and interpretation.Natural
Environment Research Council, Swindon.
, , AND B. E. BROWN. 1993. An index showing break-down of
seriation, related to disturbance, in a coral-reef assem-blage.
Mar. Ecol. Prog. Ser. 102: 153160.
CONNELL, J. H. 1978. Diversity in tropical rain forests and
coralreefs. Science 199: 13021309.
, AND M. J. KEOUGH. 1985. Disturbance and patch dynamicsof
subtidal animals on hard substrata, p. 125151. In S. T. A.Pickett
and P. S. White [eds.], The ecology of natural distur-bance and
patch dynamics. Academic.
FLACH, E. C. 1992. Disturbance of benthic infauna by
sediment-reworking activities of the lugworm Arenicola marina, p.
8189. In C. H. R. Heip and P. H. Nienhuis [eds.], Proceedings ofthe
26th European marine biology symposiumbiological ef-fects of
disturbances on estuarine and coastal marine environ-ments 30.
FROUIN, P. 2000. Effects of anthropogenic disturbances of
tropicalsoft-bottom benthic communities. Mar. Ecol. Prog. Ser.
194:3953.
GEE, J. M., R. M. WARWICK, M. SCHANNING, J. A. BERGE, AND W.G.
AMBROSE. 1985. Effects of organic enrichment on mei-ofaunal
abundance and community structure in sublittoral softsediments. J.
Exp. Mar. Biol. Ecol. 91: 247262.
GRASSLE, J. F., AND L. S. MORSE-PORTEOUS. 1987.
Macrofaunalcolonization of disturbed deep-sea environments and the
struc-ture of deep-sea benthic communities. Deep-Sea Res.
34:19111950.
GRIME, J. P. 1973a. Competitive exclusion in herbaceous
vegetation.Nature 242: 344347.
. 1973b. Control of species density in herbaceous vegetation.J.
Environ. Manag. 1: 151167.
GUIDI-GUILVARD, L. D., AND R. BUSCAIL. 1995. Seasonal surveyof
metazoan meiofauna and surface sediment organics in anon-tidal
turbulent sublittoral prodelta (northwestern Mediter-ranean). Cont.
Shelf Res. 15: 633653.
HALL, J. A., C. L. J. FRID, AND M. E. GILL. 1997. The response
ofestuarine fish and benthos to an increasing discharge of
sewageeffluent. Mar. Pollut. Bull. 34: 527535.
HALL, S. J. 1994. Physical disturbance and marine benthic
com-munities: Life in unconsolidated sediments. Oceanogr. Mar.Biol.
Annu. Rev. 32: 179239.
, D. J. BASFORD, M. R. ROBERTSON, D. G. RAFFAELLI, ANDI. TUCK.
1991. Patterns of recolonisation and the importanceof pit-digging
by the crab Cancer pagurus in a subtidal sandhabitat. Mar. Ecol.
Prog. Ser. 72: 93102.
HULTHE, G., S. HULTH, AND P. O. J. HALL. 1998. Effects of
oxygenon degradation rate of refractory and labile organic matter
incontinental margin sediments. Geochim. Cosmochim. Acta
62:13191328.
HUSTON, M. A. 1979. A general hypothesis of species
diversity.Am. Nat. 113: 81101.
. 1994. Biological diversity. Cambridge University
Press.JOHNSON, R. G. 1970. Variation in diversity within benthic
marine
communities. Am. Nat. 104: 285300.. 1973. Conceptual models of
benthic marine communities,
p. 148159. In T. J. M. Schopf [ed.], Models in
paleobiology.Freeman.
KUKERT, H., AND C. R. SMITH. 1992. Disturbance, colonization
andsuccession in a deep-sea sediment community: Artificial-mound
experiments. Deep-Sea Res. 39: 13491371.
LENNERT-CODY, C. E., AND P. J. S. FRANKS. 1999. Planktonic
patch-iness in high-frequency internal waves. Mar. Ecol. Prog.
Ser.186: 5966.
LEVIN, L. A., S. E. CHILDERS, AND C. R. SMITH. 1991.
Epibenthic,
-
1733Diversity, disturbance, and enrichment
agglutinating foraminiferans in the Santa Catalina Basin
andtheir response to disturbance. Deep-Sea Res. 38: 465483.
