ORIGINAL PAPER

Context dependence of marine ecosystem engineer invasionimpacts on benthic ecosystem functioning

Ana de Moura Queiros • Jan Geert Hiddink •

Gareth Johnson • Henrique Nogueira Cabral •

Michel Joseph Kaiser

Received: 2 November 2009 / Accepted: 2 November 2010 / Published online: 19 February 2011

� Springer Science+Business Media B.V. 2011

Abstract Introduced ecosystem engineers can

severely modify the functioning on invaded systems.

Species-level effects on ecosystem functioning (EF)

are context dependent, but the effects of introduced

ecosystem engineers are frequently assessed through

single-location studies. The present work aimed to

identify sources of context-dependence that can

regulate the impacts of invasive ecosystem engineers

on ecosystem functioning. As model systems, four

locations where the bivalve Ruditapes philippinarum

(Adams and Reeve) has been introduced were

investigated, providing variability in habitat charac-

teristics and community composition. As a measure

of ecosystem engineering, the relative contribution of

this species to community bioturbation potential was

quantified at each site. The relevance of bioturbation

to the local establishment of the mixing depth of

marine sediments (used as a proxy for EF) was

quantified in order to determine the potential for

impact of the introduced species at each site. We

found that R. philippinarum is one of the most

important bioturbators within analysed communities,

but the relative importance of this contribution at the

community level depended on local species compo-

sition. The net contribution of bioturbation to the

establishment of sediment mixing depths varied

across sites depending on the presence of structuring

vegetation, sediment granulometry and compaction.

The effects of vegetation on sediment mixing were

previously unreported. These findings indicate that

the species composition of invaded communities, and

the habitat characteristics of invaded systems, are

important modulators of the impacts of introduced

species on ecosystem functioning. A framework that

encompasses these aspects for the prediction of the

functional impacts of invasive ecosystem engineers is

suggested, supporting a multi-site approach to inva-

sive ecology studies concerned with ecosystem

functioning.

Keywords Bioturbation � Ecosystem engineer �Ecological context � Invasive

Introduction

Invasion ecology studies are frequently based on

single-location observations. However, research in

the last decade has demonstrated that the character-

istics of the invaded system play an important part in

the determination of the extent of invasion impacts

A. de Moura Queiros (&) � J. G. Hiddink �G. Johnson � M. J. Kaiser

School of Ocean Sciences, Bangor University, Menai

Bridge, Isle of Anglesey, Wales LL59 5AB, UK

e-mail: anqu@pml.ac.uk

H. N. Cabral

Instituto de Oceanografia, Campo Grande, 1749-016

Lisbon, Portugal

123

Biol Invasions (2011) 13:1059–1075

DOI 10.1007/s10530-011-9948-3

(Colautti and MacIsaac 2004; Ruesink 2003). This

has particular relevance when assessing the effects of

invasive ecosystem engineers on ecosystem function-

ing, i.e. species that can modify the physical–

chemical structure of invaded habitats (Bouma et al.

2009; Crooks 2002; Cuddington and Hastings 2004).

Two key arguments corroborate this perspective.

Firstly, invasion success depends on the charac-

teristics of the non-indigenous species, the invaded

species assemblage and the associated physical

environment (Colautti and MacIsaac 2004). These

characteristics together can be defined as ‘‘ecological

niche opportunities’’ (Shea and Chesson 2002) and

include species interactions, the suitability of the new

environment to the introduced species, and resource

availability (Fridley et al. 2007). In this sense, the

extent of success of an invasion is defined by local

conditions, and cannot be extrapolated to the scale of

the landscape. The success of establishment of a non-

indigenous species is not synonymous to its impact

(Levine and D’Antonio 1999), but it is here pertained

that some level of establishment (transient or not)

should precede any level of impact.

Secondly, species-level effects on ecosystem func-

tioning depend on environmental variation that can

lead to changes in species interactions, performance,

and community composition (Cardinale et al. 2000;

Vaughn et al. 2007). With regards to ecosystem

engineers, environmental variation related to the

physical-chemical characteristics of a habitat can also

modify species-level effects on ecosystem function-

ing. For example, the active mixing and transport of

sediment particles by bioturbating infauna can signif-

icantly enhance biogeochemical ecosystem function-

ing in aquatic sediments, but the net effect of this type

of engineering depends on the type of sediment

observed (Mermillod-Blondin and Rosenberg 2006).

Both arguments support the perspective that the

characteristics of the environment where invasion

takes place can significantly modify the impact of

ecosystem engineers on ecosystem functioning. On

one hand by limiting the extent of the invasion

success, a feature common to all types of invasion;

but on the other, by scaling the net impact of the

ecosystem engineering on the physical–chemical

properties of the invaded habitat. However, a theo-

retical framework that supports the assessment of

sources of context-dependence in order to objectively

quantify these impacts is currently lacking. The

frequent assessment of ecosystem engineer invasion

at single locations cannot be extrapolated to a wider

regional context, where environmental variation

occurs. This hampers the ability of such studies to

truly disentangle the contribution of single species to

ecosystem functioning modification. The present

study aimed at identifying characteristics of invaded

systems that can lead to variation in the impacts of

ecosystem engineers on ecosystem functioning, as

sources of context-dependency.

As model systems, four locations where the same

species has been introduced were investigated, pro-

viding a background of variability in habitat charac-

teristics and community composition against which

invasion impacts could be compared. The mixing

depth of marine coastal sediments was analysed as a

measure of ecosystem functioning. Vertical changes

in the colour of marine sediments express the a

transition in the redox state of iron, that in turn reflects

a vertical stratification in the sequence of electron

acceptors for organic matter degradation (Aller 1982;

Teal et al. 2009). Where oxygen is abundant, oxides of

iron form, and where in low availability, iron occurs in

association with sulphides (Teal et al. 2009). The

vertical location of the iron redox transition is often

used as an indicator of the redox state of marine

sediments, as is associated with the sediment depth to

which mixing occurs (Aller 1982). Sediment mixing

depth can be perceived ecosystem function sensu

Naem et al. (2002) as it depends on abiotic charac-

teristics of sediments, such as granulometry and wave

exposure (Cadee 2000; Mermillod-Blondin and

Rosenberg 2006), but also on its biota (e.g. Volken-

born et al. 2007a). Infauna influence the distribution

of oxygen in marine sediments through active mixing

of sediment particles (bioturbation, Richter 1936) and

burrow ventilation (bioirrigation, Emerson et al.

1984), enhancing solute exchange with the overlying

water column, and disrupting otherwise established

vertical chemical gradients (Pischedda et al. 2008).

Sediment mixing depths are associated with other

important aspects of benthic ecosystem functioning,

such as nutrient cycling (Biles et al. 2002), and

bioturbation as a type of ecosystem engineering is

well described (e.g. Michaud et al. 2005; Solan et al.

2004). Hence, the link between bioturbation and

sediment mixing depth represents a good model to

study the influence of particular ecosystem engineers

on ecosystem functioning.

1060 A. de Moura Queiros et al.

123

The present study investigated four European sites

where the Manila clam, Ruditapes philippinarum

(Adams and Reeve), a species native to the Indo-

Pacific region, has been introduced. Introductions in

European coastal areas have aimed at compensating

for irregular yields of the native European sister

species Ruditapes decussatus (Linnaeus), one of the

main shellfish industries in Western Europe (Goullet-

quer 2006). The Manila clam is a marine suspension

feeding bivalve and an important ecosystem engineer

known to significantly increase sediment erosion and

re-suspension rates (Sgro et al. 2005). The impact of

Manila clam bioturbation on sediment mixing depth

was quantified at each location, and potential sources

of context-dependency for this impact were investi-

gated. The first hypothesis tested was that the

importance of the contribution of the Manila clam

bioturbation at the community level depended on the

composition of each community. Secondly, it was

hypothesised that the influence of community biotur-

bation on the establishment of sediment mixing depth

depended on habitat characteristics at each site.

It is recognized that studies of an observational

nature do not allow for an unambiguous identification

of causal relationships. But the current understanding

of the effects of biological invasions on ecosystem

functioning compel a need for large-scale studies,

across environmental variation, which is difficult to

replicate experimentally.

