change in biological traits and community structure of macroinvertebrates through primary succession...

BioOne sees sustainable scholarly publishing as an inherently collaborative enterprise connecting authors, nonprofit publishers, academic institutions, researchlibraries, and research funders in the common goal of maximizing access to critical research.

Change in biological traits and community structure of macroinvertebratesthrough primary succession in a man-made Swedish wetlandAuthor(s): Albert Ruhí, Jan Herrmann, Stéphanie Gascón, Jordi Sala, Joja Geijer, and Dani BoixSource: Freshwater Science, 31(1):22-37. 2012.Published By: The Society for Freshwater ScienceDOI: http://dx.doi.org/10.1899/11-018.1URL: http://www.bioone.org/doi/full/10.1899/11-018.1

BioOne (www.bioone.org) is a nonprofit, online aggregation of core research in the biological, ecological, andenvironmental sciences. BioOne provides a sustainable online platform for over 170 journals and books publishedby nonprofit societies, associations, museums, institutions, and presses.

Your use of this PDF, the BioOne Web site, and all posted and associated content indicates your acceptance ofBioOne’s Terms of Use, available at www.bioone.org/page/terms_of_use.

Usage of BioOne content is strictly limited to personal, educational, and non-commercial use. Commercial inquiriesor rights and permissions requests should be directed to the individual publisher as copyright holder.

http://dx.doi.org/10.1899/11-018.1

http://www.bioone.org/doi/full/10.1899/11-018.1

http://www.bioone.org

http://www.bioone.org/page/terms_of_use

Change in biological traits and community structure ofmacroinvertebrates through primary succession in a man-made

Swedish wetland

Albert Ruhı1,3, Jan Herrmann2,4, Stephanie Gascon1,5, Jordi Sala1,6,Joja Geijer2,7, AND Dani Boix1,8

1 Institute of Aquatic Ecology, University of Girona, Girona, Catalonia, Spain2 School of Natural Sciences, Linnaeus University, Kalmar, Sweden

Abstract. We investigated the successional development of a benthic macrofaunal community in KalmarDamme, a man-made wetland in southeastern Sweden, over a 13-y period after construction (1997–2009).We assessed primary succession by monitoring changes in species composition, biological traits, andcommunity structure. Predictable successional changes occurred, and the structure of the community wascomplex at the advanced phase. Three successional phases were observed, each with typifying species. Atthe advanced phase, dominant biological traits shifted to favor animals with longer life-cycle durations,aerial dispersal strategies, and certain feeding types (i.e., filter-feeders, fine sediment collectors, predators,and parasites). We selected the Coleoptera assemblage to represent the macroinvertebrate community inan evaluation of the extent to which man-made wetlands in different age classes (,3 y and §10 y) weresimilar to natural wetlands (.50 y). No significant differences in Coleoptera assemblages werefound between natural wetlands and man-made wetlands §10 y. However, man-made wetlands §10 yhad significantly higher coleopteran species richness than natural and man-made wetlands ,3 y becauseof the convergence of pioneering species persisting from early successional stages with representativespecialized taxa from natural wetlands. Our results suggest that 10 y may be sufficient time for themacrofaunal communities of man-made wetlands to resemble those of natural wetlands.

Key words: colonization, succession, man-made, wetlands, invertebrates, aquatic Coleoptera.

Succession can involve large numbers of speciesand important changes in community structure(Smith 1928, Odum 1960, Gutierrez and Fey 1980).These changes usually are associated with differentsuccessional phases, each with its own stable point,moving in a particular direction (Allen et al. 1977,Niell 1981, Legendre et al. 1985). Succession has beenstudied during the process of recovery after majordisturbances, such as fires, hurricanes, or droughts(Force 1981, Smith 1982, Kaufman 1983). However,newly created aquatic habitats also are optimalsystems in which to study and test successionaldynamics. Research on man-made aquatic systemshas expanded over the last decade (Oertli et al. 2005)

because they tend to be easy to monitor and regulate.Nevertheless, most studies done in newly createdwetlands are short-term (,3 y), which is a criticallimitation when assessing ecological functioning(Zedler and Callaway 1999). For example, studiesfocused on the mitigation of wetland loss to conserveamphibians have shown that detailed, qualitative,and long-term observations are essential for correctevaluation of their conservation value (Pechmannet al. 2001, Petranka et al. 2003, 2007). Furthermore,short-term approaches are restrictive when studyinginvertebrate communities in these systems becausethe initial years are dominated by pioneer assem-blages (Bloechl et al. 2010). Hence, these studies mayassess initial colonization successfully, but usually donot detect later changes in assemblage structure andorganization (Barnes 1983, Friday 1987). A smallnumber of pioneer colonizers may explain the maindifferences between assemblages in created vs naturalwater bodies (Ruhı et al. 2009). Moreover, for someparticularly species-rich taxa with specialized feeding

3 E-mail addresses: [email protected] [email protected] [email protected] [email protected] [email protected] [email protected]

Freshwater Science, 2012, 31(1):22–37’ 2012 by Society for Freshwater ScienceDOI: 10.1899/11-018.1Published online: 6 December 2011

22

structures (e.g., Coleoptera; Lundkvist et al. 2001,Thiere et al. 2009), succession may continue for .10 y(Fairchild et al. 2000).

Increases in biomass and diversity during succes-sion were hypothesized in earlier general ecologypapers (Margalef 1968, Odum 1971, Gutierrez and Fey1980, Legendre et al. 1985), but unimodal (hump-shaped) changes in biomass and diversity over timehave been described when the scale of measurementwas long enough to include several stages ofsuccession (Guo 2003). In freshwater systems, abun-dance/density and taxonomic richness of macroin-vertebrates tend to increase during the first years ofsuccession (e.g., in new reservoirs; Voshell andSimmons 1984, Bass 1992), but asymptotic patternstend to appear when studies are longer (e.g., 20-ystudy of a deglaciated stream; Milner and Robertson2010). Despite the paucity of published research innewly created wetlands, available studies suggest thatnew taxonomic groups tend to arrive in a directionalsuccessional process (Barnes 1983, Herrmann et al.2000, Solimini et al. 2003, Ruhı et al. 2009). Thus,changes in community structure should be reflectedby measures of taxonomic distinctness (Clarke andWarwick 2001a). Successional changes usually havebeen described taxonomically, but they also shouldbe apparent in analyses of functional characteristics ofthe organisms (Hooper et al. 2005, Verberk et al. 2008).Batzer and Wissinger (1996) pointed out that changesin species composition of establishing communitiescould be a consequence of the favored life-historystrategies (i.e., dispersal capacity, desiccation resis-tance, trophic specialization). The hypotheses andrationale proposed by Townsend and Hildrew (1994)and Bonada et al. (2007) relating the presence of traitsfavoring resistance/resilience to ecological distur-bance in streams also could be relevant to under-standing succession in new wetlands.

Our study goal was to describe the main changes ofa macrofaunal community in a man-made wetlandand to analyze whether the community eventuallyresembled that of a natural one. The biological traitsand community structure of the macroinvertebratecommunity of a newly created wetland were moni-tored for 13 y. We predicted that: 1) taxa with shortlife cycles and active dispersal mechanisms woulddominate immediately after flooding and during theinitial phases of succession, whereas taxa with longlife cycles and passive dispersal mechanisms wouldincrease in importance during the later phases; 2) thedominant feeding type would shift during successionfrom taxa that consume suspended matter andprimary production to those feeding on prey, hosts,and abundant fine particulate organic matter; 3)

species richness and evenness would increase duringearly successional stages; 4) species richness andevenness would decrease and cumulative richnessand numerical abundance would reach asymptotesduring later successional stages; and 5) averagetaxonomic distinctness would be higher and variationin taxonomic distinctness would be lower in late thanearly stages during succession (Table 1).

We also compared the macroinvertebrate commu-nity in the study wetland with those in other man-made (,3 y and §10 y old) and natural wetlands. Weused aquatic Coleoptera as a surrogate for the entireinsect community because coleopterans have beenidentified as a potential biodiversity indicator incontinental European aquatic ecosystems (Fosteret al. 1989, Sanchez-Fernandez et al. 2006). Weexpected the composition of the Coleoptera in oldman-made wetlands to converge on composition innatural wetlands subsequent to the arrival of late-colonist specialist taxa to old man-made wetlands. Incontrast, we expected the community in young man-made wetlands to be dominated by pioneeringgeneralist taxa and, therefore, different from thecommunity in old man-made wetlands. Thus, wepredicted higher species richness in natural and oldman-made wetlands than in young ones.

