Ecological implications of the presence of the tube-building polychaete Lanice conchilega on soft-bottom benthic ecosystems

Page 1

Mar Biol (2008) 154:1009–1019

DOI 10.1007/s00227-008-0992-1

ORIGINAL PAPER

Ecological implications of the presence of the tube-building polychaete Lanice conchilega on soft-bottom benthic ecosystems

Gert Van Hoey · Katja Guilini · Marijn Rabaut · Magda Vincx · Steven Degraer

Received: 15 May 2007 / Accepted: 28 April 2008 / Published online: 14 May 2008© Springer-Verlag 2008

Abstract The common tube-building polychaete Laniceconchilega is known as a habitat structuring species andcan form dense aggregations. The eVects of L. conchilegaon the surrounding benthic community have received littleattention, especially in subtidal areas. Therefore, the pres-ence of L. conchilega in diVerent habitats in the North Seaand its eVect on the abundance, species richness, diversityand community structure in these habitats are evaluated inthe present paper, based on data from the ICES North SeaBenthos Survey of 2000. Lanice conchilega has a widegeographical distribution and a low habitat specialization,but optimally occurs in shallow Wne sands. In the presentstudy, the presence of L. conchilega resulted in a densityincrease and a signiWcant (positive) correlation of the ben-thos density with the density of L. conchilega. Furthermore,the species richness (number of species) increased withincreasing density of L. conchilega. This trend was, how-ever, not consistent: the number of species reached more orless an asymptotic value or even decreased after reaching acritical density of L. conchilega (>500–1,000 ind/m²), asobserved in shallow Wne sands. The same overall patternwas detected concerning the expected number of species.The N1-diversity index showed similar or slightly higher

values in L. conchilega patches compared to patches with-out L. conchilega. From the results of the community anal-ysis, it can be concluded that the species, which wereresponsible for the increase of the diversity, belonged to theoverall species-pool of that habitat. The eVects on densityand diversity diVered between the four discerned habitats(shallow muddy sand, shallow Wne sand, shallow mediumsand and deep Wne sand), and were most pronounced inshallow Wne sands. These patterns can be attributed to thehabitat structuring capacity of L. conchilega. The mecha-nisms responsible for the increase of the habitat quality inpatches of L. conchilega can be summarized as (1) changesin the hydrodynamics, (2) increases of the habitat stabilityand oxygen supply, and (3) a creation of habitat heteroge-neity in a uniform environment. In this way, L. conchilegaalters the habitat characteristics and aVects other organisms,and can therefore even be considered as an ecosystem engi-neer. In other words, L. conchilega patches are responsiblefor an increased habitat quality in an otherwise uniformhabitat, which results in a higher survival of the surround-ing benthic species.

Introduction

Biogenic habitat structures play a major role in structuringthe distribution pattern of benthic fauna by modifying thesediment (Eckman et al. 1981; Carey 1987) and hydro-dynamic parameters (Eckman 1983), or by changing inter-actions between species (Woodin 1978). Some tube-buildingpolychaetes provide considerable structures in the other-wise relatively unstructured soft-bottom sediments (Woodin1978; Zühlke et al. 1998; Zühlke 2001; Bolam and Fernan-des 2002; Callaway 2003b; Rees et al. 2005). An exampleof a structuring tube-building polychaete is the sand mason,

Communicated by M. Wahl.

G. Van Hoey · K. Guilini · M. Rabaut · M. Vincx · S. DegraerBiology Department, Marine Biology Section, Ghent University (UGent), Sterre Campus, Krijgslaan 281-S8, 9000 Ghent, Belgium

G. Van Hoey (&)ILVO Institute for Agriculture and Fishery Research, Unit Animal Sciences, Fisheries, Ankerstraat 1, 8400 Ostend, Belgiume-mail: [email protected]

123

Page 2

1010 Mar Biol (2008) 154:1009–1019

Lanice conchilega, which lives in a tube of sand or shellbreccia attached to an inner thin organic layer. The tubeitself is crowned with a sand-fringe, which protrudes 1–4 cm above the sediment surface (Ziegelmeier 1952). Thisspecies can reach densities of several thousands of individ-uals per square metre (Buhr and Winter 1976; Ropert andDauvin 2000; Van Hoey et al. 2006), is found on all Euro-pean coasts and colonizes a wide variety of intertidal andsubtidal sediments down to about 1,900 m (Hartmann-Schröder 1996; Ropert and Dauvin 2000).

Despite its wide distribution and the formation of some-times dense aggregations, the eVects of the presence ofL. conchilega on the surrounding benthic community havereceived little attention. The interaction between L. conchi-lega and the benthos was previously described by Zühlkeet al. (1998), Dittmann (1999) and Zühlke (2001) on twosandXats of the East Frisian Waddensea (the GröningerPlate and the Dornumer Nacken). These studies alsodescribed some experiments on the eVect of artiWcial tubeson the benthos. Both studies concluded that the benthos intidal Xats has a temporary and optional association with thetubes of L. conchilega and that the presence of such struc-tures enriched the Arenicola-dominated sandXat associationin abundance and species numbers. This indicates that L.conchilega is a habitat structuring species, which aVects thesurrounding benthic community. In the study of Callaway(2006), on an exposed beach in South Wales, it was con-cluded that not only groups of tubes, but also single poly-chaete tubes aVect the environment. This ability can beattributed to the following mechanisms (Callaway 2006):(1) the tubes provide a settlement surface for larval andpost-larval benthic organisms (Qian 1999), (2) there is animproved oxygen supply in the sediments surroundingL. conchilega tubes (Forster and Graf 1995), (3) the tubesaVect the current velocities in the benthic boundary layer(Eckman et al. 1981; Heuers et al. 1998; Hild and Günter1999), (4) the tubes have a stabilizing eVect on the sedi-ment, and (5) the space between tubes can serve as a refugefrom predation (Woodin 1978).

Nevertheless, these conclusions were not conWrmed forother habitats, especially in subtidal areas, where L. conchi-lega is widespread. A large-scale benthos survey, performedin the subtidal of the North Sea in 2000–2001 under the guid-ance of the Benthos Ecology Working group of ICES (Reeset al. 2002, 2007), provided an opportunity to focus on sub-tidal areas. The resulting dataset formed the basis of thedescription of the ecological implication of the presence ofL. conchilega on some soft-bottom benthic ecosystems in theNorth Sea. In other words, the present study aimed to investi-gate the eVects of the presence of L. conchilega on the abun-dance, species richness, diversity and community structure indiVerent soft-bottom habitats in the North Sea, in view of theecosystem engineering function of L. conchilega.