Methods

Four study-sites where the Manila clam has been

introduced were chosen to cover a range of habitat

characteristics that may influence sediment mixing

depth (Fig. 1), such as the presence of standing

vegetation and sediment type (granulometry and

compaction). The introduction status of the Manila

clam varies between study-sites. In Poole Harbour,

the Venice Lagoon and the Bay of Arcachon the

species has wide-spread, self-sustained populations

that support a target-fishery (Caldow et al. 2007;

Pranovi et al. 2006; Robert et al. 1993), suggesting

that the species can be classified as invasive in those

areas (Colautti and MacIsaac 2004). In the Ria

Formosa the introduction is more recent and sporadic,

and the introduction status is unknown (Campos and

Cachola 2006). An absence of official statistics on

this introduction remains, but anecdotal evidence

collected in a pilot study by the authors supports the

presence of this species in the Ria Formosa (unpub-

lished). No individuals of this species were identified

in the area sampled at the time of the present study.

However, the continuous import of Manila clams to

the Ria (Campos and Cachola 2006), the distinctness

of this site with regards to habitat characteristics, and

the culture of the sister species R.decussatus as the

most important shellfish resource regionally, made

that site relevant within the objectives proposed for

the present study.

The bioturbation activity of all species identified at

each site was quantified, to establish the relative

importance of Manila clam bioturbation within each

community. The contribution of community biotur-

bation to sediment mixing depth establishment was

then assessed in relation to habitat characteristics.

Sampling design: collection of macrofauna

and Sediment Profile Imaging data

Sampling took place between June and August 2007.

At each location (Fig. 1), 11 stations were sampled

around the low water mark to match the stations

defined haphazardly for a pilot study carried-out in the

Fig. 1 Geographical distribution of study-sites. PH: Poole

Harbour, 50.112�N 2.058�W, coarse silt. VL: Venice Lagoon,

45.405�N 12.317�W, fine sand. BA: Bay of Arcachon,

44.710�N 1.133�W, coarse silt. RF: Ria Formosa, 37.020�N

7.843�W, coarse sand. Sediment types were characterized

according to the Udden-Wentworth scale, using samples

collected at study-sides in a pilot-study developed in the

previous year (unpublished data)

Context dependence of marine ecosystem 1061

123

previous year, which aimed at collecting macrofauna

and environmental data. Sampling stations were

spaced at no less than ten meters of each other, within

the low intertidal of study areas. Each station was

defined as a haphazardly orientated transect of 3m in

length, across which a sediment profile imaging (SPI)

camera (Rhoads and Cande 1971) was manually

deployed five times. A light-weight digital sediment

profiling camera (model 3731-D L/W, Marine System

Technology, Inc) was used, which relies on a Nikon

D80 (resolution of 3,872 9 2,592 pixels), and has an

effective sediment penetration of approximately

21 cm (prism window 15.2 cm 9 21.6 cm). Sediment

profile imaging allows for an estimation of the Fe

redox transition in the sediment (used as a proxy of,

and henceforth referred to as, ‘‘sediment mixing

depth’’ or ‘‘MD’’). This variable is characterized as

the depth at which an obvious vertical colour change

in the sediment profile occurred, representing the

interface between oxidized (high reflectance) and

reduced (low reflectance) iron species in the sediment

(Teal et al. 2009). An area of 25 by 25 cm was

sampled for macrofauna around each camera insertion

point, allowing for a fine resolution correspondence to

be drawn between mixing depth and community data.

The large sample area required that macrofauna was

sampled with the use of a hand net (0.5 cm mesh) and

spade (as opposed to a corer), down to a depth of

25 cm. The net was attached to a rigid frame, which

was effectively pushed down to the required depth and

then used to scoop up a sample of consistent size. The

cohesive nature of the sediments meant that all fauna

were retained within the sediment prior to sieving.

Macrofauna samples were sieved in situ using a

0.5 cm mesh to collect large biomass bioturbators in

the sediment, and fixed in 4% buffered formaldehyde.

After 24 h, samples were rinsed and preserved in 40%

industrial methylated spirit until processing. All

samples were stained with Rose Bengal and washed,

after which all fauna was identified to the lowest

possible taxonomic level. Individual wet mass was

recorded, and converted to ash-free dry mass using the

conversion factors estimated by Brey (2001).

Estimation of bioturbation

Community bioturbation potential was estimated

from the species composition, at each sampling

point, using a simplified version of the index used

by Solan et al. (2004) such that:

BPj ¼XSj

i¼1

Ri � AFDMij: ð1Þ

BPj is the community bioturbation potential at

sampling point j, estimated as the sum of the products

of the sediment reworking mode for each species

i (Ri) and its ash-free dry mass (AFDMij). Sj is the

observed number of species (species richness) at

sampling point j. Sediment reworking modes (Ri)

were defined according to a review of marine

literature, in accordance to the recognized notion

that different species have different modes of sedi-

ment reworking (e.g. Michaud et al. 2005, 2006;

Solan et al. 2004)—Table 4 in the annexed materials

section. Ri values ranged from -1 (for sediment

stabilizers) to 4 (for regenerators).

Estimation of ecosystem functioning: sediment

mixing

Sediment mixing depth was assessed at each sampling

point using sediment profile imaging (Rhoads and

Cande 1971). Mixing depth was estimated using the

image analysis software Image J 1.37 (National

Institute of Health, USA). Picture analysis consisted

of estimating the mean mixing depth through conven-

tional threshold analysis. In summary, this image

segmentation technique converts Red–Green–Blue

image layers into binary images, which can be used

to identify the Fe redox transition depth, according to a

standardized pixel intensity threshold. A detailed

description of the protocol is given in the annexed

materials section. Stations for which MD could not be

accurately determined (e.g. image out of focus) were

excluded from all subsequent analyses as lower

resolution hampers the accuracy of MD estimates.

This significantly reduced the number of data points

for analysis. Therefore, data points (five per station)

were analyzed as individual samples throughout, for

all study-sites. The absence of a data structure

associated with the sampling design was verified by

one-way analysis of variance (ANOVA), which indi-

cated that mixing depth variance was not larger

between than within stations, in all but the Venice

dataset (Ria Formosa: F1,22=, P = 0.15, R2 = 4.99

1062 A. de Moura Queiros et al.

123

and n = 24; Arcachon: F1,38 = 1.94, P = 0.17,

R2= 4.86 and n = 40; Venice: F1,42 = 17.58, P\0.01,

R2 = 29.53 and n = 44; Poole: F1,20 = 0.01, P = 0.93,

R2 = 0.04 and n = 22). MD measurements in Venice

were significantly different for two groups of stations, as

identified by a multiple comparison Tukey-test. How-

ever, this grouping structure was not reflected by any of

the other variables included in the study, as verified by

multiple comparison of station values for other environ-

mental variables (i.e. Tukey test grouping).

Data analysis

Community bioturbation potential and sediment

mixing

To test the first hypothesis, the relative contribution

of the Manila clam, and of all other species, to

community bioturbation rates were quantified at each

sampling point (BPij). For each community, a multi-

ple linear regression model was calculated using

community bioturbation potential as the response

variable (BPj), and BPij of all individual species,

across stations, as the predictors. Individual species’

contributions were calculated as the change in the

coefficient of determination (R2) associated with the

addition of each species to the regression model.

Species were considered rare if they occurred in less

than 10% of the samples within each area, and

excluded from this analysis.

To test the second hypothesis, the contribution of

community bioturbation potential to the establish-

ment of mixing depth was quantified in relation to

habitat characteristics, across study-sites. In addition

to SPI and macrofauna data, other habitat character-

istics within each study site were measured (Table 1).