Methods

Study area

Macroinvertebrate study.—The macroinvertebratestudy was carried out at Kalmar Damme (Fig. 1A,B), which is on the southeastern coast of Sweden. Thissite consists of a man-made freshwater wetland thatwas constructed (i.e., excavated and flooded) in late1996. This freshwater system is fed by the Tornebystream and is situated 2 km from the stream outletinto the Baltic Sea (Fig. 1C). The drainage basin(48 km2) of the wetland is mainly agricultural, butalso includes large forested areas, a few urban areas,and the regional Kalmar Airport. The wetland consistsof several connected elongated waterbodies ,2 m indepth that are vegetated with belts of common reed(Phragmites australis), bulrush (Typha latifolia), commonwater plantain (Alisma plantago-aquatica), and branchedbur-reed (Sparganium erectum). Some submergedplants were planted immediately after constructionof the wetland. Initially, the most abundant sub-merged species were the Canadian waterweed (Elodeacanadensis) and the spiked Eurasian water milfoil(Myriophyllum spicatum). However, both species be-came less abundant in later years. In addition, severalspecies of fish were present from the time of the initialflooding.

2012] PRIMARY SUCCESSION IN MAN-MADE WETLANDS 23

TA

BL

E1.

Hy

po

thes

esan

dra

tio

nal

efo

rco

mm

un

ity

stru

ctu

rean

db

iolo

gic

altr

aits

con

sid

ered

inth

em

acro

inv

erte

bra

test

ud

y.

+an

d2

ind

icat

eex

pec

ted

hig

han

dlo

wv

alu

es,

resp

ecti

vel

y.

Her

b=

her

biv

oro

us.

Tra

itT

rait

stat

eS

ucc

essi

on

earl

yR

adv

ance

dR

atio

nal

eR

efer

ence

s

Bio

log

ical

trai

t

No

.re

pro

du

ctiv

ecy

cles

/y

.1

+2

Stab

leen

vir

onm

ents

shou

ldfa

vor

spec

ies

wit

hlo

ng

erlif

ecy

cle,

lon

glif

esp

ans,

and

late

mat

uri

tyT

ow

nse

nd

and

Hil

dre

w19

94,

Bo

nad

aet

al.

2007

ƒ1

2+

Dis

per

sal

stra

teg

ies

Aq

uat

ic2

+A

rriv

alp

rob

abil

itie

so

faq

uat

icd

isp

erse

rssh

ou

ldin

crea

sew

ith

tim

e;ae

rial

acti

ve

dis

per

sal

stra

teg

ies

fav

ore

dd

uri

ng

the

1stp

has

eso

fa

hy

dro

per

iod

;aer

ial

pas

siv

est

rate

gie

sre

qu

ire

mo

reti

me

Dan

ell

and

Sjo

ber

g19

82,

Bas

s19

92,

Bat

zer

and

Wis

sin

ger

1996

,B

on

ada

etal

.20

07A

eria

lac

tiv

e+

2A

eria

lp

assi

ve

2+

Fee

din

gst

rate

gie

sP

lan

tea

ters

(scr

aper

s+

her

b.

pie

rcer

s)+

2R

esp

on

seto

earl

yav

aila

bil

ity

of

per

iph

yto

nan

dm

acro

ph

yte

sB

on

ada

etal

.20

07,

Ru

hı

etal

.20

09F

ilte

r-fe

eder

s+

2R

esp

on

seto

earl

yav

aila

bil

ity

of

sest

on

and

pla

nk

ton

Vo

shel

lan

dS

imm

on

s19

84

Sh

red

der

s+

2R

esp

on

seto

earl

yav

aila

bil

ity

of

mac

rop

hy

tes

and

larg

ete

rres

tria

lli

tter

Vo

shel

lan

dS

imm

on

s19

84

Fin

ese

dim

ent

coll

ecto

rs2

+R

esp

on

seto

the

incr

ease

of

fin

ep

arti

cula

teo

rgan

icm

atte

rin

the

sed

imen

tsB

on

ada

etal

.20

07

Pre

dat

ors

(en

gu

lfer

s+

anim

alp

ierc

ers)

2+

Res

po

nse

toin

crea

sed

abu

nd

ance

of

pre

y;

mo

reim

po

rtan

tth

anth

ein

itia

ld

om

inan

ceo

fae

rial

colo

niz

ing

pre

dat

ors

(e.g

.,d

ivin

gb

eetl

esan

dh

emip

tera

ns)

Bat

zer

and

Wis

sin

ger

1996

,S

chn

eid

eran

dF

rost

1996

,B

loec

hl

etal

.20

10

Par

asit

es2

+R

esp

on

seto

the

incr

ease

of

the

abu

nd

ance

of

po

ten

tial

ho

sts

Co

mm

un

ity

par

amet

er

Nu

mer

ical

abu

nd

ance

2+

Pro

du

ctio

no

fth

esy

stem

resu

lts

inin

crea

sed

nu

mer

ical

abu

nd

ance

of

mac

roin

ver

teb

rate

sw

ith

am

axim

um

du

rin

gad

van

ced

succ

essi

on

alp

has

e

Mar

gal

ef19

68,

Od

um

1971

,G

uti

erre

zan

dF

ey19

80,

Leg

end

reet

al.

1985

,G

uo

2003

Av

erag

eta

xo

no

mic

dis

tin

ctn

ess

2+

Incr

ease

dta

xon

omic

div

ersi

tyan

dev

enn

ess

assu

cces

sion

pro

ceed

s;ar

riv

alof

new

and

taxo

nom

ical

lyd

iver

sesp

ecie

ssh

ould

be

spre

adac

ross

sev

eral

taxo

nom

icg

rou

ps

and

rich

nes

sof

afe

wsp

ecie

s-ri

chp

ion

eer

gro

up

ssh

ould

dec

reas

eov

erti

me

Infe

rred

fro

mp

ub

lish

edfa

un

alli

sts

in:

Bar

nes

1983

,C

hri

stm

anan

dV

osh

ell

1993

,H

errm

ann

etal

.20

00,

Ho

van

dW

alse

ng

2003

,So

lim

ini

etal

.20

03,

Lan

cast

eret

al.

2004

,R

uh

ıet

al.

2009

Var

iati

on

inta

xo

no

mic

dis

tin

ctn

ess

+2

Pie

lou

’sev

enn

ess

ind

ex2

+2

Ev

enn

ess

and

rare

fied

spec

ies

rich

nes

ssh

ould

atta

inm

axim

um

val

ues

du

rin

gth

ein

term

edia

tep

has

esof

succ

essi

on;a

rriv

alof

new

spec

ies

shou

ldbe

limite

dto

war

dan

asym

pto

tein

am

atu

reco

mm

un

ity,

wh

ere

ah

igh

pro

por

tion

ofth

eec

olog

ical

nic

hes

are

alre

ady

occu

pie

d

Mar

gal

ef19

68,

Bat

zer

and

Wis

sin

ger

1996

,G

uo

2003

,M

iln

eran

dR

ob

erts

on

2010

Rar

efie

dsp

ecie

sri

chn

ess

2+

2

Cu

mu

lati

ve

spec

ies

rich

nes

s2

+

24 A. RUHI ET AL. [Volume 31

FIG. 1. Study area (A), locations of the study sites (B), and a scheme of water movements and sampling locations (1–3) in KalmarDamme (KD) (C). Black circles in (B): Kalmar Damme (KD) and the 2 man-made wetlands §10 y, Ljungbyholm (LJ) and Resby (RE);white circles: man-made wetlands ,3 y, Overstatorp (OV), Rockneby (RO), and Tingby (TI); squares: selected natural wetlands(.50 y without major human disturbances) Arby Damme (AD), Bjorkerumsdammen (BJ), Bollmossen (BO), Brudhyltetjarn (BR),Glomminge (GL), Igelmossen (IG), Kolboda Hagby (KO), Lenstamossen (LE), Nytorpskarret (NY), and Tyska Bruket (TB).


The system was designed and built to reduce Nlevels to 50% of the annual amount of urea-Ndischarged at the airport. This objective was met bythe 2nd y (1998) after construction (based on monthlydata from a wetland monitoring program run by theKalmar Airport Authority). A 2nd design objectivewas to promote aquatic biodiversity (Herrmann et al.2000).

Coleoptera study.—We did the Coleoptera study in 6man-made and 10 natural wetlands, which wereselected as reference sites (Fig. 1B). To the best ofour knowledge, all wetlands were permanent. Threeman-made wetlands (Kalmar Damme, Ljungbyholm,and Resby) were §10 y old and 3 (Overstatorp,Rockneby, and Tingby) were 1 to 3 y old at the time ofsampling. We chose 10 natural wetlands with differ-ent levels of vegetation structure and substratum typefrom the mainland and the adjacent island of Oland toencompass a broad range of natural variability. The 10natural wetlands were Arby Damme, Bjorkerums-dammen, Bollmossen, Brudhyltetjarn, Glomminge,Igelmossen, Kolboda Hagby, Lenstamossen, Nytorps-karret, and Tyska Bruket. Six of these wetlands weredefinitely natural in origin, but the origin of Bjorker-umsdammen, Bollmossen, Igelmossen, and TyskaBruket was unknown. However, none of the 10wetlands had been affected by human activity for§50 y. See Table 2 for information about the creationof the man-made systems and the characteristics of all16 study sites.