Materials and methods

Study area

The study area covers most of the English Channel and theNorth Sea (delimited by Norway and Denmark in the east,the UK in the west and Germany, the Netherlands, Belgiumand northern France in the south). The North Sea (51°–61°N, 3°W–9°E) is divided into a number of loosely deW-ned areas: a relatively shallow southern North Sea (South-ern Bight and German Bight), the central North Sea(Doggerbank, Oysterground), the Northern North Sea, theNorwegian Trench and the Skaggerak, from which the lasttwo areas were not included in the present study (Fig. 1).

Data origin

Under the guidance of the Benthos Ecology Working groupof ICES, a total of 2,227 macrobenthic samples (1,405 sta-tions) were gathered in the North Sea and English Channelin the years 2000 or 2001. These data originate from vari-ous projects, including national monitoring surveys (Reeset al. 2002, 2007). The total dataset was used to describethe spatial distribution of L. conchilega in the North Sea.To enable detailed analyses on the eVect of L. conchilegaon the benthos, a uniform dataset was selected with onlysamples taken with a 0.1 m² Van Veen or Day grab andsieved alive on a 1-mm sieve. This resulted in a Wnal data-set of 1,098 samples (comprising 513 diVerent stations).

All data was incorporated into a database, and taxonomicinter-comparisons were performed (Rees et al. 2002, 2007).These data modiWcations were executed during several work-shops of the ICES study group on the North Sea BenthosProject 2000. After taxonomic clearance, a dataset consistingof 717 taxa (further referred to as species) was obtained. Thedensity of L. conchilega in the present study is based on indi-vidual counts, rather than tube counts and should thus beconsidered as minimum counts (Van Hoey et al. 2006).

The sedimentological characteristics of the diVerent sam-ples were coded according to the following sediment clas-ses: (a) mud, (b) muddy sand, (c) Wne to medium sand, (d)medium to coarse sand, (e) sand and gravel, and (f) mixedsediments (Report ICES CM 2004/E:05) (ICES 2004).Additionally, water depth at each sampling station wasrecorded. The diVerent habitat types were distinguished bysediment classes and bathymetrical information (shallow<70 m and deep >70 m) (following the benthic communityanalyses of Künitzer et al. 1992; Rees et al. 2007).

Data analysis

The eVects of L. conchilega on the benthos were investi-gated for every habitat type in which the species was found

123

Page 3

Mar Biol (2008) 154:1009–1019 1011

and for which a representative number of samples (>100)was available (Fig. 2). This number of samples was chosento exclude uncertainties in the results. The following uni-variate indices were used to describe the benthos (exclud-ing L. conchilega) in each sample: (1) density N, (2) speciesrichness S, expressed as number of species per sample (i.e.,per 0.1 m²), (3) the exponential form of the Shannon–Wiener index N1 (Hill 1973) and (4) expected number ofspecies (ES 50) (Hurlbert 1971). The relations between thoseunivariate indices and the density of L. conchilega in thediVerent habitats were visualized based on diVerent densityclasses of L. conchilega (deWned in such way that theygive the best reXection of the observed patterns). A Mann–

Whitney U test was used to test for diVerences in the univariateindices between samples with and without L. conchilegaand a Spearman rank correlation analysis was done todescribe the correlation between the univariate indices andthe density of L. conchilega. Non-parametric tests wereused because the assumptions for parametric tests, evenafter transformation, were not fulWlled (Conover 1971).

The benthic community structure within the diVerenthabitats was analyzed with non-parametric multidimen-sional scaling (MDS) on the fourth-root transformed data-set, in which the samples containing L. conchilega (group1) and the samples without L. conchilega (group 2) werelabeled a priori. Analysis of similarity (one-way ANOSIM)

Fig. 1 Density distribution of L. conchilega in the entire North Sea and English Channel. Open circle 0 ind/m², small Wlled circle 1–99 ind/m², large Wlled circle 100–499 ind/m², larger Wlled circle 500–999ind/m², largest Wlled circle >1,000 ind/m²

English Channel

Southern Bight

Oyster ground

Dogger bank German

Bight

Northern North Sea

Western

North S

ea

U.K.

Norway

Belgium

France

TheNetherlands

Germany

Den

mar

k

123

Page 4

1012 Mar Biol (2008) 154:1009–1019

was used to test for diVerences between the two groups andSIMPER was used to investigate which species contributedmost to the dissimilarity between the groups. These calcu-lations were done with the Primer 5.2.9 software package(Clarke and Warwick 2001).

Results

Distribution pattern

In 2000–2001, L. conchilega was found in the entire NorthSea and English Channel (Fig. 1) (25% of the stations). Inthe central English Channel, L. conchilega was seldomlyfound (<5% of the samples), whereas the species occurredfrequently in the entire North Sea (42% of the samples).The areas with the highest frequency of occurrence anddensities were the German Bight, the central part of theNorth Sea (east of the Dogger Bank) and along the French,Belgian and Dutch coast. In the deeper northern part of theNorth Sea, L. conchilega was frequently found, but in lowdensities (<100 ind/m²), whereas in the western North Sea,L. conchilega was seldomly found and only in very lowdensities (<100 ind/m²).

Habitat preferences

Lanice conchilega was found in most soft-bottom sedimenttypes in the North Sea, with diVerences in frequency ofoccurrence and average densities between the discernedhabitat types (Fig. 2). No deWnitive conclusion of the occur-rence of L. conchilega in shallow mud, deep muddy sandsand deep medium sands could be made, due to the low

number of samples in these habitat types (<30 samples). Asfor the other habitats, the highest percentages of occurrence(41–51%) and highest average densities (138–419 ind/m²)of L. conchilega in shallow areas were observed in mixed,muddy and Wne sand sediments. In shallow medium andcoarse sediments, the frequencies of occurrence (24 and30%, respectively) and average densities (17 and 12 ind/m²,respectively) were much lower. In deep muds and Wnesands (>70 m), L. conchilega occurred frequently (53 and45%, respectively), but in low average densities (32 and14 ind/m², respectively). Although L. conchilega was foundin all habitat types, for reasons of representativeness furtherdetailed analyses were only done for habitats containingmore than 100 samples (deep Wne sand, shallow muddysand, shallow Wne sand and shallow medium sand).