However, not all factors that could have influenced

mixing depth could be measured (e.g. hydrodynamic

factors, Huettel and Webster 2001). This is a

frequently observed limitation of observational eco-

logical studies that may lead to an heterogeneous

distribution of variance of the response variable,

hampering the use of traditional regression methods

that model its mean (MD, Cade and Noon 2003; Cade

et al. 2005). For this reason, in sites were MD

exhibited heterogeneous variance in relation to

Table 1 Habitat characteristics measured at each study-site

Site Variable Data type Assessment

Poole Harbour Nonea – –

Venice Lagoon Seaweed Presence/absence and quantitativeb In situ and image analysisb

Seagrass Presence/absence In situ

Coarse sediment fraction Presence/absence Retention in macrofauna sample

Sediment compaction Quantitative index In situc

Bay of Arcachon Seagrass Presence/absence In situ

Coarse sediment fraction Presence/absence Retention in macrofauna sample

Ria Formosa Seaweed Presence/absence In situ

Seagrass Presence/absence In situ

Coarse sediment fraction Presence/absence Retention in macrofauna sample

We assessed the presence of structuring vegetation (i.e. seaweeds and seagrasses), and sediment characteristics. The later included

variation in the resistance of sediment to the penetration of the SPI prism (i.e. sediment compaction), and the presence of a coarse

sediment fraction (i.e.[5 mm). Variables absent from table (for each location) indicate homogenous conditions (e.g. compaction) or

absence (seaweeds, seagrasses and coarse sediment fraction)a Habitat was homogenous across sampling stations, and seaweeds, seagrasses and coarse sediments were absent from the sampled

areab At this location, due to the presence of an ephemeral seaweed mat (Spyridia filamentosa), the variable was assessed through image

analysis, in addition to the presence/absence data obtained in situ. Canopy height was estimated from sediment profile images and

used in all subsequent analysisc Values ranged from 1 to 4, indicating increasing resistance of sediment to SPI penetration

Context dependence of marine ecosystem 1063

123

community bioturbation potential, the relationships

between these variables (and habitat characteristics)

were modelled by quantile regression (Koenker and

Bassett 1978). This method was chosen because (1) it

enables the analysis of data with heterogeneous

variance; (2) it does not assume a particular vari-

ance–covariance structure, which seemed appropriate

for the analysis of observational data; and (3) it

enables regression models to be fitted to other

quantiles of the response variable than the mean, as

opposed to classical regression methods (Cade and

Noon 2003; Hiddink 2005; Valavanis et al. 2008). In

such cases, the selection of the best linear quantile

models began with locating the quantile of the

univariate relationship between BPj and MD at which

the relationship was significant, based on the confi-

dence intervals and the partial significance of the

coefficients of univariate models, for all quantiles

(Koenker and Bassett 1982). Standard errors were

calculated using bootstrapping. The selected univar-

iate model was compared to the null model (i.e.

including intercept only) using the local variant of the

Wald test. This tests for the significance of a predictor

(model with n ? 1 parameters) in relation to a

simpler model structure (n parameters, i.e. nested)

at a specified quantile (Koenker and Bassett 1982). A

full model was then constructed for each site, using

the selected quantile of the response, and including

also all the variables in Table 1 as main effects and

first order interactions. Selection of the most parsi-

monious model proceeded by removing variables in

all possible orders, using the Wald test to compare

nested models.

For study sites where data variance for the uni-

variate model was homogenous, quantile regressions

were not appropriate, as it was possible to model the

mean of the response (MD). Instead, backward

stepwise regression model selection was used, based

on Akaike’s Information Criterion (Venables and

Ripley 2002). As before, models included MD

(response) and BPj and the other measured habitat

characteristics as main effects and first order

interactions.

The correlation between variables selected for

each model was verified a posteriori using the

Pearson’s product-moment correlation coefficient.

All regression analyses were computed in R (R

Development Core Team, 2009; Vienna, Austria;

http:/www.R-project.org).

Results

Manila clam and community bioturbation

The average community bioturbation potential per

gram of community biomass (average BPj/AFDMj)

was similar across study sites: ABP Poole = 2.52 ±

0.94 g-1, ABPVenice = 2.22 ± 0.72 g-1, ABPArca-

chon = 2.73 ± 0.54 g-1, ABPRia = 2.56 ± 0.56 g-1.

In the three areas where the Manila clam has self-

sustaining populations (Poole, Venice and Arcachon),

it was always selected by the regression models as

one of the species contributing the most to commu-

nity bioturbation rates (Table 2). In Poole harbour,

the Manila clam ranked third out of 21 species

identified in total, accounting for 22.4% of BPj

variability. The most important bioturbator was Mya

arenaria (Linnaeus, 57.8%), and Cerastoderma edule

(Linnaeus, 19.3%) was the second—two other sus-

pension feeding bivalves. In the Venice lagoon, out of

44 species, the Manila clam contributed 17.8% to

BPj, ranking second after Cerastoderma glaucum

(Poiret, 61.1%)—another suspension feeding bivalve.

The deposit-feeding gastropod Cyclope neritea (Lin-

naeus) significantly explained another 14.0% of BPj

variability, and another four species significantly

accounted for an additional 1.4% (Table 2). In the

Bay of Arcachon, out of 36 species, the Manila clam

was the third most important bioturbator in the

community, explaining 24.2% of BPj variability, after

the scavenging polychaete Diopatra neapolitana

(Delle Chiage, 27.8%) and Cerastoderma edule

(Linnaeus, 26.0%). The facultative deposit-feeder

polichaete Melinna palmata (Malmgren) significantly

explained another 18.2% of BPj variability. In the Ria

Formosa, where the Manila clam was not present, BPj

was significantly related to three suspension feeding

bivalves: Cerastoderma edule (Linnaeus, 97.0%),

Ruditapes decussatus (Linnaeus, 2.4%) and Veneru-

pis senegalensis (Gmelin, 0.1%). Pripapulids

explained a further 0.1% of the variation of BPj at

this study-site, where a total of 41 species was

identified.

Bioturbation and mixing depth relationships

across different habitats

Average sediment mixing depth (AMD) estimates

varied between sites, despite similar estimates

1064 A. de Moura Queiros et al.

123

Table 2 Contribution of individual species to community bioturbation potential: ordinary least-squares regression of BPj on BPij

Site Species R2 changea (%) F change Df1, df2b P

Poole Harbour Mya arenaria 57.80 26.04 1, 19 \0.01

Cerastoderma edule 19.30 15.22 1, 18 0.001

Ruditapes philippinarum 22.40 881.40 1, 17 \0.01

Venice Lagoon Cerastoderma glaucum 61.10 65.88 1, 42 \0.01

Ruditapes philippinarum 17.80 34.63 1, 41 \0.01

Cyclope neritea 14.00 454.33 1, 40 \0.01

Musculista senhousia 0.40 13.49 1, 39 \0.01

Neanthes succinea 0.60 35.82 1, 38 \0.01

Ensis siliqua 0.30 18.42 1, 37 \0.01

Calliostoma spp. 0.10 9.93 1, 36 0.03

Bay of Arcachon Diopatra neapolitana 27.80 15.03 1, 39 \0.01

Cerastoderma edule 26.00 21.38 1, 38 \0.01

Ruditapes philippinarum 24.20 40.67 1, 37 \0.01

Melinna palmata 18.20 44.83 1, 36 \0.01

Ria Formosa Cerastoderma edule 97.00 722.34 1, 22 \0.01

Ruditapes decussatus 2.40 96.06 1, 21 \0.01

Venerupis senegalensis 0.10 5.59 1, 20 0.03

Priapulids 0.10 4.89 1, 19 0.04

a R2 change represents the amount of variability of BP explained by each species in the modelb Ratio test degrees of freedom (df1: regression; df2: error)

Table 3 Relationship between mixing depth, bioturbation and habitat characteristics: linear and quantile regression models

Site tau Variable Coefficient t p R2 Fa P

Venice Lagoonb – Intercept 17.68 10.65 \0.01 53.00% F3, 40 = 14.79 \0.01

BPj -1.05 -2.60 0.01

Canopy height -0.58 -4.20 \0.01

Compaction -1.50 -2.81 \0.01

Bay of Arcachonc 0.19 Intercept 11.61 9.69 \0.01 – F3, 21 = 21.58 \0.01

BPj 1.47 1.08 0.29

Coarse -5.96 -1.48 0.15

BPj*Coarse 14.98 140.71 \0.01

Ria Formosad 0.29 Intercept 2.99 2.18 0.04 – F1, 22 = 13.06 \0.01

BPj 0.53 2.24 0.04

a The local variant of the Wald test produces an F-like statistic. Presented values compare the most parsimonious multivariate model

with the null model (Koenker and Bassett 1982)b Backward stepwise linear regression using Akaike’s information criterionc Quantile regression using the data subset defined outside of seagrass Z. noltii patches. Inside the patches, MD could not be

significantly related to any of the measures habitat characteristics, or bioturbation potentiald Quantile regressione Bioturbation modes: B: biodiffuser; C: upward or downward-conveyor; G: gallery-diffuser; R: regenerator; S: surficial-modifier;