Sampling and sample processing

Macroinvertebrate study.—We visited KalmarDamme during 5 periods from early 1997 to late2009. Period I was the 1st y (1997), period II the 2nd y(1998), period III the 3rd y (1999), period IV the 7th y(2003), and period V the 13th y (2009) after construc-tion. These 5 periods were intended to represent theprogressive development of the changing macroin-vertebrate community. In each period, we visited thewetland in spring (March–April), summer (June–July), and autumn (September–October) and sampledmacroinvertebrates at 3 fixed sites (Fig. 1B). Duringeach period, personnel from the Kalmar AirportAuthority did monthly analyses of dissolved inor-ganic N (DIN = NH4

+ + NO22 + NO32), and soluble

reactive P (SRP = PO432). Monthly meteorological

variables were obtained from the Swedish Meteoro-logical and Hydrological Institute.

We sampled aquatic macrofauna with a Surbersampler (33 3 33 cm) equipped with an extra 15-cm-high front metal frame and 0.5-mm mesh on bothparts. We collected 5 subsamples at each site over a

TA

BL

E2.

Ch

arac

teri

stic

so

fw

etla

nd

ssa

mp

led

du

rin

gth

eC

ole

op

tera

stu

dy

.M

ax=

max

imu

m.

Ty

pe

Sit

e(a

bb

rev

iati

on

)C

reat

edS

amp

led

Max

area

(m2)

Max

dep

th(m

)

Em

erg

edv

eget

atio

nco

ver

Su

bm

erg

edv

eget

atio

nco

ver

Su

bst

rate

typ

eF

ish

Man

-mad

e§

10y

Kal

mar

Dam

me

(KD

)19

9620

06–2

008

180,

000

1.8

50%

80%

Det

ritu

sP

rese

nt

Lju

ng

by

ho

lm(L

J)19

9820

08–2

009

300

2.0

20%

20%

Co

arse

gra

vel

Ab

sen

tR

esb

y(R

E)

1998

2008

–200

950

001.

515

%50

%C

lay

,d

etri

tus

Un

kn

ow

nM

an-m

ade

,3

yO

ver

stat

orp

(OV

)20

0520

06–2

008

20,0

00.

3.0

25%

25%

San

dP

rese

nt

Ro

ckn

eby

(RO

)20

0220

03–2

005

160

1.0

70%

70%

Cla

y,

san

dA

bse

nt

Tin

gb

y(T

I)20

0320

04–2

006

1000

.2.

510

%10

%S

and

,d

etri

tus

Pre

sen

tN

atu

ral

.50

yA

rby

Dam

me

(AD

)–

2004

–200

580

2.0

10%

0%C

lay

Ab

sen

tB

jork

eru

msd

amm

en(B

J)–

2006

–200

810

01.

510

%50

%S

and

,cl

ayP

rese

nt

Bo

llm

oss

en(B

O)

–20

05–2

006

5000

2.0

70%

50%

Det

ritu

sP

rese

nt

Bru

dh

ylt

etja

rn(B

R)

–20

07–2

009

3000

.2.

015

%15

%P

eat

Pre

sen

tG

lom

min

ge

(GL

)–

2007

–200

912

00.

550

%0%

Cla

yA

bse

nt

Igel

mo

ssen

(IG

)–

2006

–200

810

,000

1.5

10%

20%

San

d,

gra

vel

Pre

sen

tK

olb

od

aH

agb

y(K

O)

–20

06–2

008

2500

0.5

50%

0%D

etri

tus

Ab

sen

tL

enst

amo

ssen

(LE

)–

2004

–200

510

,000

2.0

70%

20%

Det

ritu

sP

rese

nt

Ny

torp

skar

ret

(NY

)–

2007

–200

924

000.

570

%0%

Pea

tA

bse

nt

Ty

ska

Bru

ket

(TB

)–

2004

–200

572

0.5

50%

0%D

etri

tus

Ab

sen

t


range of depths (10–40 cm). We preserved samples insitu in 70% ethanol. We sorted samples and identifiedmacroinvertebrates in the laboratory. Identificationwas carried out mostly to species level, except forOligochaeta and Chironomidae.

Coleoptera study.—We visited each of the 16 wet-lands 4 times with §1 spring and §1 autumnsampling visit over 2–3 consecutive years between2003 and 2009. We designed the sampling procedureto capture the maximum number of aquatic Coleop-tera. We sampled with a dip net (0.5-mm mesh size).The maximum sampling depth was 1 m and samplescovered all microhabitats. We applied a basic sam-pling effort of 10 min with a sampling rationale towork the site until no new morphospecies werefound. We devoted equal effort to recording commonand rarer species (Eyre et al. 1986). We passedsamples through a 0.5-mm mesh, transferred themto 70% ethanol, and sorted and identified coleopter-ans to species level in the laboratory.

Data analysis

Macroinvertebrate study.—We assessed taxonomicchanges over time with analysis of similarities(ANOSIM). This test operates on a resemblancematrix and is similar to a standard univariate analysisof variance (ANOVA), but requires neither normalitynor homoscedasticity of data. We compared speciesabundance matrices among years. The analysisreturns a global R and a p-value expressed as apercentage. When R is negative or close to 0,similarities within groups (samples within the sameperiod) and among groups (samples in differentperiods) are equivalent. In contrast, when R valuesapproach 1, samples are more similar within thanamong groups, and groups of samples from differentperiods are differentiated (Clarke and Warwick2001b). We built the distance matrix with the Bray–Curtis similarity measure, based on a sample 3

species matrix of abundance data, originally stan-dardized and 4!(x)-transformed. We used similarity ofpercentages analysis (SIMPER) to identify the speciesthat characterized the community of each period. Weordered species from more to less contribution to thetotal within-period similarity and established a cut-offfor a global contribution of 70%. Within the selectedspecies, we considered as typifying those thatcontributed a large proportion to the total within-period similarity and were found at consistentabundance among samples in that period (criterion:mean contribution/SD . 2). A full description ofSIMPER analysis is provided by Clarke (1993). Weused 2-dimensional nonparametric multidimensional

scaling (MDS) to visualize community patterns (99restarts, 0.01 minimum stress, Kruskal fit scheme 1).We standardized and 4!(x)-transformed abundancesand used Bray–Curtis distances. Groups of sampleswith several thresholds of similarity (12.5%, 25%,50%) were calculated by means of clustering (com-plete linkage), and the best threshold was representedover the MDS diagram.

We analyzed 3 categories of biological traits fromthose put forward by Tachet et al. (2002): reproductivecycles/y, type of dispersion, and feeding strategy. Weused affinity scores (Tachet et al. 2002) for each genusto evaluate 11 traits within these categories. Affinityscores ranged from 0 (i.e., no affinity) to 5 (i.e., highaffinity). This fuzzy coding technique (Chevenet et al.1994) allowed us to build a biological traits abundancematrix. We tested for changes in biological traits overtime with generalized linear mixed models (GLMMs).We used GLMM to overcome autocorrelation be-tween observations (more similarity is expectedamong sites within a visit than among different visits)and nested data structures (each period encompassed3 visits). We built a model in which the responsevariable was the modality of each biological trait(abundance/sample), the fixed explanatory variablewas a nested structure (period factor [year] was fixedand visit factor [visits within each period] was nestedwithin the period), and sampling site was random toovercome spatial pseudoreplication. In mixed models,random effects account for pseudoreplication bymodelling the covariance structure introduced in therandom part of the model (Crawley 2002). We startedwith a model without interactions. Validation of themodel showed no significant patterns in the residuals,so we followed the current consensus (Zuur et al.2009) and decided not to include interactions thatwould have improved the model at the expense ofincreasing the difficulty of biological interpretation ofthe results. We applied the Bonferroni procedure tocompensate for the effects of multiple tests on Type Ierrors (adjusted p = 0.0045). We used post hoc Scheffetests to examine pairwise differences among periods.