EVect of Lanice conchilega on the benthic characteristics

Presence/absence of Lanice conchilega

A highly signiWcant diVerence (P < 0.0001) in benthic den-sity and species richness (excluding L. conchilega) wasfound between L. conchilega samples and samples withoutL. conchilega in shallow muddy sands, Wne sands andmedium sands (Table 1). Those diVerences in density andspecies richness were signiWcant in deep Wne sands(P = 0.0115 and P = 0.0027). The N1-diversity index inL. conchilega samples diVered signiWcantly in shallow Wnesands (P < 0.0001), medium sands (P = 0.0012) and deepWne sands (P = 0.0225). Only in shallow muddy sands, nosigniWcant diVerence was found (P = 0.1299). The ES(50)was only signiWcantly diVerent in shallow Wne sands andmedium sands (P < 0.0001).

Fig. 2 Percentage of occur-rence (bars, left axis) and average density (log ind/m²) (squares, right axis) of L. conchilega in the diVerent habitat types (with indication of the total number of samples). The four habitats, which were represented by more than 100 samples in the database, were encircled

0

10

20

30

40

50

60

70

80

Sha

llow

mud

Dee

p m

ud

Sha

llow

mud

dy s

and

Dee

p m

uddy

san

d

Sha

llow

fine

san

d

Dee

p fin

e sa

nd

Sha

llow

med

ium

sa

nd

Dee

p m

ediu

m

sand

Sha

llow

coa

rse

sand

Sha

llow

mix

ed

sedi

men

ts

%

0,1

1

10

100

1000

log

(ind/

m² )

236/12083/44

27/1

309/133192/87

131/38

10/7

44/13

48/20

14/2

123

Page 5

Mar Biol (2008) 154:1009–1019 1013

Correlation between benthic univariate indices and density of Lanice conchilega

In the four habitats, the density of the surrounding benthosincreased with increasing density of L. conchilega (Fig. 3a).The increasing trend of the density was comparable in thefour habitats. The correlation between the density of thebenthic fauna and the density of L. conchilega was positiveand signiWcant in all habitats, but was strongest in shallowWne sands (Spearman R: 0.63) and was lowest in deep Wnesands (Spearman R: 0.23) (Table 1).

Although species richness diVered strongly betweenhabitats, a signiWcant positive correlation was foundbetween the species richness and the density of L. conchi-lega in all habitats, with the highest value in shallow Wnesands (Spearman R: 0.65) and the lowest in deep Wne sands(Spearman R: 0.27) (Table 1). In shallow muddy sands, thecorrelation was atypical: the species richness decreasedwith higher densities of L. conchilega. In shallow muddysands, the species richness decreased when the density of

L. conchilega exceeded 1,000 ind/m², while in shallow Wnesands, the species richness leveled oV at 500 ind/m² ofL. conchilega (Fig. 3b).

The N1-diversity index and its relation with L. conchi-lega density diVered between the habitats (Fig. 4a). In shal-low muddy sands, the N1-diversity index did not increasewith the L. conchilega density and did not show a signiW-cant correlation (Spearman R: 0.08; P = 0.22) (Table 1),whereas a minor, through signiWcant to very high signiW-cant correlation was observed in the other three habitats.The strongest correlation was found in shallow Wne sands(Spearman R: 0.39) (Table 1).

The trend in the ES(50) was comparable with that of thespecies richness (Fig. 4b), with some small diVerences: (1)in shallow muddy sands and deep Wne sand no increase andno signiWcant correlation in ES(50) with the L. conchilegadensity was observed, (2) in shallow Wne and medium sandsan increase and a signiWcant correlation (Spearman R: 0.39–0.34, respectively) was found, but the curve leveled oV at100 ind/m² in medium sands and decreased in Wne sandswhen the density of L. conchilega exceeded 1,000 ind/m².

EVect of Lanice conchilega on the community structure

When the community structure in the diVerent habitats wasvisualized by MDS, it was clear that the samples containingL. conchilega individuals (group 1) were not clearly sepa-rated from the samples without L. conchilega (group 2),due to their central position in the MDS (Fig. 5). The one-way ANOSIM analysis revealed that, for the four habitats,the two groups could be signiWcantly distinguished(P < 0.05). The R value was low, indicating a high overlapbetween the groups in all habitats [R = 0.125 for shallowWne sands (P = 0.001), R = 0.098 for shallow mediumsands (P = 0.039), R = 0.097 for shallow muddy sands(P = 0.001) and R = 0.018 for deep Wne sands (P = 0.048)].Based on the SIMPER results (Table 2), it became clearthat the two groups were dominated by the same species,but with diVerences in their densities between the twogroups. For most species their density was higher in thesamples containing L. conchilega individuals. The averagedensity of the species was 3–10 times higher in the sampleswith L. conchilega compared to the samples withoutL. conchilega, except in deep Wne sand where the densitydiVerences were much lower (1.4 times) (Table 2).

Discussion

Distribution and habitat preferences

Lanice conchilega has a cosmopolitan distribution, as it isfound from the Arctic to the Mediterranean, in the Arabian

Table 1 First, the diVerences tested in benthic density, species rich-ness, N1-diversity and ES(50) by Mann–Whitney U test, between sam-ples with and without L. conchilega for the diVerent habitats. Second,the Spearman rank correlation between the benthic density, speciesrichness, N1-diversity and ES(50) and the density of L. conchilega forthe diVerent habitats. The number of observations (n) within each hab-itat where 236 for shallow muddy sand, 309 for shallow Wne sand, 192for deep Wne sand and 131 for shallow medium sand

Habitats Mann–Whitney U test

Spearman rank correlation

R P

Density

Shallow muddy sand P < 0.0001 0.45 P < 0.0001

Shallow Wne sand P < 0.0001 0.63 P < 0.0001

Deep Wne sand P = 0.0115 0.23 P = 0.0013

Shallow medium sand P < 0.0001 0.39 P < 0.0001

Species richness

Shallow muddy sand P < 0.0001 0.45 P < 0.0001

Shallow Wne sand P < 0.0001 0.65 P < 0.0001

Deep Wne sand P = 0.0027 0.27 P = 0.0001

Shallow medium sand P < 0.0001 0.5 P < 0.0001

N1

Shallow muddy sand P = 0.1299 0.08 P = 0.22

Shallow Wne sand P < 0.0001 0.39 P < 0.0001

Deep Wne sand P = 0.0225 0.158 P = 0.028

Shallow medium sand P = 0.0012 0.36 P < 0.0001

ES(50)