St: stabilizerf Feeding modes: DF: deposit-feeder; G: grazer; IG: interface grazer; O: omnivorous; P: predator; S: scavenger; SDF: surface

deposit-feeder; SF: suspension-feeder; SSDF: sub-surface deposit-feeder

Context dependence of marine ecosystem 1065

123

of average community bioturbation potential:

AMDPoole = 6.0 ± 2.2 cm, AMDVenice = 8.7 ±

4.7 cm, AMDArcachon = 14.5 ± 3.0 cm, AMDRia

= 6.6 ± 3 cm. The importance of BPj in determining

mixing depth appeared to vary between sites, depend-

ing on habitat characteristics. In Poole Harbour,

sediment mixing could not be significantly related to

BPj, across all quantiles of that variable. In all other

study-sites, MD increased significantly with BPj

(Table 3; Fig. 2). In the Venice lagoon, MD variance

was homogenous across the range of BPj estimates, so

the relationship between the two variables was

modelled by AIC stepwise linear regression (Table 3;

Fig. 2). This suggests that bioturbation is always a

determinant factor of MD in this area, regardless of

the value of BPj. Bioturbation potential was selected

in the most parsimonious model of mixing depth in

this area, as were sediment compaction (physically

and visually assessed, Table 1) and Spyridia filamen-

tosa cover (canopy height (cm), Table 1). The vari-

ables selected for this model were not significantly

correlated (Pearson’s product-moment correlation:

qBPj/canopy = 0.16, P = 0.29; qBPj/compaction = 0.02,

P = 0.91; q canopy/compaction = 0.24, P = 0.12). In the

Bay of Arcachon and the Ria Formosa, the variance of

MD was heterogeneous across the range of BPj

estimates: deep MD values were observed across the

ranges of BPj, but the MD values corresponding to the

lower quantile(s) increased linearly and significantly

in relation to BPj, (quantile regressions, Table 3;

Fig. 2). Wedge shaped plots suggest that in these two

areas, the relative influence of BPj on MD increases as

bioturbation increases in value. In the Bay of Arca-

chon the effect of bioturbation on MD was restricted

to areas where the seagrass Zostera noltii (Horne-

mann) did not occur (presence/absence data, Table 1).

In these stations, from the available measured habitat

characteristics (Table 1), mixing depth was best

explained by the positive interaction between BPj

and the presence of a coarse sediment fraction

(absence: gray circles and dashed line, Fig. 2; pres-

ence: black circles and dashed line, Fig. 2; Tables 1

and 3). The two variables selected for this model were

not significantly correlated (q = -0.31 with

P = 0.13). Inside the seagrass patch, sediment mixing

could not be significantly related to bioturbation, or

any of the measured habitat characteristics In the Ria

Formosa, the model for MD included BPj (Table 3;

Fig. 2), and none of the other measured variables

(Table 1).

Fig. 2 Relationship between estimated sediment mixing depth

(cm), community bioturbation potential, and measured habitat

characteristics. Lines represent fitted models (Table 3). a Ria

Formosa; b Bay of Arcachon, stations outside of seagrass

patches: black for stations where a coarse sediment fraction

([5 mm) was present, and grey for stations where it was not;

c Poole Harbour; (d.i-d.iv) Venice Lagoon—lines indicate

mixing depth values (cm) predicted by fitted model; bubble

size indicates changes in the relative value of observed mixing

depth values

1066 A. de Moura Queiros et al.

123

Discussion

The findings presented strongly corroborate the

perspective that the characteristics of the system

where invasion takes place can regulate the impacts

of ecosystem engineering invasion on ecosystem

functioning. Both the composition of the invaded

communities, and the measured habitat characteris-

tics were successfully identified as sources of contex-

dependency of this impact, that explained the

observed variation across study sites.

The first tested hypothesis was confirmed, as the

contribution of Manila clam to bioturbation at the

community level was weighted down by the presence

of other high-impact bioturbators in the community.

The Manila clam, where present (Poole Harbour,

Venice Lagoon and Bay fo Arcachon) was always

selected as one of the species in the community

contributing the most to bioturbation, and the fraction

of community bioturbation attributable to this intro-

duced species was fairly consistent across sites

(around 20% of BPj). However, the importance of

this relative contribution within each community was

markedly different. In some areas, community bio-

turbation was heavily dominated by the activity of

one single species that contributed to more than

50% of BPj (e.g. M. arenaria in Poole Harbour, and

C. glaucum the Venice Lagoon). Conversely, in

Arcachon, community bioturbation potential was

more evenly dependent on the activiy of several

species, with contributions comparable to that of the

Manila clam. This implies that the contribution of the

Manila clam to bioturbation at the community level is

more important in Arcachon, than in Venice or Poole

Harbour. This result agrees with the perspective that

the impact of non-indigenous ecosystem engineers

should depend on differences between the ‘‘strength’’

of the engineering carried out by the introduced

species, and that of those that compose the native

community (Bouma et al. 2009). ‘‘Strength’’ has been

characterized as the number of habitats that an

ecosystem engineer can modify to express the

potential of a species to impact the physical charac-

teristics of an environment (Bouma et al. 2009). In

the present study, ‘‘strength’’ can be characterized

directly in relation to the bioturbation mode of each

species, and its biomass, expressing different scales

of impact of different functional groups of bioturba-

tors in sediment particle transport. It is reasonable to

suggest that variation in the ‘‘strength’’ of the

engineering carried out by the introduced species

and by those that compose the native community

should explain variation in the relevance, at the

community level, of ecosystem engineer introduc-

tions in different systems.

The contribution of bioturbation, as a type of

ecosystem engineering, to sediment mixing depth

varied between sites, according to variation in habitat

characteristics. These results confirmed the second

hypothesis that the impact of ecosystem engineers on

ecosystem functioning may be scaled down by the

characteristics of the environment where invasion

takes place. While in the Ria Formosa sediment

mixing depth was only related to bioturbation, in the

Bay of Arcachon and Venice it was also affected by

other habitat characteristics, and in Poole it did not

relate to bioturbation potential at all. These results

support the argument for a need to disentangle the

contribution of organisms to particular processes (e.g.

bioturbation), from net effects on the functioning of

an ecosystem (Solan et al. 2008). This differentiation

seems reasonably essential to objectively quantify the

impact of ecosystem engineers, and may not be

obvious when single location assessments are carried

out. Environmental factors that can act locally to

limit or enhance the effect of the engineering on a

particular aspect of ecosystem functioning may not

be as readily identified unless different systems are

compared between which those factors vary. These

environmental factors, or habitat characteristics, that

also contribute to the determination of a particular

aspect of ecosystem functioning may be character-

ized as the ‘‘functional context’’ against which the

impact of an ecosystem engineer can be measured.

The relative importance of the engineering in relation

to the environmental factors determining the func-

tional context should largely explain the difference

between the performance of a particular species of

ecosystem engineer, and the measured effect on

ecosystem functioning.

It is reasonable to observe that further information

with regards to the variation of the performance of

each analyzed species in relation to variation in the

environmental conditions, between study-sites, could

have enriched the interpretation of the present study.

Changes in the overall depth of bioturbation and the

intensity of burrowing activity have been observed in

relation to environmental variation (Maire et al. 2007;

Context dependence of marine ecosystem 1067

123

Przeslawski et al. 2009). However, bioturbation

reworking modes (Ri), as here portrayed, are not

comparable to the latter because they are not

expected to significantly change in relation to envi-

ronmental conditions. Ri concern the way in which

different organisms rework sediments with particular

emphasis on burrow type (e.g. permanence, size,

depth), and feeding strategy (e.g. head up or down;

suspension or deposit feeder). The existence of

different sediment reworking modes has been amply

validated in the literature as indicative of distinct

effects of organisms on sediment particle transport

(Michaud et al. 2006; Solan et al. 2004)—their

validity is therefore not thought to be here at stake.