In addition, we selected 6 measures of communitystructure for analysis: numerical abundance (N),average taxonomic distinctness (D+), variation intaxonomic distinctness (L+), Pielou’s evenness index(J’), rarefied species richness (S1), and cumulativespecies richness (S2). Biodiversity analyses are oftenbased on species richness alone, but taxonomicrelatedness also should be considered because assem-blages composed only of taxonomically relatedspecies should be regarded as less diverse than thosethat host more distantly related species (Warwick andClarke 1995, Abellan et al. 2006). Average taxonomic


distinctness (D+) represents the average path length inthe phylogenetic tree connecting 2 random species ofthose collected, whereas variation in taxonomicdistinctness (L+) measures the variance in pairwiselengths between each pair of species and reflects theunevenness of the taxonomic tree (Clarke and War-wick 2001a, Clarke and Gorley 2006). We took intoaccount species, genus, family, order, class, subphy-lum, and phylum and used the same branch length toweight each taxonomic level. Abundance was variableamong samples, so we calculated richness using ararefaction procedure with the function rarefy deriv-ing from the vegan library (Oksanen et al. 2010). Thisfunction, which was inspired by Heck et al. (1975),calculates the estimated species richness for a numberof individuals by simulating random draws of a fixednumber of individuals based on the least abundantsample. S1 is not an estimate of the total communityrichness, but it permits an unbiased comparisonbetween samples of unequal abundance. Thus, it canbe regarded as a measure of diversity because, for agiven number of taxa, it will increase with theevenness of species abundance (Lods-Crozet andCastella 2009). We calculated S2 by considering allspecies present at least once within each period andaccumulating species from one period to the next. Wetested community variables for changes over timewith GLMM and used the same model design as forbiological traits. We explored but did not test S2statistically because of the lack of replicates. Welog(x)-transformed N to ensure a better fit of errors toa normal distribution. For each variable, we usedScheffe post hoc tests to compare values among the 5periods. We used general linear models (GLMs) withperiod as a fixed factor to examine differences inenvironmental variables (air temperature, precipita-tion, and water-chemistry variables) among periods.Because of the poor taxonomical resolution ofOligochaeta and Chironomidae, we took these groupsinto account only when their inclusion did not createa bias caused by taxonomic lumping. In the case ofbiological traits, we included these groups only in theanalyses of reproductive cycles per year and dispers-al. In the case of community variables, we includedthese groups only in determination of N.

Coleoptera study.—We analyzed coleopteran assem-blages based on a single matrix of 186 species(variables) and 64 samples (16 sites 3 4 visits/site).We analyzed 2 factors: 1) wetland type (man-madewetlands ,3 y, man-made wetlands §10 y, andnatural wetlands) and 2) wetland site (each of the 16wetlands). We used ANOSIM to test for differences inColeoptera composition across wetland types and sites.We used a 2-way nested layout in which the 2 factors

were hierarchic (site factor nested within type factor).We built a rectangular matrix based on originallystandardized and posteriorly 4!(x)-transformed datausing the Bray–Curtis similarity measure. We usedSIMPER with wetland type as factor to detect whichspecies contributed .1% to the similarity of eachwetland type. We calculated Coleoptera richness (Sc)by summing all recorded species for each wetland in agiven visit. We examined varability in Sc with a GLMwith type as a fixed factor. To assess the statisticalconsequences of the unbalanced design, we ran 5GLMs in which we selected 3 natural-wetland sites atrandom. The results of these GLMs were similar tothose obtained with the unbalanced design. Therefore,we report the results for the unbalanced modelsbecause they included data from more wetlands thanthe balanced GLMs. We used post hoc Scheffe tests toexamine pairwise differences among wetland types.Last, we used 2-dimensional MDS to visualize patternsin coleopteran assemblages among wetland types (99restarts, 0.01 minimum stress, Kruskal fit scheme 1).

We used PRIMER (version 6.0 for Windows;PRIMER-E, Plymouth, UK) for cluster analysis,ANOSIM, SIMPER, MDS, and calculations of allcommunity variables except for S1 richness. We usedPASW (version 18; SPSS, Chicago, Illinois) to runGLMMs, GLMs, and Scheffe post hoc tests.

Results

Macroinvertebrate study

Overview.—No environmental or water-chemistryvariables varied significantly among the 5 periods.Across the study, annual mean air temperature variedbetween 7.3 and 8.2uC (F4,55 = 0.324, p = 0.861),annual mean precipitation between 36.8 and 46.6 mm/mo (F4,55 = 0.187, p = 0.944), water pH between 7.1and 7.6 (F4,55 = 1.725, p = 0.158), DIN concentrationbetween 4.75 and 6.46 mg/L (F4,55 = 0.562, p = 0.691),and SRP concentration between 0.037 and 0.057 mg/L(F4,55 = 1.307, p = 0.279).

One hundred taxa were found at Kalmar Damme.The groups with the highest species richness were:Coleoptera (23), Trichoptera (15), and Heteroptera(14). Six taxonomic groups dominated the communityacross the 5 periods spanning 13 y (Chironomidae,Asellidae, Corixidae, Baetidae, Oligochaeta, andPlanorbidae) and always contributed .80% of theabundance. The relative composition of taxonomicgroups was different for each period. For example,periods I and V were dominated by Chironomidae(with secondary groups being Corixidae in period Iand Oligochaeta in period V). However, duringperiods II, III, and IV Baetidae were abundant, the


numbers of Asellidae increased, and the numbers ofCorixidae declined (Fig. 2A). Eighteen families werepresent in period I. Eleven new families arrived inperiod II, 5 in period III, and 2 in period IV, whereasno new families appeared in period V.

Species composition.—Each of the 5 periods wascharacterized by a significantly different speciescomposition (ANOSIM global R = 0.778, p = 0.001;all pairwise R . 0.600, p , 0.05). MDS and clusteranalysis showed that the samples were different ineach period and that the samples were distributedalong a gradient related to time. The threshold of12.5% similarity was selected for grouping samplescorresponding to: 1) an early phase (period I), 2) anintermediate phase (periods II, III and IV), and 3) anadvanced phase (period V) (Fig. 3).

Changes in community composition were furtherconfirmed by SIMPER. Of the 17 species selected bythis method, only 8 were categorized as typifyingspecies (Table 3). These typifying species distin-guished one period from the next. However, periodsII, III, and IV shared the same main species. Thus, theearly phase was mainly represented by Sigara striata,the intermediate phase by Asellus aquaticus, and theadvanced phase by Sialis lutaria. Period III presented 6different species that typified the community (insteadof a range of 1–3 obtained in the other periods). Also,the mean similarity in composition of each periodgrew from the lowest level in the early phase (39.98%)to the highest level in the advanced phase (56.43%),with a mean of 51.69% during the intermediate phase.

Biological traits.—Most biological traits presentedsignificant differences across periods, but the rela-tionship between these trends and our hypotheseswas not always clear (Table 4). Species with short lifecycles did not differ significantly among periods,whereas species with long life cycles increased overtime (i.e., during periods III, IV, and V) until theybecame dominant (as expected). Species with aquaticdispersal did not differ significantly among periods.Species with active aerial dispersal were significantlymore abundant in period V than in earlier periods (theinverse trend to that expected), and the abundance ofpassive aerial species increased over time (as expect-ed). The abundance of herbivores did not differamong periods, and filter-feeders were most abun-dant in period V (the inverse trend to that expected).Shredders decreased over time (as expected), whereasthe remaining feeding types (fine sediment collectors,predators, and parasites) presented low values duringthe early periods and became more common overtime with high values in period V.

Community variables.—Most community variableschanged as hypothesized. GLMMs and post hoc tests

showed that N was low during periods I, II, III, andIV, and was significantly higher in period V (Fig. 2B).D+ and L+ showed that period V contained moretaxonomically unrelated taxa than earlier periods, andthe tree composition was more even than duringpreceding periods (Fig. 4A, B). J’ was significantlyhigher in period I than in period III (inverse trend tothat expected; Fig. 4C). The highest values for S1occurred during the intermediate phase and S2tended toward an asymptote (as expected) thatapproached 100 species (Fig. 4D).

Coleoptera study

The composition of each wetland was distinctive,regardless of wetland type (ANOSIM, R = 0.912, p =

0.001). Significant compositional differences wereobtained when comparing young (,3 y) and old(§10 y) man-made wetlands (R = 0.518, p = 0.001)and young man-made wetlands with natural wet-lands (R = 0.324, p = 0.001). In contrast, compositiondid not differ between old and natural wetlands (R =

20.150, p = 0.986). In the MDS plot, samples obtainedfrom old man-made wetlands were intermediatebetween samples from young man-made wetlandsand from natural wetlands (Fig. 5). Furthermore, oldman-made wetlands shared the highest number of the45 species selected by SIMPER (56% of total species)with other wetland types, i.e., 17.7% of the 45 specieswere shared with young man-made wetlands, 23.6%

were shared with natural wetlands, and 14.7% wereshared with both types of wetlands. Moreover, allspecies that were common to young man-madewetlands and natural wetlands also were present inold man-made wetlands. Exclusive species occurredmore often (48% of species) in natural wetlands thanin other wetland types (Table 5).

Sc differed significantly among wetland types (F2,61

= 14.356, p , 0.001). Old man-made wetlands hadsignificantly higher Sc than young man-made ornatural wetlands, but Sc did not differ between youngman-made and natural wetlands (Fig. 6).