Shallow muddy sand P = 0.07 0.08 P = 0.22

Shallow Wne sand P < 0.0001 0.39 P < 0.0001

Deep Wne sand P = 0.16 0.17 P = 0.17

Shallow medium sand P < 0.0001 0.34 P < 0.0001

123

Page 6

1014 Mar Biol (2008) 154:1009–1019

Gulf and the PaciWc, from the low water neap tide markdown to 1,900 m (Hartmann-Schröder 1996). In our survey,L. conchilega was found in the entire North Sea down to adepth of 180 m (deepest record in the dataset was 380 m).This tube-building polychaete is known to live mainly insandy sediments from mud to coarse sand (Hartmann-Schröder 1996; Degraer et al. 2006), as was conWrmed bythe present study. Yet, shallow muddy and Wne sands werestrongly preferred: L. conchilega showed its highest fre-quencies of occurrence and densities in those sediments(more than 1,000 ind/m² compared to maximal 575 ind/m²in shallow medium sands). In the deeper habitats,L. conchilega was frequently encountered but only in lowabundance (max. 170 ind/m² in deep Wne sand).

The distribution of L. conchilega is mainly determinedby the sedimentology as was shown in Willems et al.(2008). This study tried to model the habitat preferences ofL. conchilega based on several types of environmental vari-ables (granulometrics, hydrodynamics, pigments and nutri-

ents), and only granulometric variables were selected in theWnal model. However, the hydrodynamics were assumed tobe more important following the study of Buhr (1976) andHeuers et al. (1998), but sedimentology and hydrodynamicswere more or less related. From the distribution map ofL. conchilega (Fig. 1), it can be deduced that the highestdensities and percentages of occurrence were observed inthe coastal areas of the North Sea (German Bight, French,Belgian and Dutch coast) and in the central part of theNorth Sea (east of the Dogger Bank). Those areas werecharacterized as zones with very high primary productionin the North Sea (McGlade 2002; Peters et al. 2005). Nextto physical factors (sediment type, Xow regime), whichmainly determine the distribution of benthic species, theavailability of food might also have a positive inXuence onthe abundance and occurrence of L. conchilega. Addition-ally, the occurrence of L. conchilega also depends on therecruitment success, which is highly variable (Van Hoey2006), but seemed to be successful in 2000–2001.

Fig. 3 a The density (with exclusion of L. conchilega) of the benthicspecies, versus the diVerent L. conchilega density classes with indica-tion of the standard deviation, and b the species richness (with exclu-sion of L. conchilega) of the benthic species, versus the diVerent

L. conchilega density classes with indication of the standard deviation.Square shallow muddy sand, rhombus shallow Wne sand, triangle deepWne sand, circle shallow medium sand

Lanice density classes (ind/m²)

0

10

20

30

40

50

60

Mac

robe

nthi

c sp

e cie

s ric

hnes

s(s

pp./0

,1m

²)

0 1-99 100-499 500-999 1000-4999 >5000

(b)

0 1-99 100-499 500-999 1000-4999 >5000

Lanice density classes (ind/m²)

0

2000

4000

6000

8000

10000

12000

14000

Mac

robe

nthi

cde

nsity

(ind/

m²)

(a)

Fig. 4 a The N1-diversity (with exclusion of L. conchilega) of the ben-thic species, versus the diVerent L. conchilega density classes withindication of the standard deviation, and b the ES(50) (with exclusionof L. conchilega) of the benthic species, versus the diVerent

L. conchilega density classes with indication of the standard deviation.Square shallow muddy sand, rhombus shallow Wne sand, triangle deepWne sand, circle shallow medium sand

Lanice density classes (ind/m²)

0

2

4

6

8

10

12

14

16

18

20

22

24

ES

(50

)

0 1-99 100-499 500-999 1000-4999 >5000

(b)

Lanice density classes (ind/m²)

0

2

4

6

8

10

12

14

16

18

20

22

24

26

28

30

N1

dive

rsity

1-99 100-499 500-999 1000-4999 >50000

(a)

123

Page 7

Mar Biol (2008) 154:1009–1019 1015

Hence, it can be concluded that L. conchilega has a widegeographical distribution and a low habitat specialization(i.e., eurytopic species), but optimally occurs in shallowWne sands and shallow muddy sands in the subtidal.

Ecological implications of the presence of Lanice conchilega

The results of the present study clearly show that L. conchi-lega has the potential to positively aVect the surroundingbenthos, which is reXected in the signiWcant and positivecorrelation between the benthic density and the density ofL. conchilega. Furthermore, the species richness increasedwith increasing density of L. conchilega. This trend was,however, not consistent: the number of species no longerincreased or even decreased after reaching a critical densityof L. conchilega (>500–1,000 ind/m²), as observed in shal-low Wne sands. A similar, but weaker trend was observedconcerning the expected number of species and indicatedan enrichment of species in L. conchilega patches. TheN1-diversity index, which takes into account species domi-nance and richness, showed similar or slightly higher val-ues in L. conchilega patches compared to patches withoutL. conchilega. These diversity patterns imply that mainlyspecies with low abundance contribute to the higher speciesrichness in samples containing L. conchilega. In otherwords, the chance to encounter a certain species increasesin L. conchilega patches, due to the higher density of a lotof benthic species in those patches (see SIMPER results,Table 2), compared to the surroundings. The observedincreases in species richness and abundances recorded inL. conchilega patches have also been discerned around the

tubes of other polychaetes (Woodin 1978; Luckenbach1986), in L. conchilega patches in intertidal areas (Zühlkeet al. 1998; Zühlke 2001; Callaway 2003a, b, 2006) andeven around artiWcial tubes (Zühlke et al. 1998; Dittmann1999).