For instance, a small-bodied superficial modifier such

as a Hydrobia sp. could be expected to have a much

smaller effect on sediment particle transport than a

polychaete such as Nereis sp., which forms a burrow

of much larger depth and permanence. Ri are

expected to be consistent, regardless of environmen-

tal variation, unless animals are very much at the

limit of their physiological tolerance. However, in

such cases, it is also expected that data collected from

an observational study should reflect this. Commu-

nity bioturbation potential (as calculated in this

study) accounts for Ri but also for the biomass of

each species in the community. This could be

expected to be low if environmental conditions

would be so physiologically intolerable for a partic-

ular species to the point of causing severe modifica-

tion of its burrowing behavior, leading concomitantly

to a low contribution to BPj. Hence, the use of BPij as

a weight system through which the functional rele-

vance of different bioturbating species in each

community can be compared is here seen as a sound

methodology. However, the current understanding of

how ecosystem engineer invasions impact ecosystem

functioning could be enriched if more information

becomes available about the effects environmental

gradients (e.g. temperature) and inter-specific inter-

actions on the functional performance of individual

species.

Some of the habitat characteristics that appeared to

influence the effects of bioturbation on sediment

mixing in this study represent novel findings.

Observed effects of standing vegetation in the Bay

of Arcachon and the Venice lagoon had not previ-

ously been reported. The significant negative effect of

the density of the seaweed S. filamentosa on MD

observed in Venice was not entirely unexpected. The

occurrence of ephemeral algal mats alternate biomass

explosions with deposition (and decomposition) of

large amounts of algae materials to the bottom,

promoting anoxia of superficial sediments (Pihl et al.

1996). Algal mats affect the burrowing behaviour of

benthic infauna by increasing hypoxic and anoxic

conditions in the sediment—e.g. crustaceans and

bivalves tend to move to the surface of sediments in

response to physiological stress (Marsden and

Bressington 2009). The effects of bioturbating

infauna on sediment mixing could be severely limited

for surfacing infauna. The negative coefficient asso-

ciated with BPj in the Venice model for mixing depth

could possibly reflect this effect, suggesting that

potential positive effects of bioturbation on mixing

depth could be impaired in the presence of seaweed

mats. Similarly, in the Bay of Arcachon, bioturbation

significantly affected mixing depth outside of sea-

grass meadows, but not inside. In spite of a lack of

conclusive references on this matter, seagrasses have

the potential to modify sediment mixing depths both

positive and negatively. The three dimensional

structure of meadows can reduce water current

speeds, increasing organic matter deposition and

therefore anoxia in the superficial sediments of

seagrass beds (Blanchet et al. 2004). On the other

hand, oxygen released in the rhizosphere promotes

redox heterogeneity in the sediment (Borum et al.

2006). The present study was unable to efficiently

quantify the potential effects of these processes

(which are expected to be density dependent),

because seagrass distribution was recorded as pres-

ence/absence data. As bioturbation significantly

modified mixing depths outside the meadows but

not inside, it is suggested that those or other processes

intrinsic to the seagrass assemblage play a greater

role in the definition of sediment mixing depths than

bioturbation.

Other habitat characteristics here found to influ-

ence the relationship between bioturbation and mixing

depth establishment agree with the findings of other

authors. The effects of sediment granulometry, as

observed in the Bay of Arcachon, had been previously

reported (e.g. Volkenborn et al. 2007b). This occurs

because the positive effect of bioturbation on solute

and porewater transport can be enhanced in permeable

(coarse) sediments, where advective transport can

occur. In more cohesive (finer) sediment types,

1068 A. de Moura Queiros et al.

123

transport occurs at slower rates, being more limited to

molecular diffusion (Huettel and Webster 2001). In

Arcachon, the presence of coarse sediments appeared

to enhance the effect of bioturbation on the establis-

hement of sediment mixing depths. The effect of

granulometry may potentially also explain why bio-

turbation had no apparent effect on mixing depth in

Poole Harbour, as this is the study site where sediment

is (on average) the finest (Fig. 1). Negative effects of

sediment compaction on sediment mixing, as

observed in the Venice lagoon, had also been

previously reported. Badino et al. (2004) found that

increasing sediment compaction correlated with shal-

lower sediment mixing depths. The dredging gear

used for clam harvesting in that study re-suspended

the fine sediment fraction. This settled near the

sediment surface, increasing sediment compaction

and promoting superficial anoxia. The use of the same

type of fishing gear has been described for the area in

the Venice lagoon sampled in this study, and cited in

association with sediment destabilization (Aspden

et al. 2004). However, no data on the distribution of

the use of fishing gears could be analysed in this study,

and it is possible that bioturbation may have a more

determinant effect on mixing depth establishment in

areas less affected by fisheries in the Venice lagoon.

In summary, the findings presented by this study

suggest that the identification of sources of context-

dependency of ecosystem engineer impacts on eco-

system functioning can be efficiently identified by

multi-site assessments. It is pertained that the iden-

tification of the sources of context-dependency is an

essential step in disambiguating (1) the ecosystem

engineering effects of invasive species; from (2) the

net effects of that type of ecosystem engineering on

ecosystem functioning. While dealing with observa-

tional data, it is thought that the present study

illustrates that observing changes in the functional

effects of ecosystem engineers across real systems

can provide novel information. Specifically, about

species-level effects on ecosystem functioning. It is

suggested that this disambiguation can significantly

and quantitatively explain variation in the impacts of

ecosystem engineers in different systems, and that it

should therefore be a crucial part of invasive ecology

studies concerned with ecosystem functioning.

Acknowledgments This project was funded by the

Foundation for Science and Technology (Ministry of Science,

Technology and Higher Education, Portugal), under contract

BD/SFRH/21338/2005. The authors thank Dr. Jean-Paul

Dreno, Dr. Isabelle Auby and Dr. Martin Plus at the Institut

Francais de Recherche pour l’Exploitation de la Mer

(Arcachon, France), Dr.Sasa Raicevich and Dr. Otello

Giovanardi at the Istituto Centrale per la Ricerca Acientifica

e Tecnologica Applicata al Mare (Chioggia, Italy), and the

Instituto para a Conservacao da Natureza (Portugal), for all the

kind work and facilities made available. The authors also thank

Augusto da Paz at the Cooperativa de Viveiristas da Ria

Formosa for support provided during field work in Portugal.

Dr. Camille Saurel and Vasco Candido are kindly thanked for

all the help provided with field work. The authors thank Martin

Solan, the editor and two anonymous referees for constructive

comments on an earlier version of the manuscript.

Appendix

See Table 4.

Table 4 Estimation of Ri—a review of burrowing behaviour and feeding modes

Species Bioturbation

modeaFeeding

modebRi Reference

Abra segmentum B DF 3 Maire et al. (2006)

Abra spp. B DF 3 Maire et al. (2006)

Acanthochitona spp. S G 1 Raffaelli (1985)

Alitta succinea G O; P; S 3 Ouellette et al. (2004)

Ampelisca brevicornis G SDF 3 Coyle and Highsmith (1994), Schaffner and Boesch (1982)

Aphelochaeta marioni C SDF 2 Coyle and Highsmith (1994), Gaudencio and Cabral (2007)

Aphelochaeta vivipara C SDF 2 Antoniadou and Chintiroglou (2006), Fauchald and Jumars

(1979), Gaudencio and Cabral (2007), Laima et al. (2002)

Aponuphis bilineata C SDF 2 Antoniadou and Chintiroglou (2006), Fauchald and Jumars

(1979)

Bivalve B S; O 3 Marine Biological Association of the United Kingdom (2009b)

Context dependence of marine ecosystem 1069

123

Table 4 continued

Species Bioturbation

modeaFeeding

modebRi Reference

Buccinidae S S; O 1 Marine Biological Association of the United Kingdom (2009b)

Buccinumhumphreysianum

S S; O 1 Marine Biological Association of the United Kingdom (2009b)

Capitellidae C SDF 2 Fauchald and Jumars (1979)

Carcinus spp. S S; O 1 Marine Biological Association of the United Kingdom

(2009b), Atkinson and Naylor 1973, Simmers and Bush 1983

Carcinus maenas S P; S 1 Marine Biological Association of the United Kingdom

(2009b), Atkinson and Naylor (1973), Simmers and Bush

(1983)

Caulleriella zetlandica B SDF; IG; SF 3 Marine Biological Association of the United Kingdom

(2009b), Fauchald and Jumars (1979)

Cerastoderma edule B SF 3 Marine Biological Association of the United Kingdom

(2009b), Mermillod-Blondin et al. (2004)

Cerastoderma glaucum B SF 3 Marine Biological Association of the United Kingdom

(2009b), Mermillod-Blondin et al. (2004)