Discussion

Succession of biological traits

The biological traits and method used (based onfuzzy coding) in our study provided meaningfulinformation about the ecological functioning of thisnewly created freshwater system despite the unavoid-able arbitrariness of assigning species to multipletraits groups (Hooper et al. 2002, Wright et al. 2006).In the macroinvertebrate study, both taxonomiccomposition and biological traits analyses displayed


a 3-phase pattern, and certain life-history strategieswere characteristic of the early, intermediate, andadvanced phases of succession. As expected, thenumber of individuals with long life spans and latematurities increased over time. These features areassociated with ecological stability rather than envi-ronmental stress (Townsend and Hildrew 1994,Bonada et al. 2007). Also as expected (Batzer andWissinger 1996), aerial passive dispersal was more

common in advanced rather than at early phases ofsuccession. Contrary to expectations, active dispersersalso dominated in the advanced successional phase.Initial dominance of this strategy is expected whendispersal distances can be covered efficiently by flight(Barnes 1983, Velasco et al. 1993, Bloechl et al. 2010).The high abundance of chironomids (active dispers-ers) in the advanced successional phase in our studywetlands might explain the lack of support for thishypothesis.

The abundances of most functional feeding groupsincreased over time. The results supported ourpredictions for fine sediment collectors, predators,and parasites, but contradicted our predictions forfilter-feeders, which were expected to decrease.Hydrological conditions of man-made wetlandsstrongly affect the resultant communities (Mitschand Wilson 1996). The unexpectedly high abundanceof filter-feeders in the late successional phase mighthave been a consequence of the continuous inputof nutrient-rich water, which may have sustainedindividuals in this guild throughout the study period(Gallardo et al. 2009). As expected based on theirdependence on macrophytes and large terrestrial litter(Voshell and Simmons 1984), shredders dominated atthe early rather than at the advanced successionalphase. Plant cover strongly influences macroinverte-brate communities (de Szalay and Resh 2000), andsubmerged macrophytes (mainly Elodea) covered

FIG. 2. Dominant taxonomic groups (A) and mean (61 SD) numerical abundance (B) of macroinvertebrates in Kalmar Dammefor each of the 5 study periods. Bars with the same letter are not significantly different (Scheffe post hoc test, p . 0.05).

FIG. 3. Plot of sites ordinated by 2-dimensional (2D)nonparametric multidimensional scaling analysis ofmacroinvertebrate species composition. Circles enclosegroups of sites with similarity .12.5% (similaritypercentages analysis).


large areas of the wetland during the early succes-sional stage and into the intermediate stage butdeclined mid-way through our study. Predators arean important component of macroinvertebrate com-munities in the initial phases of succession in bothtemporary and newly created water bodies (e.g.,Layton and Voshell 1991, Boix et al. 2006). Moreover,predator abundance increases over time as wetlandsundergo succession (Batzer and Wissinger 1996).Thus, predation is an important mechanism increating community structure in freshwater habitats(reviewed by Wellborn et al. 1996).

Community structure

In the macroinvertebrate study, maximum N wasrecorded at the advanced phase, as expected (Gutier-rez and Fey 1980, Legendre et al. 1985). However, thisincrease was mainly explained by the dominance ofchironomids. Thus, the hypothesized accompanyingincrease in evenness was not supported. Communitystructure was clearer under the prism of taxonomicdistinctness indices than when viewed in terms ofdiversity. On the one hand, the initial low values ofD+ and L+ showed that only a few taxonomic groupscolonized the wetland, a pattern that has beenobserved frequently in other systems (Velasco et al.1993, Heino et al. 2005, Ruhı et al. 2009). On the otherhand, despite a decline in S1, the increase in D+ near

the end of the study showed that the accumulation ofspecies during succession (S2) contributed to a moretaxonomically diverse community at the advancedphase. Moreover, the decrease in L+ demonstratedthat the taxonomic structure of the communitybecame more complex and more even in the advancedsuccessional phase. The arrival of new and taxonom-ically unrelated species took place in several taxo-nomic groups and largely compensated for the loss ofspecies from the few species-rich pioneer groups.Overall, the uneven community structure with lowtaxonomic distinction in the early phase developedprogressively into a richer, more even, and regularstructure. Maximum S1 was attained during theintermediate period, and increases in S2 showed thatnew species continued to colonize the area, trendsthat were expected (Margalef 1968, Batzer and Wis-singer 1996, Guo 2003, Milner and Robertson 2010).However, the rate of colonization declined over timeuntil it was almost negligible between periods IV andV, .10 y after creation of the wetland.

Man-made wetlands vs natural wetlands

Use of surrogate or model groups with biodiversitymetrics often has raised the criticism that thesurrogate group might not be representative of othergroups within the community of interest (Bonn andGaston 2005, Rodrigues and Brooks 2007). However,

TABLE 3. Similarity of percentages (SIMPER) analysis of macroinvertebrate community composition in Kalmar Damme. Valuesrepresent the percentage contribution of each species to the within-period similarity (only the most important species contributingglobally up to 70% of the within-period similarity are shown). Bold indicates species that consistently typified the community bycontributing a large proportion to total within-period similarity and being found at consistent abundance among samples in thatperiod (criterion: mean contribution/SD . 2).

Species

Period

I II III IV V

Average similarity (%) 39.98 50.38 58.00 46.71 56.43

Sigara striata 38.63* 8.83* 5.95Asellus aquaticus 18.88 27.11* 25.29* 24.20*Callicorixa praeusta 13.93Cloeon inscriptum 16.72* 9.09* 15.81Planorbis sp. 9.02 7.96*Erpobdella octoculata 7.73*Glyphotaelius pellucidus 7.17*Ischnura elegans 5.09*Gyraulus sp. 4.83 10.47Helobdella stagnalis 4.63 3.65Phryganea bipunctata 5.56Coenagrion puella/pulchellum 5.59Brachytron pratense 3.49Notonecta glauca 3.37Sialis lutaria 30.71*Pisidium sp. 15.39Agrypnia sp. 14.72


aquatic Coleoptera have been identified as potentialbiodiversity indicators in continental ecosystems(Foster et al. 1989, Sanchez-Fernandez et al. 2006).Moreover, this group is particularly species rich innorthern Europe (Nilsson 1984, Lundkvist et al. 2001,Thiere et al. 2009). Therefore, it is a useful surrogate.We were unable to determine whether the totalmacroinvertebrate community in the advanced phasetruly resembled a community that would be presentin a natural wetland. The Coleoptera study providedinteresting evidence to address this question becauseColeoptera usually require a long time to form matureassemblages (Valladares et al. 1994). Taxonomicsimilarity develops only after specialist taxa, whichrequire environmental stability and arrive afterpioneer generalist colonizers, can settle in the man-made wetland (Barnes 1983, Fairchild et al. 2000).Under the prism of composition (ANOSIM), 10 y ofsuccession seemed sufficient for Coleoptera assem-blages in the studied man-made wetlands to becomesimilar to those found in natural wetlands.

The significant differences in macroinvertebratecommunity composition found among wetlands ofsimilar type may be attributed to several factors.Idiosyncratic responses of species to different temporaland patch-specific factors are the most plausibleexplanation. Eventual critical episodes (e.g., an unex-pected dry period or the growth of different plantspecies) may strongly influence the final composition ofthe community (Jeffries 2003). Thus, similar wetlands inclose proximity to one another may differ in habitatheterogeneity, which affects the potential maximumnumber of coexisting taxa (Savage et al. 1998, Della Bellaet al. 2005). Differences in vegetation structure betweennatural and older man-made wetlands also may explaindifferences in macroinvertebrate species richness withina wetland type (Nilsson and Soderberg 1996).

Natural and old man-made wetlands were similarin terms of Coleoptera composition and structure, butyoung wetlands hosted an important proportion ofcharacteristic species (many of them from the familyDytiscidae). Some of these characteristic species (e.g.,Scarodytes halensis, Nebrioporus canaliculatus, Hygrotusconfluens) showed the pioneering nature of theassemblage (Ruhı et al. 2009, Bloechl et al. 2010) andwere absent from older wetlands. Other species (e.g.,Laccophilus minutus, Hydroglyphus geminus, Hydroporuspalustris) remained until later in the successionalprocess and contributed to the community presentin old wetlands. Thus, a substantial core group of taxamay explain the similarity of communities in oldman-made wetlands. Moreover, after enough time haspassed for the establishment of a viable and sustain-able community, old man-made wetlands may, in

TA

BL

E4.

Dif

fere

nce

sin

abu

nd

ance

of

mac

roin

ver

teb

rate

s’b

iolo

gic

altr

aits

amo

ng

per

iod

san

dth

ere

sult

so

fth

ere

spec

tiv

eg

ener

aliz

edli

nea

rm

ixed

mo

del

s(G

LM

Ms)

(sig

nif

ican

cew

asju

dg

edat

pƒ

0.00

45ac

cord

ing

toB

on

ferr

on

ico

rrec

tio

n).

Fo

rth

esi

gn

ific

ant

cate

go

ries

,p

erio

ds

wit

hth

esa

me

lett

erw

ere

no

tsi

gn

ific

antl

yd

iffe

ren

tac

cord

ing

toS

chef

fep

ost

ho

cte

sts

(p.

0.05

):a

=lo

wv

alu

es,

b=

med

ium

val

ues

,an

dc

=h

igh

val

ues

of

the

par

ticu

lar

trai

tca

teg

ory

atth

atp

erio

d.