The MDS results visualized that in every investigatedhabitat the two groups [samples with (group 1) and without(group 2) L. conchilega] consisted mostly of species fromthe same species pool. This was conWrmed by the ANOSIMand SIMPER results, where a signiWcant diVerence betweenthe two groups was found, but with a very low R value anda similar species dominance in the two groups. This indi-cates that there was a high overlap in species compositionbetween the two groups, but the density of the speciesdiVered. These results conWrmed the hypothesis that thespecies, which are aVected by L. conchilega belong to theoverall species pool of that habitat. This aspect is describedmore elaborately in Rabaut et al. (2008). It was thus dem-onstrated that L. conchilega is aVecting the benthos presentin a particular habitat in the subtidal, rather than forming itsown community (see also Zühlke et al. 1998; Dittmann1999). In this way, it seems that the eVect of L. conchilegatubes on the benthic fauna is highly dependent on the nativespecies present in the surrounding sands at any moment andon their susceptibility to tube eVects. This could be a reasonwhy species richness and diversity leveled oV in some hab-itats: almost no new species for that habitat were attracted.Lanice conchilega was considered to improve the habitatquality (e.g., habitat heterogeneity, food availability, Xowvelocity reduction), which led to increases of the densitiesof otherwise rare species in that habitat. In contrast, thedecrease in species richness and diversity from a critical

Fig. 5 Two-dimensional MDS (multi dimensional scaling) plot of similarities for the four habi-tats between samples with L. conchilega individuals (Wlled triangles) and samples without L. conchilega individuals (open triangles), with exclusion of the L. conchilega individuals

Stress: 0,21 Stress: 0,17Shallow fine sand Deep fine sand

Stress: 0,21Shallow muddy sand Stress: 0,21Shallow medium sand

123

Page 8

1016 Mar Biol (2008) 154:1009–1019

Tab

le2

SIM

PER

ana

lysi

s. T

he W

rst 2

5 sp

ecie

s of

the

SIM

PER

spe

cies

list

, whi

ch c

ontr

ibut

e to

the

diss

imil

arit

y be

twee

n gr

oup

1 (s

ampl

es w

ith

Lan

ice

conc

hile

ga) a

nd g

roup

2 (s

ampl

es w

ithou

tL

. con

chil

ega)

for

eac

h ha

bita

t with

indi

cati

on o

f th

eir

aver

age

abun

danc

e (i

nd/m

²) p

er g

roup

Sha

llow

Wne

san

dD

eep W

ne s

and

Shal

low

mud

dy s

and

Shal

low

med

ium

san

d

Spe

cies

Gro

up 1

A

v.A

bund

Gro

up 2

A

v.A

bund

Spec

ies

Gro

up 1

A

v.A

bund

Gro

up 2

A

v.A

bund

Spec

ies

Gro

up 1

A

v.A

bund

Gro

up 2

A

v.A

bund

Spe

cies

Gro

up 1

A

v.A

bund

Gro

up 2

A

v.A

bund

Spio

phan

es60

1.01

66.1

5M

yrio

chel

e49

7.47

452.

76A

mph

icte

is g

unne

ri48

4.5

222.

07Sp

ioph

anes

651.

0651

.23

Mag

elon

a24

6.42

117.

13O

wen

ia fu

sifo

rmis

335.

0623

1.43

Spio

phan

es38

8.08

102.

4G

astr

osac

cus

spin

ifer

49.3

557

.22

Bat

hypo

reia

150.

4768

.4P

ecti

nida

e27

9.66

220.

38M

ysel

la33

1.86

127.

61A

onid

es p

auci

bran

chia

ta76

.71

8.78

Tel

lina

145.

629

.2Sp

ioph

anes

252.

317

2.95

Nuc

ula

niti

dosa

125.

0823

4.05

Bat

hypo

reia

45.1

27.7

1

Ens

is d

irec

tus

263.

1823

.73

Am

phic

teis

gun

neri

138.

0510

2.19

Abr

a al

ba13

0.5

102.

69N

epht

ys c

irro

sa35

.58

51.5

1

Pho

roni

da11

5.49

124.

32P

horo

nida

100

44.9

5C

orbu

la g

ibba

75.5

92.5

9Sc

olop

los

33.6

127

.08

Ech

iuri

da34

.14

47.9

Par

amph

inom

e jeV

reys

ii42

.07

28.5

7P

horo

nida

136.

1732

.84

Spio

30.4

813

.28

Uro

thoe

pos

eido

nis

72.2

919

.02

Axi

nulu

s cr

ouli

nens

is32

.87

23.9

Pect

inid

ae10

9.12

29.0

1U

roth

oe b

revi

corn

is14

.13

12.3

5

Mys

ella

105.

0311

.44

Cha

etoz

one

33.1

24.1

Mag

elon

a49

.37

48.1

3T

elli

na9.

2620

.8

Am

phic

teis

gun

neri

71.2

26.9

3H

exac

oral

lia

29.8

912

Myr

ioch

ele

54.2

548

.79

Oph

elia

rat

hkei

17.7

78.

96

Scol

oplo

s33

.32

14.0

7T

hyas

ira

32.3

23.9

Pho

loe

balt

ica

78.8

811

.47

Gon

iade

lla

14.8

44.

73

Pec

tini

dae

45.2

12.0

5A

onid

es p

auci

bran

chia

ta22

.64

14.3

8T

elli

na62

.98

25.1

3H

ydro

ides

nor

vegi

ca20

.32

0.1

Acr

ocni

da b

rach

iata

64.9

63.

58M

ugga

wah

rber

gi38

.62

13.2

4Sc

olop

los

82.8

111

.07

Ech

iuri

da15

.71

3.99

Nep

htys

cir

rosa

15.6

118

.07

Pri

onos

pio

29.7

720

.86

Cha

etoz

one

22.2

737

.6T

hrac

ia15

.48

0.81

Cor

bula

gib

ba72

.33

7.84

Lao

nice

sar

si28

.05

17.7

1M

onta

cuta

45.6

410

.05

Mag

elon

a18

.55

13.4

2

Nem

erti

na36

.32

4.33

Nem

erti

na33

.91

27.0

5P

haxa

s pe

lluc

idus

95.5

810

.69

Spis

ula

11.7

75.

56

Abr

a al

ba45

.79

15.0

1M

ysel

la5.

9814

.19

Nep

htys

hom

berg

ii42

.83

32.7

1A

bra

alba

13.8

417

.16

Nep

htys

hom

berg

ii20

.61

12.7

7P

seud

omys

tide

s li

mba

ta21

.49

19.8

1O

wen

ia fu

sifo

rmis

31.6

314

.18

Uro

thoe

pos

eido

nis

11.2

36.

13

Hex

acor

alli

a31

.25.

99U

roth

oe e

lega

ns19

.218

.86

Har

pini

a an

tenn

aria

27.5

88.