Chamelea gallina B SF 3 Marine Biological Association of the United Kingdom (2009b)

Cirratulidae C SDF;IG;SF 2 Marine Biological Association of the United Kingdom

(2009b), Laima et al. (2002)

Cirratulus cirratus C DF 2 Marine Biological Association of the United Kingdom

(2009b), Laima et al. (2002)

Clymenura clypeata G SSDF; G 3 Marine Biological Association of the United Kingdom

(2009b), Wlodarska-Kowalczuk and Pearson (2004)

Corophium spp. G SDF; IG;SF 3 Marine Biological Association of the United Kingdom

(2009b), Mermillod-Blondin et al. (2004)

Crassostrea angulata S SF -1 Marine Biological Association of the United Kingdom

(2009b), Murray et al. (2002)

Cyathura carinata B – 3 Olafsson and Persson (1986)

Cyclope neritea B – 3 Pischedda et al. (2008)

Cymodoce truncata S SSDF; G 1 Marine Biological Association of the United Kingdom (2009b)

Diopatra neapolitana C O; P; S 2 Fauchald and Jumars (1979)

Dosinia lupinus B SF 3 Marine Biological Association of the United Kingdom (2009b)

Ensis siliqua R SF 4 Drew (1907)

Eteone picta G P; S 3 Marine Biological Association of the United Kingdom

(2009b), Michaud et al. (2006)

Euclymene oerstedi G SSDF; G 3 Marine Biological Association of the United Kingdom

(2009b), Clavier (1984)

Eunice harassii C O; P S 2 Marine Biological Association of the United Kingdom

(2009b), Fauchald and Jumars (1979)

Eunicidae C O; P S 2 Marine Biological Association of the United Kingdom (2009b)

Euspira catena B O; P S 3 Marine Biological Association of the United Kingdom

(2009b), Kabat (1990)

Exogone spp. B SSDF; IG; SF 3 Fauchald and Jumars (1979)

Gammarella spp. S – 1 Marine Biological Association of the United Kingdom (2009b)

Gammarus spp. S O; P S 1 Marine Biological Association of the United Kingdom

(2009b), Bousfield (1970)

Glycera alba B S 3 Marine Biological Association of the United Kingdom (2009b)

Glycera sp. B S 3 Marine Biological Association of the United Kingdom (2009b)

1070 A. de Moura Queiros et al.

123

Table 4 continued

Species Bioturbation

modeaFeeding

modebRi Reference

Gibbula cineraria S SSDF; G; O; P;

S

1 Marine Biological Association of the United Kingdom (2009b)

Hamynoea hydatis S SDF 1 Marine Biological Association of the United Kingdom

(2009b), Malaquias et al. (2004)

Hedistes diversicolor G SDF; S-SDF;

O; S; SF

3 Marine Biological Association of the United Kingdom

(2009b), Michaud et al. (2006)

Hinia incrassata S O; P; S 1 Marine Biological Association of the United Kingdom

(2009b), Tallmark (1980)

Hydrobia sp S SSDF; G 1 Marine Biological Association of the United Kingdom

(2009b), Biles et al. (2002)

Lekanesphaera levii S S 1

Lepidochitona cinerea S SSDF; G 1 Marine Biological Association of the United Kingdom

(2009b), Evans (1951)

Loripes lacteus B DF 3 Koulouri et al. (2006)

Loripes lucinallis B DF 3 Koulouri et al. (2006)

Lucinella divaricata B SF 3 Koulouri et al. (2006)

Macoma balthica B DF; SF 3 Michaud et al. (2005)

Marphysa sanguinea B O 3 Marine Biological Association of the United Kingdom

(2009b), Gerino et al. (2007)

Melinna palmata G SDF;IG; SF 3 Marine Biological Association of the United Kingdom

(2009b), Olafsson and Persson (1986)

Musculista senhousia S SF -1 Crooks (1996)

Mya arenaria B SF 3 Marine Biological Association of the United Kingdom

(2009b), Michaud et al. (2005)

Nassarius reticulatus S SDF; S 1 Marine Biological Association of the United Kingdom

(2009b), Tallmark 1980

Nemertea B P; S 3 Marine Biological Association of the United Kingdom

(2009b), Murray et al. (2002)

Nephtys spp. B O; P; S 3 Marine Biological Association of the United Kingdom

(2009b), Davie (1993)

Notomastus latericeus G SSDF; SDF; G 3 Marine Biological Association of the United Kingdom

(2009b), D’Andrea and Lopez (1997)

Onuphidae C O; P; S 2 Fauchald and Jumars (1979)

Onuphis eremita C O; P; S 2 Marine Biological Association of the United Kingdom

(2009b), Fauchald and Jumars (1979)

Ostrea edulis S SF -1 Murray et al. (2002)

Pagurus spp. S O; P; S 1 Marine Biological Association of the United Kingdom (2009b)

Paphia aurea B SF 3 Gerino et al. (2007)

Perinereis cultrifera G O; P; S 3 Marine Biological Association of the United Kingdom

(2009b), Scaps (1995)

Philine aperta B O; P; S 3 Marine Biological Association of the United Kingdom

(2009b), Morton and Chiu (1990)

Piromis eruca S SDF 1 Fauchald and Jumars (1979)

Polynoinae B O; P; S 3 Marine Biological Association of the United Kingdom

(2009b), Pernet (2000)

Priapulida C DF; SF 2 Powilleit et al. (1994)

Pseudopolydorapulchra

C SDF 2 Marine Biological Association of the United Kingdom

(2009b), Fauchald and Jumars (1979)

Context dependence of marine ecosystem 1071

123

Sediment profile image analysis: estimation

of mixing depth

Each image was analysed as follows: (1) image was

split into Red–Green–Blue layers, of which only the red

layer was used subsequently, as it produced the best

contrast between the oxidized and the reduced sediment

fractions; (2) the sediment–water interface was elimi-

nated from analysis by manually drawing a polygon

over this area which was defined as background, (3) the

picture was converted into a binary image (foregroung/

background) using a pixel intensity threshold, such that

the oxidized sediment layer is defined as foreground; (4)

the area defined as foreground was measured and

divided by the width of the picture to produce the mean

depth of the Fe redox transition, i.e., mixing depth (cm).

Differences in sediment colour between study-sites may

affect the estimation of this parameter. For this reason,

we developed a standardization procedure to define the

threshold within each study-site that minimized mixing

depth estimation errors within site, and made the data

comparable between sites. The procedure consisted of

the following for each dataset: (1) intensity threshold

was defined manually for each picture to optimize the

contrast between oxidized and reduced sediment frac-

tions; (2) ten pictures were selected to cover the

observed threshold range; (3) of those, five thresholds

were selected to estimate mixing depth (as described)

for each of the ten pictures, covering the observed

threshold range, so that five estimates were obtained for

each picture; (4) all threshold values were plotted

against mixing depth estimates for each picture; (5)

standard threshold intensity was defined, per area,

within the range for which mixing depth estimates

varied the least for the maximum number of pictures.

References

Aller RC (1982) The effects of macrobenthos on chemical

properties of marine sediment and overlying water. In:

McCall PL (ed) Animal-sediment relations. The biogenic

Table 4 continued

Species Bioturbation

modeaFeeding

modebRi Reference

Pseudopolydora spp. C SDF; IG; SF 2 Marine Biological Association of the United Kingdom

(2009b), Fauchald and Jumars (1979)

Rhithropanopeusharrisii

S S; O; P 1 Turoboyski (1973)

Ruditapes decussatus B SF 3 Kanaya et al. (2005)

Ruditapesphilippinarum

B SF 3 Kanaya et al. 2005

Sabellidae C S 2 Fauchald and Jumars (1979)

Scolelepis spp. C SF 2 Marine Biological Association of the United Kingdom

(2009b), Dauer (1983)

Scrobicularia plana B DF; SF 3 Marine Biological Association of the United Kingdom

(2009b), Orvain (2005)

Spionidae C – 2 Fauchald and Jumars (1979)

Streblospio benedicti C SDF; IG;SF 2 Fauchald and Jumars (1979)

Tanais dulongii G – 3 Hassack and Holdich (1987)

Tellina donacina B SDF; IG;SF 3 Marine Biological Association of the United Kingdom (2009b)

Tellina tenuis B DF 3 Marine Biological Association of the United Kingdom (2009b)

Turritella communis S SF 1 Marine Biological Association of the United Kingdom

(2009b), Kennedy (1995)

Venerupis senegalensis B SF 3 Marine Biological Association of the United Kingdom (2009b)

a Bioturbation modes: B: biodiffuser; C: upward or downward-conveyor; G: gallery-diffuser; R: regenerator; S: surficial-modifier;

St: stabilizerb Feeding modes: DF: deposit-feeder; G: grazer; IG: interface grazer; O: omnivorous; P: predator; S: scavenger; SDF: surface

deposit-feeder; SF: suspension-feeder; SSDF: sub-surface deposit-feeder

1072 A. de Moura Queiros et al.

123

alteration of sediments. Plenum Press, New York,

pp 53–102

Antoniadou C, Chintiroglou C (2006) Trophic relationships of

polychaetes associated with different algal growth forms.