Arr

ow

sin

dic

ate

the

dir

ecti

on

of

dif

fere

nce

.si

g=

sig

nif

ican

t,n

s=

no

nsi

gn

ific

ant,

her

b.

=h

erb

ivo

rou

s.

Bio

log

ical

trai

t

Hy

po

thes

isth

rou

gh

succ

essi

on

earl

yR

adv

ance

d

GL

MM

Per

iod

F1

4,3

0p

Per

iod

effe

ctI

IIII

IIV

V

No

.re

pro

du

ctiv

ecy

cles

/y

.1

+2

2.38

70.

022

ns

(no

tp

erfo

rmed

)ƒ

12

+3.

266

0.00

3si

gQ

ab

qc

qc

qc

Dis

per

sal

stra

teg

ies

Aq

uat

ic2

+2.

402

0.02

1n

s(n

ot

per

form

ed)

Aer

ial

acti

ve

+2

7.55

2,

0.00

1si

gb

bQ

ab

qc

Aer

ial

pas

siv

e2

+18

.289

,0.

001

sig

Qa

Qa

Qa

Qa

qc

Fee

din

gst

rate

gie

sS

crap

ers

+h

erb

.p

ierc

ers

+2

2.53

70.

017

ns

(no

tp

erfo

rmed

)F

ilte

r-fe

eder

s+

236

.934

,0.

001

sig

Qa

bb

bq

cS

hre

dd

ers

+2

4.45

1,

0.00

1si

gq

cq

cq

cQ

aQ

aF

ine

sed

imen

tco

llec

tors

2+

14.2

17,

0.00

1si

gQ

ab

bq

cq

cE

ng

ulf

ers

+an

imal

pie

rcer

s2

+6.

653

0.00

1si

gQ

aQ

aQ

aq

cq

c

Par

asit

es2

+12

.360

,0.

001

sig

Qa

Qa

Qa

Qa

qc


some cases, support higher levels of biodiversity thannatural wetlands (Mitsch and Wilson 1996). However,this conclusion should be treated with cautionbecause the number of species found exclusively innatural wetlands was high in our study. Moreover,many Coleoptera species are rare and only occur innonimpacted or natural sites (Nilsson and Soderberg1996, Lundkvist et al. 2002).

Conclusions

We observed 3 distinct phases of succession, duringwhich taxonomic composition and some measures ofcommunity structure changed according to our

hypotheses. A 3-phase pattern has been presentedpreviously as a theoretical framework within which todescribe secondary succession in temporary waterbodies (Lake et al. 1989, Bazzanti et al. 1996, Boix et al.2004). These phases were described as initial allogen-ic, middle autogenic, and final allogenic. We proposethat the early successional phase in Kalmar Dammewas allogenic as a direct consequence of the creationof the wetland. However, the important communitychanges that occurred in Kalmar Damme during theintermediate and advanced phases were not related toshifts in environmental variables, which were stableduring our study. Instead, we suggest that competi-tive exclusion led to a more stable community in theadvanced successional phase (Guo 2003). In a tempo-rary water body, successional changes in the finalphase have been described as allogenic because thewater body is drying. However, the advancedsuccessional phase of a community in a man-madewetland would not be allogenic but autogenic becauseof ongoing primary succession, as long as the systemremains stable. Our results suggest that a communityin a stage of advanced succession might be moretaxonomically robust and biodiverse and would alsoprovide the ecosystem with higher functionality thana community in an earlier successional stage (Hooperet al. 2005, Graham et al. 2006). Higher multifunction-ality in an ecosystem has been related to higher levelsof stability (Chapin et al. 2000, Hector and Bagchi2007). Therefore, we suggest that primary successiondynamics in a newly created wetland may beregarded as a process with successive increasingthresholds of community complexity and stability.

FIG. 4. Mean (61 SD) taxonomic distinctness (D+) (A), variation in taxonomic distinctness (L+) (B), Pielou’s evenness index (J’)(C), and rarefied species richness (S1) and cumulative species richness (S2) (D) of macroinvertebrate communities in KalmarDamme across the 5 study periods. Degrees of freedom, F statistics and p-values from generalized linear mixed models (GLMMs)are also shown. Dots with the same letter are not significantly different (Scheffe post hoc test, p . 0.05).

FIG. 5. Plot of sites ordinated by 2-dimensional (2D)nonparametric multidimensional scaling analysis of Cole-optera species composition in 3 wetland types.


Acknowledgements

This work was supported by a PhD grant and aScientific Research grant (CGL2008 05778/BOS) fromthe Ministerio de Ciencia y Tecnologıa of the SpanishGovernment. The authors thank Anders Bostrom andIsabell Petersson for assistance with sampling, sort-ing, and identification of samples from 1997–2003;Claes Kungberg for chemistry data; Nuria Pla for herwork on Fig. 1; David Estany-Tigerstrom for helpwith the GLMM models; Darold P. Batzer for generalcomments; David Butterfield, Gail Schofield, andJeffrey Garnett for English revision; and anonymousreviewers and associate editor Heikki Mykra forcomments that improved the initial manuscript.

Literature Cited

ABELLAN, P., D. T. BILTON, A. MILLAN, D. SANCHEZ-FERNANDEZ,AND P. M. RAMSAY. 2006. Can taxonomic distinctnessassess anthropogenic impacts in inland waters? A casestudy from a Mediterranean river basin. FreshwaterBiology 51:1744–1756.

ALLEN, T. F. H., S. M. BARTELL, AND J. F. KOONCE. 1977. Multiplestable configurations in ordination of phytoplanktoncommunity change rates. Ecology 58:1076–1084.

BARNES, L. E. 1983. The colonization of ball-clay ponds bymacroinvertebrates and macrophytes. Freshwater Biol-ogy 13:561–578.

BASS, D. 1992. Colonization and succession of benthicmacroinvertebrates in Arcadia Lake, a South-CentralUSA reservoir. Hydrobiologia 242:123–131.

BATZER, D. P., AND S. A. WISSINGER. 1996. Ecology of insectcommunities in nontidal wetlands. Annual Review ofEntomology 41:75–100.

TA

BL

E5.

Sim

ilar

ity

of

per

cen

tag

es(S

IMP

ER

)an

aly

sis

of

Co

leo

pte

rasp

ecie

sco

mp

osi

tio

nin

3w

etla

nd

typ

es.

Val

ues

rep

rese

nt

per

cen

tag

eco

ntr

ibu

tio

ns

of

spec

ies

toea

chw

etla

nd

typ

e(o

nly

spec

ies

wit

h.

1%co

ntr

ibu

tio

nto

each

wit

hin

-wet

lan

d-t

yp

esi

mil

arit

yar

esh

ow

n).

MM

=m

an-m

ade

wet

lan

d,

Nat

=

nat

ura

lw

etla

nd

.

Sp

ecie

sM

M,

3y

MM

§10

yN

at.

50y

Sp

ecie

sM

M,

3y

MM

§10

yN

at.

50y

Sp

ecie

sM

M,

3y

MM

§10

yN

at.

50y

Lac

cop

hil

us

min

utu

s13

.01.

3–

Hal

iplu

sh

eyd

eni

1.7

5.1

–H

elop

hor

us

min

utu

s–

2.0

2.6

Sca

rod

yte

sh

alen

sis

12.9

––

Lac

cobi

us

min

utu

s1.

71.

3–

Hy

dro

por

us

ery

thro

cep

hal

us

–2.

05.

4N

ebri

opor

us

can

alic

ula

tus

10.6

––

Aci

liu

sca

nal

icu

latu

s1.

5–

–E

noc

hru

ste

stac

eus

–1.

7–

Hy

dro

gly

ph

us

gem

inu

s7.

22.

9–

Rh

antu

ssu

tura

lis

1.2

3.2

2.0

Gra

pto

dy

tes

bili

nea

tus

––

6.3

Hy

dro

por

us

pal

ust

ris

5.2

2.0

2.3

Hal

iplu

sru

fico

llis

–1.

6–

Och

theb

ius

min

imu

s–

–4.

4H

yg

rotu

sco

nfl

uen

s4.

8–

–N

oter

us

clav

icor

nis

–6.

92.

0G

rap

tod

yte

sg

ran

ula

ris

––

4.2

Hy

gro

tus

inae

qual

is3.

84.

93.

2H

elop

hor

us

brev

ipal

pis

–5.

13.

0H

yg

rotu

sd

ecor

atu

s–

–4.

2A

nac

aen

alu

tesc

ens

3.3

8.6

5.2

Hyg

rotu

sim

pres

sopu

nct

atu

s–

4.5

–D

ryop

sau

ricu

latu

s–

–4.

1

Gy

rin

us

nat

ator

2.6

––

Lac

cobi

us

bip

un

ctat

us

–4.

0–

Dry

ops

sim

ilar

is–

–3.

8L

imn

ebiu

str

un

cate

llu

s2.