LEVIN, S. A. 1994. Patchiness in marine and terrestrial
systems:From individuals to populations. Philos. Trans. R. Soc.
Lond.,B 343: 99103.
, AND R. T. PAINE. 1974. Disturbance, patch formation
andcommunity structure. Proc. Natl. Acad. Sci. USA 71:
27442747.
OUG, E., T. E. LEIN, R. KUFNER, AND I. B. FALK-PETERSEN.
1991.Environmental effects of a herring mass mortality in
NorthernNorway. Impact on and recovery of rocky-shore and
soft-bot-tom biotas. Sarsia 76: 195207.
PAINE, R. T. 1966. Food web complexity and species diversity.
Am.Nat. 100: 6575.
PEARSON, T. H., AND R. ROSENBERG. 1978. Macrobenthic succes-sion
in relation to organic enrichment and pollution of the ma-rine
environment. Oceanogr. Mar. Biol. Annu. Rev. 16: 229311.
POSEY, M., W. LINDBERG, T. ALPHIN, AND F. VOSE. 1996.
Influenceof storm disturbance on an offshore benthic community.
Bull.Mar. Sci. 59: 523529.
RHOADS, D. C., P. L. MCCALL, AND J. Y. YINGST. 1978.
Disturbanceand production on the estuarine seafloor. Am. Sci. 66:
577586.
RITZ, D. A., M. E. LEWIS, AND M. SHEN. 1989. Response to
organicenrichment of infaunal macrobenthic communities under
sal-monid seacages. Mar. Biol. 103: 211214.
SAMPOU, P., AND C. A. OVIATT. 1991. A carbon budget for a
eu-trophic marine ecosystem and the role of sulfur metabolism
insedimentary carbon, oxygen and energy dynamics. J. Mar. Res.44:
825844.
SCHRATZBERGER, M., AND R. M. WARWICK. 1998. Effects of
theintensity and frequency of organic enrichment on two
estuarinenematode assemblages. Mar. Ecol. Prog. Ser. 164: 8394.
THRUSH, S. F., R. D. PRIDMORE, J. E. HEWITT, AND V. J.
CUMMINGS.1991. Impact of ray feeding disturbances on sandflat
macro-benthos: Do communities dominated by polychaetes or
shell-fish respond differently? Mar. Ecol. Prog. Ser. 69:
245252.
TUCK, I. D., S. J. HALL, M. R. ROBERTSON, E. ARMSTRONG, ANDD. J.
BASFORD. 1998. Effects of physical trawling disturbancein a
previously unfished sheltered Scottish sea loch. Mar. Ecol.Prog.
Ser. 162: 227242.
VALDERHAUG, V. A., AND J. S. GRAY. 1984. Stable
macrofaunacommunity structure despite fluctuating food supply in
subtidalsoft sediments of Oslofjord, Norway. Mar. Biol. 82:
307322.
WIDDICOMBE, S. 2001. Disturbance and diversity in marine
benthiccommunities. Ph.D. thesis. University of Plymouth.
, AND M. C. AUSTEN. 1999. Mesocosm investigation into theeffects
of bioturbation on the diversity and structure of a sub-tidal
macrobenthic community. Mar. Ecol. Prog. Ser. 189: 181193.
, AND . 1998. Experimental evidence for the role ofBrissopsis
lyrifera (Forbes, 1841) as a critical species in themaintenance of
benthic diversity and the modification of sed-iment chemistry. J.
Exp. Mar. Biol. Ecol. 228: 241255.
WRIGHT, D. H. 1983. Species-energy theory: An extension of
spe-cies-area theory. Oikos 41: 496506.
YAGER, P. L., A. R. M. NOWELL, AND P. A. JUMARS. 1993. En-hanced
deposition to pits: A local food source for benthos. J.Mar. Res.
51: 209236.
ZAJAC, R. N., R. B. WHITLATCH, AND S. F. THRUSH. 1998.
Recol-onization and succession in soft-sediment infaunal
communi-ties: The spatial scale of controlling factors.
Hydrobiologia375/376: 227240.
Received: 20 November 2000Accepted: 19 June 2001
Amended: 6 July 2001