Helgoland Mar Res 60:39–49

Aspden RJ, Vardy S, Perkins RG et al (2004) The effects of

clam fishing on the properties of surface sediments in the

Lagoon of Venice, Italy. Hydrol Earth Syst Sci 8:160–169

Atkinson RJA, Naylor E (1973) Activity rhythms in some

burrowing decapods. Helgolander wiss Meeresunters

24:192–201

Badino G, Bona F, Maffiotti A et al (2004) Impact of

mechanical clam harvesting on a benthic habitat: evalu-

ation by means of sediment profile imaging. Aquat Con-

serv 14:S59–S67

Biles CL, Paterson DM, Ford RB et al (2002) Bioturbation,

ecosystem functioning and community structure. Hydrol

Earth Syst Sci 6:999–1005

Blanchet H, de Montaudouin X, Lucas A et al (2004) Het-

erogeneity of macrozoobenthic assemblages within a

Zostera Noltii Seagrass Bed: diversity, abundance, bio-

mass and structuring factors. Estuar Coast Shelf S 61:

111–123

Borum J, Sand-Jensen K, Binzer T et al (2006) Oxygen

movement in seagrasses. In: Larkum AWD, Orth RJ (Ed)

Seagrasses: biology, ecology, and conservation. Springer,

Netherlands, pp 255–270

Bouma TJ, Olenin S, Reise K et al (2009) Ecosystem engi-

neering and biodiversity in coastal sediments: posing

hypothesis. Helgol Mar Res 63:95–106

Bousfield EL (1970) Adaptive radiation in sand-burrowing

amphipod crustaceans. Chesap Sci 11:143–154

Brey T (2001). In: Population dynamics in benthic invertebrates.

A virtual handbook. Wegener Institute for Polar and Marine

Research, Germany. http://www.Awi-Bremerhaven.De/

Benthic/Ecosystem/Foodweb/Handbook/Main.Html.Alfred

Cade BS, Noon BR (2003) A gentle introduction to quantile

regression for ecologists. Front Ecol Environ 1:412–420

Cade BS, Noon BR, Flather CH (2005) Quantile regression

reveals hidden bias and uncertainty in habitat models.

Ecology 86:786–800

Cadee GC (2000) Sediment dynamics by bioturbating organ-

isms. In: Reise K (ed) Ecological comparisons of sedi-

mentary shores. Springer, Berlin, pp 127–148

Caldow RWG, Stillman RA, Durell SEAlVd et al (2007)

Benefits to shorebirds from invasion of a non-native

shellfish. Proc R Soc Lond B Biol Sci 274:1449–1455

Campos CJA, Cachola RA (2006) The introduction of the

Japanese carpet shell in coastal lagoon systems of the

Algarve (South Portugal): a food safety concern. Internet J

Food Saf V:1–2

Cardinale BJ, Nelson K, Palmer MA (2000) Linking species

diversity to the functioning of ecosystems: on the

importance of environmental context. Oikos 91:175–183

Clavier J (1984) Production due to regenaration by Euclymene

Oerstedi (Clapadde) (Polychaeta: Maldanidae) in the

Maritime Basin of the Rance (Northern Britanny). J Exp

Mar Biol Ecol 75:97–106

Colautti RI, MacIsaac HJ (2004) A neutral terminology to

define ‘invasive’ species. Divers Distrib 10:135–141

Coyle KO, Highsmith RC (1994) Benthic amphipod commu-

nity in the Northern Bering Sea: analysis of potential

structuring mechanisms. Mar Ecol-Prog Ser 107:233–244

Crooks JA (1996) The population ecology of an exotic mussel,

Musculista Senhousia, in a Southern California Bay.

Estuaries 19:42–50

Crooks JA (2002) Characterizing ecosystem-level conse-

quences of biological invasions: the role of ecosystem

engineers. Oikos 97:153–166

Cuddington K, Hastings A (2004) Invasive engineers. Ecol

Modell 178:335–347

D’Andrea AF, Lopez GR (1997) Benthic macrofauna in a

shallow water carbonate sediment: major bioturbators at

the Dry Tortugas. Geo-Marine Lett 17:276–282

Dauer DM (1983) Functional morphology and feeding

Behavior of Scolelepis Squamata (Polychaeta: Spionidae).

Mar Biol 77:279–285

Davie JT (1993) Macrofaunal community bioturbation along

an estuarine gradient. Aqua Ecol 27:147–153

Drew GA (1907) The habits and movements of the razor-shell

clam, Ensis Directus, Con. Biol Bull 13:127–140

Emerson S, Jahnke R, Heggie D (1984) Sediment-water

exchange in shalow estuarine sediments. J Mar Res 42:

709–730

Evans FGC (1951) An analysis of the behaviour of Lepido-chitona Cinereus in response to certain physical features

of the environment. J Anim Ecol 20:1–10

Fauchald K, Jumars PA (1979) The diet of worms: a study of

polychaet feeding guilds. Oceanogr Mar Biol Annu Rev

17:193–284

Fridley JD, Stachowicz JJ, Naeem S et al (2007) The invasion

paradox: reconciling pattern and process in species inva-

sions. Ecology 88:3–17

Gaudencio MJ, Cabral HN (2007) Trophic structure of mac-

robenthos in the Tagus Estuary and adjacent coastal shelf.

Hydrobiologia 587:241–251

Gerino M, Frignani M, Munai C et al (2007) Bioturbation in

the Venice Lagoon: rates and relationship to organisms.

Acta Oecol Int J Ecol 32:14–25

Goulletquer P (2006). In: Ruditapes Philippinarum. FAO

fisheries and aquaculture department. http://www.fao.org/

fishery/culturedspecies/Ruditapes_philippinarum/en#tc

BioFea. April

Hassack E, Holdich DM (1987) The tubicolous habit amongst

the tanaidacea (Crustacea, Peracarida) with particular

reference to deep-sea species. Zool Scipta 16:223–233

Hiddink JG (2005) Implications of Liebig’s law of the mini-

mum for the use of ecological indicators based on abun-

dance. Ecography 28:264–271

Huettel M, Webster IT (2001) Porewater flow in permeable

sediments. In: Boudreau BP (ed) The benthic boundary

layer. Transport processes and biogeochemistry. Oxford

University Press, Oxford

Kabat AR (1990) Species of Naticidae (Mollusca, Gastropoda)

described by Linnaeus in the Systema-Naturae (1758).