6–

–H

yd

rop

oru

su

mbr

osu

s–

3.6

1.0

Hal

iplu

sv

arie

gat

us

––

3.7

Ily

biu

sfu

lig

inos

us

2.3

3.6

–H

yd

robi

us

fusc

ipes

–3.

63.

1B

eros

us

luri

du

s–

–3.

6A

gab

us

bip

ust

ula

tus

2.2

3.7

2.4

Hy

dro

por

us

tris

tis

–3.

54.

4H

yd

rop

oru

sg

yll

enh

alii

––

3.1

Ag

abu

sn

ebu

losu

s2.

1–

–H

eloc

har

esob

scu

rus

–2.

32.

9Il

ybi

us

chal

con

atu

s–

–2.

3A

gab

us

stu

rmii

1.9

4.9

–N

oter

us

cras

sico

rnis

–2.

2–

Cy

mbi

ody

tam

arg

inel

la–

–1.

7A

cili

us

sulc

atu

s1.

9–

–H

yd

rop

oru

sst

riol

a–

2.9

–C

opel

atu

sh

aem

orrh

odal

is–

–1.

5FIG. 6. Box-and-whisker plots of Coleoptera richness (Sc)

in 3 wetland types. Bars with the same letter are notsignificantly different (Scheffe post hoc test, p . 0.05). Linesin boxes show medians, box ends show quartiles, whiskersshow 95% confidence intervals, circles show outliers.


BAZZANTI, M., S. BALDONI, AND M. SEMINARA. 1996. Inverte-brate macrofauna of a temporary pond in Central Italy:composition, community parameters and temporalsuccession. Archiv fur Hydrobiologie 137:77–94.

BLOECHL, A., S. KOENEMANN, B. PHILIPPI, AND A. MELBER. 2010.Abundance, diversity and succession of aquatic Cole-optera and Heteroptera in a cluster of artificial ponds inthe North German Lowlands. Limnologica 40:215–225.

BOIX, D., J. SALA, S. GASCON, AND S. BRUCET. 2006. Predation ina temporary pond with special attention to the trophicrole of Triops cancriformis (Crustacea: Branchiopoda:Notostraca). Hydrobiologia 571:341–353.

BOIX, D., J. SALA, X. D. QUINTANA, AND R. MORENO-AMICH.2004. Succession of the animal community in aMediterranean temporary pond. Journal of the NorthAmerican Benthological Society 23:29–49.

BONADA, N., S. DOLEDEC, AND B. STATZNER. 2007. Taxonomicand biological trait differences of stream macroinverte-brate communities between mediterranean and temper-ate regions: implications for future climatic scenarios.Global Change Biology 13:1658–1671.

BONN, A., AND K. J. GASTON. 2005. Capturing biodiversity:selecting priority areas for conservation using differentcriteria. Biodiversity and Conservation 14:1083–1100.

CHAPIN, F. S., E. S. ZAVALETA, V. T. EVINER, R. L. NAYLOR, P. M.VITOUSEK, H. L. REYNOLDS, D. U. HOOPER, S. LAVOREL, O. E.SALA, S. E. HOBBIE, M. C. MACK, AND S. DIAZ. 2000.Consequences of changing biodiversity. Nature 405:234–242.

CHEVENET, F., S. DOLEDEC, AND D. CHESSEL. 1994. A fuzzycoding approach for the analysis of long-term ecologicaldata. Freshwater Biology 31:295–309.

CHRISTMAN, V. D., AND J. R. VOSHELL. 1993. Changes in thebenthic macroinvertebrate community in two years ofcolonization of new experimental ponds. InternationaleRevue der gesamten Hydrobiologie und Hydrographie78:481–491.

CLARKE, K. R. 1993. Non-parametric multivariate analyses ofchanges in community structure. Australian Journal ofEcology 18:117–143.

CLARKE, K. R., AND R. N. GORLEY. 2006. PRIMER v6: usermanual/tutorial. PRIMER-E, Plymouth, UK.

CLARKE, K. R., AND R. M. WARWICK. 2001a. A furtherbiodiversity index applicable to species lists: variationin taxonomic distinctness. Marine Ecology ProgressSeries 216:265–278.

CLARKE, K. R., AND R. M. WARWICK. 2001b. Changes in marinecommunities: an approach to statistical analysis andinterpretation. 2nd edition. PRIMER-E, Plymouth, UK.

CRAWLEY, M. 2002. Statistical computing. An introduction todata analysis using S-Plus. John Wiley and Sons Ltd.,Chichester, UK.

DANELL, K., AND K. SJOBERG. 1982. Successional patterns ofplants, invertebrates and ducks in a man-made lake.Journal of Applied Ecology 19:395–409.

DE SZALAY, F. A., AND V. H. RESH. 2000. Factors influencingmacroinvertebrate colonization of seasonal wetlands:responses to emergent plant cover. Freshwater Biology45:295–308.

DELLA BELLA, V., M. BAZZANTI, AND F. CHIAROTTI. 2005.Macroinvertebrate diversity and conservation status ofMediterranean ponds in Italy: water permanence andmesohabitat influence. Aquatic Conservation: Marineand Freshwater Ecosystems 15:583–600.

EYRE, M. D., S. G. BALL, AND G. N. FOSTER. 1986. An initialclassification of the habitats of aquatic Coleoptera innorth-east England. Journal of Applied Ecology 23:841–852.

FAIRCHILD, G. W., A. M. FAULDS, AND J. F. MATTA. 2000. Beetleassemblages in ponds: effects of habitat and site age.Freshwater Biology 44:523–534.

FORCE, D. C. 1981. Postfire insect succession in SouthernCalifornia chaparral. American Naturalist 117:575–582.

FOSTER, G. N., A. P. FOSTER, M. D. EYRE, AND D. T. BILTON.1989. Classification of water beetle assemblages inarable fenland and ranking of sites in relation toconservation value. Freshwater Biology 22:343–354.

FRIDAY, L. E. 1987. The diversity of macro invertebrate andmacrophyte communities in ponds. Freshwater Biology18:87–104.

GALLARDO, B., S. GASCON, M. GARCIA, AND F. A. COMIN. 2009.Testing the response of macroinvertebrate functionalstructure and biodiversity to flooding and confinement.Journal of Limnology 62:315–326.

GRAHAM, N. J., S. K. WILSON, S. JENNINGS, N. V. C. POLUNIN,J. P. BIJOUX, AND J. ROBINSON. 2006. Dynamic fragility ofoceanic coral reef ecosystems. Proceedings of theNational Academy of Sciences of the United States ofAmerica 103:8425–8429.

GUO, Q. 2003. Temporal species richness-biomass relation-ships along successional gradients. Journal of Vegeta-tion Science 14:121–128.

GUTIERREZ, L. T., AND W. R. FEY. 1980. Ecosystemsuccession: a general hypothesis and a test modelof a grassland. MIT Press, Cambridge, Massachusetts.

HECK, K. L., G. VAN BELLE, AND D. SIMBERLOFF. 1975. Explicitcalculation of the rarefaction diversity measurementand the determination of sufficient sample size. Ecology56:1459–1461.

HECTOR, A., AND R. BAGCHI. 2007. Biodiversity and ecosystemmultifunctionality. Nature 448:188–190.

HEINO, J., J. SOININEN, J. LAPPALAINEN, AND R. VIRTANEN. 2005.The relationship between species richness and taxo-nomic distinctness in freshwater organisms. Limnologyand Oceanography 50:978–986.

HERRMANN, J., A. BOSTROM, AND I. BOHMAN. 2000. Invertebratecolonisation into the man-made Kalmar Damme wet-land dam system. Verhandlungen der InternationalenVereinigung fur theoretische und angewandte Lim-nologie 27:1653–1656.

HOOPER, D. U., F. S. CHAPIN, J. J. EWEL, A. HECTOR,P. INCHAUSTI, S. LAVOREL, J. H. LAWTON, D. M. LODGE,M. LOREAU, S. NAEEM, B. SCHMID, H. SETALA, A. J. SYMSTAD,J. VANDERMEER, AND D. A. WARDLE. 2005. Effects ofbiodiversity on ecosystem functioning: a consensus ofcurrent knowledge. Ecological Monographs 75:3–35.

HOOPER, D. U., M. SOLAN, A. SYMSTAD, S. DıAZ, M. O. GESSNER,N. BUCHMANN, V. DEGRANGE, P. GRIME, F. HULOT,


F. MERMILLOD-BLONDIN, J. ROY, E. SPEHN, AND L. VAN PEER.2002. Species diversity, functional diversity and ecosys-tem functioning. Oxford University Press, Oxford, UK.

HOV, A. M., AND B. WALSENG. 2003. Suksesjon av ferskvann-sinvertebrater i et nyetablert damsystem i Trogstadkommune. Norsk Institutt for Naturforskning (NINA),Oslo, Norway.

JEFFRIES, M. 2003. Idiosyncratic relationships between pondinvertebrates and environmental, temporal and patch-specific predictors of incidence. Ecography 26:311–324.