71St

enot

hoe

mar

ina

21.0

60

Pol

inic

es19

.49

3.02

Gon

iada

12.7

618

Scal

ibre

gma

12.2

521

.72

Not

omas

tus

15.0

61.

48

Cha

etoz

one

24.9

7.24

Abr

a pr

ism

atic

a16

.09

17.1

4H

exac

oral

lia

16.9

417

.94

Gly

cera

lapi

dum

8.9

3.16

Mon

tacu

ta18

.65

5.86

Scol

oplo

s15

.63

15.5

2B

athy

pore

ia19

.58

8.18

Pis

ione

rem

ota

4.81

5.94

Oph

elia

rat

hkei

8.14

7.24

Ari

cide

a w

assi

7.36

14.6

7N

otom

astu

s18

.65

20.2

7H

exac

oral

lia

9.65

8.11

Gon

iada

17.9

75.

28T

imoc

lea

ovat

a15

.98

11.6

2E

udor

ella

em

argi

nata

204.

48P

oeci

loch

aetu

s se

rpen

s10

.71

4.59

Spio

13.6

36.

9E

udor

ella

em

argi

nata

4.48

8.48

Cha

mel

ea g

alli

na21

.17

3.88

Bra

nchi

osto

ma

lanc

eola

tum

6.06

3.11

Den

sity

of

spec

ies

at a

vera

ge W

ve ti

mes

hi

gher

in g

roup

1 c

ompa

red

to g

roup

2D

ensi

ty o

f sp

ecie

s at

ave

rage

1.4

tim

es

high

er in

gro

up 1

com

pare

d to

gro

up 2

Den

sity

of

spec

ies

at a

vera

ge th

ree

tim

es

high

er in

gro

up 1

com

pare

d to

gro

up 2

Den

sity

of

spec

ies

at a

vera

ge te

n ti

mes

hi

gher

in g

roup

1 c

ompa

red

to g

roup

2

123

Page 9

Mar Biol (2008) 154:1009–1019 1017

density of L. conchilega can be related to the competitionfor space and food in the L. conchilega patches.

It can also be argued that underlying factors (e.g., foodavailability) determine the densities of L. conchilega andtherefore also the densities of other benthic species. How-ever, the results of the present study, the studies of Rabautet al. (2008), Zühlke et al. (1998) and Callaway (2003a,2003b, 2006) clearly show that L. conchilega has the poten-tial to aVect the surrounding benthic species.

Nevertheless, diVerences in the eVect of the presenceof L. conchilega on the surrounding benthic species inthe trends of density, species richness and diversity wereobserved between the investigated soft-bottom habitats inthe North Sea. The strongest expression of the trend wasobserved in shallow Wne sands, and the weakest in deep Wnesands. The positive trend in shallow Wne sands, can beattributed to the fact (1) that Wne sands were the optimalhabitat for L. conchilega and (2) that many species canproWt from the habitat structuring capacity of L. conchilegain that environment. Shallow coastal areas were character-ized by strong dynamics and a lot of disturbance and it canbe hypothesized that L. conchilega patches create a certainstability that increases the survival of other benthic species.In deep Wne sands, the eVect of L. conchilega on benthicspecies was minimal. This might relate to the naturallyhigher benthic diversity (Künitzer et al. 1992) and thelower impact of the habitat modifying capacity of L.conchilega on the other benthic species in deep soft-bot-toms. It has to be mentioned that L. conchilega was foundin low densities, which make it impossible to predict theeVect of dense patches (not yet found in those areas).Lanice conchilega had an eVect on the density of some ben-thic species in shallow muddy sands, but no real increasesof the species richness and diversity were observed. On thecontrary, very high densities of L. conchilega (>1,000 ind/m²) had a negative eVect. The reasons for this were notclear and further investigation is needed to draw conclu-sions for this habitat. In contrary, the habitat structuringcapacity is more eVective in shallow medium sands, wherethe benthic density and diversity increased even by lowerdensities of L. conchilega. This can be attributed to the factthat the occurrence of L. conchilega creates a 3D structurein the otherwise poor sandy environment.

It can be concluded that the presence of L. conchilegahas ecological implications on the benthos in soft-bottomsediments, expressed in an increase of density and diversityof the benthos in the nearness of L. conchilega.

Lanice conchilega as ecosystem engineer?

The mechanisms responsible for the increase of the habitatquality in patches of L. conchilega can be summarized as(1) changes in the hydrodynamics, (2) increases of the

habitat stability and oxygen supply, and (3) a creation ofhabitat heterogeneity in a uniform environment.

High densities of L. conchilega can inXuence the hydro-dynamics, as has been shown in Xume experiments, inwhich dense assemblages of tubes signiWcantly reduced thecurrent velocity of the near-bottom Xow and in which nor-mal, laminar near-bottom Xow was deXected around andacross the assemblages (turbulence eVect) (Heuers et al.1998). These hydrodyamical changes have an eVect on thesedimentation of particles, detrital food (Feral 1989; Heuerset al. 1998; Seys and Musschoot 2001; Degraer et al. 2002)and on the settling of larvae and benthic species (Heuerset al. 1998; Qian 1999; Zühlke 2001; Callaway 2003a, b).The patches of L. conchilega caused sedimentation, some-times leading to elevations of the sediment surface and toan increase of the bottom roughness. These processes indi-cate that dense aggregations cause a “skimming Xow”(Morris 1955) with reduced shear stress near the bottom(Heuers et al. 1998) leading to a higher stability in the soft-bottom sediments. Tube-building species are also known tocontrol the pumping of water into and out of the bottom, by“piston pumping” in the case of L. conchilega, and provideoxygen to the adjacent sediment along the whole length ofthe tube (Forster and Graf 1995). Consequently, some spe-cies might beneWt from an improved oxygen supply in thesediment surrounding L. conchilega tubes (Callaway 2006).Due to the creation of tubes, extending out of the sediment,the habitat heterogeneity of the environment will increase,which leads to more niches for a wider variety of species.SpeciWc species will not only interact with the tubes, butsome species (predators) will be attracted by the higherfood availability.