Zool J Linn Soc-Lond 100:1–25

Kanaya G, Nobata E, Toya T et al (2005) Effects of different

feeding habits of three bivalve species on sediment

characteristics and benthic diatom abundance. Mar Ecol-

Prog Ser 299:67–78

Context dependence of marine ecosystem 1073

123

Kennedy JJ (1995) The courtship, pseudo-copulation behaviour

and spermatophore of Turritella Communis Risso 1826

(Prosobranchia:Turritellidae). J Molluscan Stud 61:

421–434

Koenker R, Bassett G (1978) Regression quantiles. Econome-

trica 46:33–50

Koenker R, Bassett G (1982) Robust-tests for heteroscedas-

ticity based on regression quantiles. Econometrica

50:43–61

Koulouri P, Dounas C, Arvanitidis C et al (2006) Molluscan

diversity along a mediterranean soft bottom sublittoral

ecotone. Sci Mar 70:573–583

Laima M, Maksymowska-Brossard D, Sauriau PG et al (2002)

Fluff deposition on intertidal sediments: effects on benthic

biota, ammonium fluxes and nitrification rates. Biogeo-

chemistry 61:115–133

Levine JM, D’Antonio CM (1999) Elton revisited: a review of

evidence linking diversity and invasibility. Oikos 87:15–26

Maire O, Duchene JC, Rosenberg R et al (2006) Effects of food

availability on sediment reworking in Abra Ovata and A.Nitida. Mar Ecol-Prog Ser 319:135–153

Maire O, Duchene JC, Gremare A et al (2007) A comparison of

sediment reworking rates by the surface deposit-feeding

bivalve Abra Ovata during summertime and wintertime,

with a comparison between two models of sediment

reworking. J Exp Mar Biol Ecol 343:21–36

Malaquias MAE, Condinho S, Cervera JL et al (2004) Diet and

feeding biology of haminoea orbygniana (Mollusca :

Gastropoda: Cephalaspidea). J Mar Biol Ass UK 84:

767–772

Marine Biological Association of the United Kingdom (2009)

In: European register of marine species. http://www.

marbef.org/data. May

Marsden ID, Bressington MJ (2009) Effects of macroalgal mats

and hypoxia on burrowing depth of the New Zealand

Cockle (Austrovenus Stutchburyi). Estuar Coast Shelf S

81:438–444

Mermillod-Blondin F, Rosenberg R (2006) Ecosystem engi-

neering: the impact of bioturbationon biogeochemical

processes in marine and freshwater benthic habitats.

Aquat Sci 68:434–442

Mermillod-Blondin F, Rosenberg R, Francois-Carcaillet F et al

(2004) Influence of bioturbation by three benthic infaunal

species on microbial communities and biogeochemical

processes in marine sediment. Aquat Microb Ecol

36:271–284

Michaud E, Desrosiers G, Mermillod-Blondin F et al (2005)

The functional group approach to bioturbation: the effects

of biodiffusers and gallery-diffusers of the macoma

balthica community on sediment oxygen uptake. J Exp

Mar Biol Ecol 326:77–88

Michaud E, Desrosiers G, Mermillod-Blondin F et al (2006)

The functional group approach to bioturbation: Ii. The

effects of the macoma balthica community on fluxes of

nutrients and dissolved organic carbon across the sedi-

ment–water interface. J Exp Mar Biol Ecol 337:178–189

Morton B, Chiu ST (1990) The diet, prey size and consumption

of Philine Orientalis (Opistobranchia: Philinidae) in Hong

Kong. J Molluscan Stud 56

Murray JMH, Meadows A, Meadows PS (2002) Biogeomor-

phological implications of microscale interactions

between sediment geotechnics and marine benthos: a

review. Geomorphology 47:15–30

Olafsson EB, Persson LE (1986) Distribution, life-cycle and

demography in a brackish water population of the isopod

Cyathura-Carinata (Kroyer) (Crustacea). Est Coast Shelf

Sci 23:673–687

Orvain F (2005) A model of sediment transport under the

influence of surface bioturbation: generalisation to the

facultative suspension-feeder Scrobicularia Plana. Mar

Ecol-Prog Ser 43:43–56

Ouellette D, Desrosiers G, Gagne JP et al (2004) Effects of

temperature on in vitro sediment reworking processes by a

gallery biodiffusor, the polychaete Neanthes Virens. Mar

Ecol-Prog Ser 266:185–193

Pernet B (2000) A scaleworm’s setal snorkel. Invertebr Biol

119:147–151

Pihl L, Magnusson G, Isaksson I et al (1996) Distribution and

growth dynamics of ephemeral macroalgae in shallow

bays on the Swedish West Coast. J Sea Res 35:169–180

Pischedda L, Poggiale JC, Cuny P et al (2008) Imaging oxygen

distribution in marine sediments. The importance of bio-

turbation and sediment heterogeneity. Acta Biotheor

56:123–135

Powilleit M, Kitlar J, Graf G (1994) Particle and fluid biotur-

bation caused by the priapulid worm Halicryptus-Spinu-

losis (Seibold, V). Sarsia 79:109–117

Pranovi F, Franceschini G, Casale M et al (2006) An ecological

imbalance induced by a non-native species: the manila

clam in the Venice Lagoon. Biol Invasions 595–609

Przeslawski R, Zhu Q, Aller R (2009) Effects of abiotic

stressors on infaunal burrowing and associated sediment

characteristics. Mar Ecol-Prog Ser 392:33–42

Raffaelli D (1985) Functional feeding groups of some intertidal

molluscs defined by gut content analysis. J Molluscan

Stud 51:233–239

Rhoads DC, Cande S (1971) Sediment profile camera for in

situ study of organism-sediment relations. Limnol Ocea-

nogr 16:110–114

Richter R (1936) Marken und spruren im hunsriick-schiefer. Ii.

Schichtung Und Grund-Leben. Senckenbergiana 18:

215–244

Robert R, Trut G, Laborde JL (1993) Growth, reproduction and

gross biochemical-composition of the manila clam Rudi-

tapes-Philippinarum in the Bay of Arcachon, France. Mar

Biol 116:291–299

Ruesink JL (2003) One fish, two fish, old fish, new fish: which

invasions matter? In: Kareiva PM, Levin SA (ed) The

importance of species: perspectives on expendabiltity and

triage. Princeton University Press, Princeton, pp 161–178

Scaps P (1995) Intraspecific agonistic behaviour in the polychaete

perinereis cultrifera (Grube). Vie Milieu 45:123–128

Schaffner LC, Boesch DF (1982) Spatial and temporal resource

use by dominant benthic amphipoda (Ampeliscidae and

Corophiidae) on the middle atlantic bight outer conti-

nental-shelf. Mar Ecol-Prog Ser 9:231–243

Sgro L, Mistri M, Widdows J (2005) Impact of the infaunal

manila clam, Ruditapes Philippinarum, on sediment sta-

bility. Hydrobiologia 550:175–182

Shea K, Chesson P (2002) Community ecology theory as a

framework for biological invasions. Trends Ecol Evol

17:170–176

1074 A. de Moura Queiros et al.

123

Simmers J, Bush BMH (1983) Motor programme switching in

the ventilatory system of Carcinus Maenas: the neuronal

basis of bimodal scaphognathite beating. J Exp Biol

104:163–181

Solan M, Cardinale BJ, Downing AL et al (2004) Extinction

and ecosystem function in the marine benthos. Science

306:1177–1180

Solan M, Batty P, Bulling MT et al (2008) How biodiversity

affects ecosystem processes: implications for ecological

revolutions and benthic ecosystem function. Aquat Biol

2:289–301

Tallmark B (1980) Population dynamics of Nassarius Retic-ulatus (Gastropoda, Prosobranchia) in Gullmar Fjord,

Sweden. Mar Ecol Prog Ser 3:51–62

Teal LR, Parker R, Fones G et al (2009) Simultaneous deter-

mination of in situ vertical transitions of color, pore-water

metals, and visualization of infaunal activity in marine

sediments. Limnol Oceanogr 54:1801–1810

Turoboyski K (1973) Biology and ecology of the crab Rhith-ropanopeus Harrisii Ssp. Tridentatus. Mar Biol 23:

303–313

Valavanis VD, Pierce GJ, Zuur AF et al (2008) Modelling of

essential fish habitat based on remote sensing, spatial

analysis and gis. Hydrobiologia 612:5–20

Vaughn CC, Spooner DE, Galbraith HS (2007) Context-

dependent species identity effects within a functional

group of filter-feeding bivalves. Ecology 88:1654–1662

Venables WN, Ripley BD (2002) Statistics complements to

modern applied statistics with S. Springer, New York

Volkenborn N, Hedtkamp SIC, JEEv Beusekom et al (2007a)

Effects of bioturbation and bioirrigation by lugworms

(Arenicola Marina) on physical and chemical sediment

properties and implications for intertidal habitat succes-

sion. Estuar Coast Shelf S 74:331–343

Volkenborn N, Polerecky L, Hedtkamp SIC et al (2007b) Biotur-

bation and bioirrigation extend the open exchange regions in

permeable sediments. Limnol Oceanogr 52:1898–1909

Wlodarska-Kowalczuk M, Pearson TH (2004) Soft-bottom

macrobenthic faunal associations and factors affecting

species distributions in an Arctic Glacial Fjord (Kon-

gsfjord, Spitsbergen). Polar Biol 27:155–167

Context dependence of marine ecosystem 1075

123