KAUFMAN, L. S. 1983. Effects of Hurricane Allen on reef fishassemblages near Discovery Bay, Jamaica. Coral Reefs 2:43–47.

LAKE, P. S., I. A. E. BAYLY, AND D. W. MORTON. 1989. Thephenology of a temporary pond in western Victoria,Australia, with special reference to invertebrate succes-sion. Archiv fur Hydrobiologie 115:171–202.

LANCASTER, J., F. EDWARDS, K. HALL, A. HOGAN, N. TOWERS, AND

S. TOWNSEND. 2004. Duloch Park remediation pondsbiomonitoring 1999–2003. Institute of Cell, Animal andPopulation Biology, University of Edinburgh, Edin-burgh, UK.

LAYTON, R. J., AND J. R. VOSHELL. 1991. Colonization of newexperimental ponds by benthic macroinvertebrates.Environmental Entomology 20:110–117.

LEGENDRE, P., S. DALLOT, AND L. LEGENDRE. 1985. Succession ofspecies within a community: chronological clustering,with applications to marine and freshwater zooplank-ton. American Naturalist 125:257–288.

LODS-CROZET, B., AND E. CASTELLA. 2009. Colonisation bymidges (Chironomidae, Diptera) of recently-createdshallow ponds: implications for the restoration oflacustrine fringing wetlands. Annales de Limnologie45:257–266.

LUNDKVIST, E., J. LANDIN, AND F. KARLSSON. 2002. Dispersingdiving beetles (Dytiscidae) in agricultural and urbanlandscapes in south-eastern Sweden. Annales ZoologiciFennici 39:109–123.

LUNDKVIST, E., J. LANDIN, AND P. MILBERG. 2001. Diving beetle(Dytiscidae) assemblages along environmental gradi-ents in an agricultural landscape in southeasternSweden. Wetlands 21:48–58.

MARGALEF, R. 1968. Perspectives in ecological theory.University of Chicago Press, Chicago, Illinois.

MILNER, A. M., AND A. L. ROBERTSON. 2010. Colonization andsuccession of stream communities in Glacier Bay,Alaska. What has it contributed to general successionaltheory? River Research and Applications 26:26–35.

MITSCH, W. J., AND R. F. WILSON. 1996. Improving the successof wetland creation and restoration with know-how,time, and self-design. Ecological Applications 6:77–83.

NIELL, F. X. 1981. Estructuras disipativas en la organizacion delos sistemas bentonicos. Oecologia Aquatica 5:239–245.

NILSSON, A. N. 1984. Species richness and succession ofaquatic beetles in some kettle-hole ponds in northernSweden. Ecography 7:149–156.

NILSSON, A. N., AND H. SODERBERG. 1996. Abundance andspecies richness patterns of diving beetles (Coleoptera,

Dytiscidae) from exposed and protected sites in 98northern Swedish lakes. Hydrobiologia 321:83–88.

ODUM, E. P. 1960. Organic production and turnover in oldfield succession. Ecology 41:34–49.

ODUM, E. P. 1971. Fundamentals of ecology. Saunders,Philadelphia, Pennsylvania.

OERTLI, B., J. BIGGS, R. CEREGHINO, P. GRILLAS, P. JOLY, AND J.-B.LACHAVANNE. 2005. Conservation and monitoring ofpond biodiversity: introduction. Aquatic Conservation:Marine and Freshwater Ecosystems 15:535–540.

OKSANEN, J., F. G. BLANCHET, R. KINDT, P. LEGENDRE, R. B.O’HARA, G. L. SIMPSON, P. SOLYMOS, M. H. H. STEVENS, AND

H. WAGNER. 2010. Vegan package v. 1.17-3. R Project forStatistical Computing, Vienna, Austria. (Available from:http://cc.oulu.fi/,jarioksa/softhelp/vegan.html)

PECHMANN, J., R. ESTES, D. SCOTT, AND J. GIBBONS. 2001.Amphibian colonization and use of ponds created fortrial mitigation of wetland loss. Wetlands 21:93–111.

PETRANKA, J. W., E. M. HARP, C. T. HOLBROOK, AND J. A.HAMEL. 2007. Long-term persistence of amphibianpopulations in a restored wetland complex. BiologicalConservation 138:371–380.

PETRANKA, J., C. KENNEDY, AND S. MURRAY. 2003. Response ofamphibians to restoration of a southern Appalachianwetland: a long-term analysis of community dynamics.Wetlands 23:1030–1042.

RODRIGUES, A. S. L., AND T. M. BROOKS. 2007. Shortcuts forbiodiversity conservation planning: the effectiveness ofsurrogates. Annual Review of Ecology, Evolution, andSystematics 38:713–737.

RUHI, A., D. BOIX, J. SALA, S. GASCON, AND X. D. QUINTANA.2009. Spatial and temporal patterns of pioneer macro-fauna in recently created ponds: taxonomic and func-tional approaches. Hydrobiologia 634:137–151.

SANCHEZ-FERNANDEZ, D., P. ABELLAN, A. MELLADO, J. VELASCO,AND A. MILLAN. 2006. Are water beetles good indicatorsof biodiversity in Mediterranean aquatic ecosystems?The case of the Segura river basin (SE Spain). Biodiver-sity and Conservation 15:4507–4520.

SAVAGE, A. A., J. H. MATHEWS, AND D. L. BEAUMONT. 1998.Community development in the benthic macroinverte-brate fauna of a Lowland Lake, Oak Mere, from 1994 to1996. Archiv fur Hydrobiologie 143:295–305.

SCHNEIDER, D. W., AND T. M. FROST. 1996. Habitat durationand community structure in temporary ponds. Journalof the North American Benthological Society 15:64–86.

SMITH, K. G. 1982. Drought-induced changes in aviancommunity structure along a montane sere. Ecology63:952–961.

SMITH, V. G. 1928. Animal communities of a deciduous forestsuccession. Ecology 9:479–500.

SOLIMINI, A. G., A. RUGGIERO, V. BERNARDINI, AND G. CARCHINI.2003. Temporal pattern of macroinvertebrate diversityand production in a new man made shallow lake.Hydrobiologia 506–509:373–379.

TACHET, H., P. RICHOUX, M. BOURNAUD, AND P. USSEGLIO-POLATERA. 2002. Invertebres d’Eau Douce (2nd correctededition). CNRS editions, Paris, France.


THIERE, G., S. MILENKOVSKI, P.-E. LINDGREN, G. SAHLEN,O. BERGLUND, AND S. E. B. WEISNER. 2009. Wetland creationin agricultural landscapes: biodiversity benefits on localand regional scales. Biological Conservation 142:964–973.

TOWNSEND, C. R., AND A. G. HILDREW. 1994. Species traits inrelation to a habitat templet for river systems. Fresh-water Biology 31:265–275.

VALLADARES, L. F., J. GARRIDO, AND B. HERRERO. 1994. The annualcycle of the community of aquatic Coleoptera (Adephagaand Polyphaga) in a rehabilitated wetland pond: theLaguna de La Nava (Palencia, Spain). Annales de Limno-logie – International Journal of Limnology 30:209–220.

VELASCO, J., A. MILLAN, AND L. RAMIREZ-DIAZ. 1993. Coloniza-cion y sucesion de nuevos medios acuaticos I. Composi-cion y estructura de las comunidades de insectos.Limnetica 9:73–85.

VERBERK, W. C. E. P., H. SIEPEL, AND H. ESSELINK. 2008. Life-history strategies in freshwater macroinvertebrates.Freshwater Biology 53:1722–1738.

VOSHELL, J. R., AND G. SIMMONS. 1984. Colonization andsuccession of benthic macroinvertebrates in a newreservoir. Hydrobiologia 112:27–39.

WARWICK, R. M., AND K. R. CLARKE. 1995. New ‘biodiversity’measures reveal a decrease in taxonomic distinctnesswith increasing stress. Marine Ecology Progress Series129:301–305.

WELLBORN, G. A., D. K. SKELLY, AND E. E. WERNER. 1996.Mechanisms creating community structure across afreshwater habitat gradient. Annual Review of Ecologyand Systematics 27:337–363.

WRIGHT, J. P., S. NAEEM, A. HECTOR, C. LEHMAN, P. B. REICH,B. SCHMID, AND D. TILMAN. 2006. Conventional functionalclassification schemes underestimate the relationshipwith ecosystem functioning. Ecology Letters 9:111–120.

ZEDLER, J. B., AND J. C. CALLAWAY. 1999. Tracking wetlandrestoration: do mitigation sites follow desired trajecto-ries? Restoration Ecology 7:69–73.

ZUUR, A. F., E. N. IENO, N. J. WALKER, A. A. SAVELIEV, AND

G. M. SMITH. 2009. Mixed effects models and extensionsin ecology with R. Springer, New York.

Received: 2 March 2011Accepted: 7 September 2011


change in biological traits and community structure of macroinvertebrates through primary succession...

Documents