In this way (changing hydrodynamics, increasing thehabitat stability and oxygen supply, habitat heterogeneity),L. conchilega alters the habitat characteristics and aVectsother organisms. Therefore, the species can be consideredas an ecosystem engineer (Jones et al. 1994). Laniceconchilega patches can even considered as ‘biogenic reefs’,because L. conchilega is sometimes found in patches,which rise from the sea bed (10–40 cm), in both intertidaland subtidal areas (Van Hoey 2006). ‘Biogenic reefs’ weredeWned as biological concretions that rise from the sea bedand were created by the animals themselves (Holt et al.1998). The L. conchilega reefs were formed by sedimenttrapping in dense aggregations of L. conchilega tubes,which is a diVerent mechanism than, for example, in Sabel-laria alveolata reefs (real concretions of animal tubes)(Holt et al. 1998). Lanice conchilega aggregations werealso characterized by a constant renewal of the populationdue to the high turn-over of L. conchilega (Van Hoey2006). This is diVerent from the real biogenic reef builderswhere the reef increases with settling juveniles on the olderstatic structures. However, the biogenic structures of

123

Page 10

1018 Mar Biol (2008) 154:1009–1019

L. conchilega aVect the density and species richness of thesurrounding benthos, even at low densities (few ind/m²)(this study; Callaway 2006). Although, in many cases, it isprobably more realistic to refer to these aggregations as L.conchilega beds rather than reefs, although their character-istics and eVects are likely to be very similar to those ofreally protruding ‘biogenic reefs’. Consequently, L. conchi-lega beds can be considered as important habitat structuringfeatures in the soft-bottom sediments of the North Sea. Inother words, L. conchilega patches were responsible for theincreased habitat quality in an otherwise uniform habitatand result in a higher survival of the surrounding benthicspecies.

Acknowledgments This study would have been impossible withoutthe consent to use the data of the North Sea Benthos Survey of 2000.Therefore, the authors want to thank the data contributors and aYliatedinstitutes for their data contributions: J. Aldridge (Cefas), S. Cochrane(NIVA), S. Degraer (Ghent University), N. Desroy (IFREMER), J.-M.Dewarumez (Wimereux/Lille University), G. Duineveld (NIOZ), H.Hillewaert (ILVO), I. Kröncke (Senckenberg), S. Nehring (AeT-um-weltplanung), R. Newell (MES Ltd on behalf of a dredging consor-tium), E. Oug (NIVA), T. Pohlmann (Hamburg University), E. Rachor(AWI), H. Rees (Cefas), M. Robertson (FRS), H. Rumohr (Kiel Uni-versity), J. Van Dalfsen (TNO). Besides, the following persons, M.Bergman (NIOZ), T. Bolam (Cefas), J. Craeymeersch (IMARES), G.Duineveld (NIOZ), J. Eggleton (Cefas), H. Hillewaert (ILVO), G. Irion(Senckenberg), P. Kershaw (Cefas), I. Kröncke (Senckenberg), M.Lavaleye (NIOZ), C. Mason (Cefas), E. Rachor (AWI), H. Rees (Ce-fas), H. Reiss (Gröningen University), H. Rumohr (Kiel University),M. Schratzberger (Cefas), R. Smith (Cefas), E. Vanden Berghe (VL-IZ), W. Willems (Ghent University), have contributed via the ICES SGNSBP 2000 to the development of this paper. The Wrst author acknowl-edges a grant from the Institute for the Promotion of Innovationthrough Science and Technology in Flanders (IWT, Flemish Govern-ment). The third author acknowledges an aspirant grant provided bythe FWO-Vlaanderen, Belgium. The authors also want to thank SoWeVandendriessche and the reviewers for making improvements toearlier drafts.

References

Bolam SG, Fernandes TF (2002) Dense aggregations of tube-buildingpolychaetes: response to small-scale disturbances. J Exp Mar BiolEcol 269:197–222

Buhr K-J (1976) Suspension-feeding and assimilation eYciency inLanice conchilega (Polychaeta). Mar Biol 38:373–383

Buhr K-J, Winter JE (1976) Distribution and maintenance of a Laniceconchilega association in the Weser estuary (FRG), with specialreference to the suspension-feeding behaviour of Lanice conchi-lega. In: Keegan BF, Ceidigh PO, Boaden PJS (eds) Biology ofbenthic organisms. 11th European symposium on marine biology,Gallaway. Pergamon Press, Oxford, pp 101–113

Callaway R (2003a) Juveniles stick to adults: recruitment of the tube-dwelling polychaete Lanice conchilega (Pallas, 1766). Hydrobio-logia 503:121–130

Callaway R (2003b) Long-term eVects of imitation polychaete tubes onbenthic fauna: they anchor Mytilus edulis (L.) banks. J Exp MarBiol Ecol 283:115–132

Callaway R (2006) Tube worms promote community change. MarEcol Prog Ser 308:49–60

Carey AD (1987) Sedimentological eVects and palaeoecological impli-cations of the tube-building polychaete Lanice conchilega (Pal-las). Sedimentology 34:49–66

Clarke KR, Warwick RM (2001) Change in marine communities: anapproach to statistical analysis and interpretation, 2nd edn. PRIM-ER-E Ltd, Plymouth Marine Laboratory, UK

Conover WJ (1971) Practical non parametric statistics. Wiley, NewYork

Degraer S, Van Lancker V, Moerkerke G, Van Hoey G, Vanstaen K,Vincx M, Henriet J-P (2002) Evaluation of the ecological value ofthe foreshore: habitat-model and macrobenthic side-scan sonarinterpretation: extension along the Belgian Coastal Zone. Techni-cal report. Ministry of the Flemish Community, Environment andInfrastructure. Department Waterways and Marine AVairsAdministration, Coastal Waterways, Belgium

Degraer S, Wittoeck J, Appeltans W, Cooreman K, Deprez T, Hillewa-ert H, Hostens K, Mees J, Vanden Berghe E, Vincx M (2006) Themacrobenhos atlas from the Belgian Part of the North Sea. Feder-aal Wetenschapsbeleid D/2005/1191/5, pp 164

Dittmann S (1999) Biotic interactions in a Lanice conchilega-domi-nated tidal Xat. In: Dittmann S (ed) The Wadden Sea ecosystem.Springer, Berlin, pp 153–162

Eckman JE (1983) Hydrodynamic processes aVecting benthic recruit-ment. Limnol Oceanogr 28:241–257

Eckman JE, Nowell ARM, Jumars PA (1981) Sediment destabilizationby animal tubes. J Mar Res 39:361–374

Feral P (1989) InXuence des populations de Lanice conchilega (Pallas)(Annelida, Polychaeta) sur la sedimentation sableuse intertidalede deux plages bas-nomandes (France). Bull Soc Geol Fr 8:1193–1200

Forster S, Graf G (1995) Impact of irrigation on oxygen Xux into thesediment: intermittent pumping by Callianassa subterraneai and‘piston-pumping’ by Lanice conchilega. Mar Biol 123:335–346

Hartmann-Schröder G (1996) Annelida, Borstenwürmer, Polychae-ta—Tierwelt Deutschlands, Teil 58. Veb Gustav Fischer Verlag,Jena, Hamburg

Heuers J, Jaklin S, Zühlke R, Dittmann S, Günther C-P, HildenbrandtH, Grimm V (1998) A model on the distribution and abundanceof the tube-building polychaete Lanice conchilega (Pallas, 1766)in the intertidal of the Wadden Sea. Verh Ges Okol 28:207–215

Hild A, Günter C-P (1999) Ecosystem engineers: Mytilus edulis andLanice conchilega. In: Dittmann S (ed) The Wadden Sea ecosys-tem. Springer, Berlin, pp 43–49

Hill MO (1973) Diversity and evenness: a unifying notation and itsconsequences. Ecology 54:427–432

Holt TJ, Rees EI, Hawkins SJ, Seed R (1998) Biogenic reefs, vol IX.An overview of dynamic and sensitivity characteristics for con-servation management of marine SACs. Technical report, Scot-tish Association for Marine Science (UK Marine SACs Project)

Hurlbert SH (1971) The non concept of species diversity: a critique andalternative parameters. Ecology 52:577–586

ICES (2004) Report of the study group on North Sea benthos project2000 (SGNSBP), 29 March–1 April 2004, Wilhelmshaven,Germany. ICES report CM 2004/E:05

Jones GJ, Lawton JH, Shachak M (1994) Organisms as ecosystemengineers. OIKOS 69:373–386

Künitzer A, Duineveld GCA, Basford D, Dewarumez JM, Dörjes J,Eleftheriou A, Heip C, Herman PMJ, Kingston P, Niermann U,Rumohr H, de Wilde PAWJ (1992) The benthic infauna of theNorth Sea: species distribution and assemblages. ICES J Mar Sci49:127–143

Luckenbach MW (1986) Sediment stability around animal tubes: theroles of hydrodynamics processes and biotic activity. LimnolOceanogr 31(4):779–787

McGlade JM (2002) The North Sea large marine ecosystem. In: Sher-man K, Skjoldal HR (eds) Large marine ecosystems of the North

123

Page 11

Mar Biol (2008) 154:1009–1019 1019

Atlantic—changing states and sustainability, vol Forthcomingvolume. Elsevier, Amsterdam, pp 339–412

Morris HM (1955) A new concept of Xow in rough conduits. Trans AmSoc Civil Eng 120:373–398

Peters SWM, Eleveld M, Pasterkamp R, van der Woerd H, DeVolderM, Jans S, Park Y, Ruddick K, Block T, Brockmann C, DoerVerR, Krasemann H, Schoenfeld W, Jorgenson PV, Tilstone G,Moore G, Sorensen K, Hokedal J, Aas E (2005) Atlas of chloro-phyll—a concentration for the North Sea based on MERIS imag-ery of 2003. Technical report. Vrije Universiteit Amsterdam, TheNetherlands

Qian PY (1999) Larval settlement of polychaetes. Hydrobiology402:239–253

Rabaut M, Guilini K, Van Hoey G, Vincx M, Degraer S (2008) A bio-engineered soft-bottom environment: the impact of Laniceconchilega on the benthic species-speciWc densities and commu-nity structure. Estuar Coast Shelf Sci 75:525–536

Rees H, Cochrane S, Craeymeersch J, de Kluijver M, Degraer S, Des-roy N, Dewarumez J-M, Duineveld G, Essink K, Hillewaert H,Kilbride R, Kröncke I, Nehmer P, Rachor E, Reiss H, RobertsonM, Rumohr H, Vanden Berghe E, Van Hoey G (2002) The NorthSea benthos project: planning, management and objectives. ICESreport, C.M. L(9)

Rees EIS, Bergmann M, Galanidi M, Hinz H, Shucksmith R, KaiserMJ (2005) An enriched Chaetopterus tube mat biotope in the east-ern English Channel. J Mar Biol Assoc UK 85:323–326

Rees HL, Eggleton JD, Rachor E, Vanden Berghe E (eds) (2007)Structure and dynamics of the North Sea benthos. ICES coopera-tive research report no. 288, p 258

Ropert M, Dauvin J-C (2000) Renewal and accumulation of a Laniceconchilega (Pallas) population in the bay des Veys, western Bayof Seine. Oceanol Acta 23(4):529–546

Seys L, Musschoot T (2001) Lanice conchilega: het leven en de vo-edingswijze van een kokerbouwende polychaet, report. Univer-sity of Ghent , Belgium

Van Hoey G (2006) Spatio-temporal variability within the macroben-thic Abra alba community, with emphasis on the structuring roleof Lanice conchilega. Ph.D. thesis. University of Ghent, Belgium

Van Hoey G, Vincx M, Degraer S (2006) Some recommendations foran accurate estimation of Lanice conchilega density based on tubecounts. Helgoland Mar Res 60:317–321

Willems W, Goethals P, Van den Eynde D, Van Hoey G, Van LanckerV, Verfaille E, Vincx M, Degraer S (2008) Where is the worm?Predictive modelling of the habitat preferences of the tube-build-ing polychaete Lanice conchilega (Pallas, 1766). Ecol Model212:74–79

Woodin SA (1978) Refuges, disturbance, and community structure:a marine soft-bottom example. Ecology 59:274–284

Ziegelmeier E (1952) Bedachtungen über den Röhrenbau von Laniceconchilega (Pallas) im experiment und am natürlichen standort.Helgoland Meeresuntersuchingen 4:107–129

Zühlke R (2001) Polychaete tubes create ephemeral community pat-terns: Lanice conchilega (Pallas, 1766) associations studied oversix years. J Sea Res 46:261–272

Zühlke R, Blome D, Heinz van Bernem K, Dittmann S (1998) EVectsof the tube-building polychaete Lanice conchilega (Pallas) onbenthic macrofauna and nematodes in an intertidal sandXat. Senc-kenb Marit 29(1):131–138

123