Yoccozse9999se.com

AQUATICBIOLOGYAquatBiolVol.
23:15–28,2014doi:10.
3354/ab00600PublishedonlineDecember2INTRODUCTIONWithrapidurbanizationandeconomicdevelopment,China,especiallythecoastalareas,hasattainedremarkableeconomicsuccess.
However,thisrapideconomicgrowthhasresultedinaseriesofsevereenvironmentalproblems(Fuetal.
2007)suchasincreasesineutrophication(Qinetal.
2007,Smith2003),organicpollutants(An&Hu2006),heavymet-als(Zhangetal.
2009),andhabitatdegradation(Lietal.
2010),whichhaveplacednewpressuresonnationalsustainabledevelopment.
AswaterqualityTheauthors2014.
OpenAccessunderCreativeCommonsbyAttributionLicence.
Use,distributionandreproductionareun-restricted.
Authorsandoriginalpublicationmustbecredited.
Publisher:Inter-Research·www.
int-res.
com*Correspondingauthor:caiyj@niglas.
ac.
cnMacroinvertebrateassemblagesinstreamsandriversofahighlydevelopedregion(LakeTaihuBasin,China)YouZhang1,2,LingLiu1,LongCheng1,3,YongjiuCai2,3,*,HongbinYin2,JunfengGao2,YongnianGao21StateKeyLaboratoryofHydrology-WaterResourcesandHydraulicEngineering,HohaiUniversity,Nanjing210098,PRChina2StateKeyLaboratoryofLakeScienceandEnvironment,NanjingInstituteofGeographyandLimnology,ChineseAcademyofSciences,Nanjing210008,PRChina3NanjingHydraulicResearchInstitute,Nanjing210029,PRChinaABSTRACT:TheLakeTaihuBasin,oneofthemosthighlydevelopedregionsinChina,suffersfromseverewaterpollutionandhabitatdegradationasaresultofeconomicdevelopmentandurbanization.
However,littleresearchhasbeenconductedonthebenthicmacroinvertebrateassemblagesandtheirenvironmentalrelationships.
Westudiedthecommunitystructureofben-thicmacroinvertebratesinthisareaandexploredpossiblefactorsregulatingtheirstructure,diver-sity,anddistribution.
Werecordedatotalof104taxa;Bellamyaaeruginosahadthehighestfre-quencyofoccurrence(71outof93sites),followedbyLimnodrilushoffmeisteri(49)andCorbiculafluminea(42).
Oligochaetawasthemostabundanttaxonomicgroup,accountingfor89.
42%oftotalmacroinvertebrateabundance.
Benthiccommunitiesweremainlycharacterizedbypollution-toleranttaxa,suchasL.
hoffmeisteri(Oligochaeta),B.
aeruginosa(Gastropoda),andPolypedilumscalaenum(Chironomidae),indicatingsevereanthropogenicdisturbanceandhabitatdegrada-tion.
Themacroinvertebratediversitydecreasedfromthewesternhillstotheeasternplainsaquaticecoregions;theShannon-WienerandMargalefindicesdifferedsignificantlybetweenthetwoecoregions.
Theabundances(%)ofgatheringcollectorsincreasedfromupstreamtodown-stream,butscrapersshowedtheoppositetrend,consistentwiththerivercontinuumconcept.
CommunitystructureandspatialpatternsofmacroinvertebratesintheLakeTaihuBasinwerestronglycorrelatedwithhabitatdiversity,nutrientloads,andaquaticvegetationcoverage.
Theseresultsprovidevaluableinformationforeffectivemanagementpracticesofbiodiversityconserva-tioninstreamandriverecosystemsoftheLakeTaihuBasin.
KEYWORDS:Anthropogenicdisturbance·Habitatdegradation·Nutrientenrichment·Ecoregion·Sub-ecoregion·YangtzeRiverDeltaOPENPENACCESSCCESSAquatBiol23:15–28,2014mayonlypartiallyreflectenvironmentalimpact(Birketal.
2012),moreattentionshouldbegiventothestatusofbiologicalcommunitiesandconservationbiodiversity(Morseetal.
2007)inthemanagementofaquaticecosystems.
Macroinvertebratesareanimportantcomponentofstreamandriversystemsandplaycrucialrolesinmaintainingthestructuralandfunctionalintegrityoffreshwaterecosystems(Wallace&Webster1996).
Theyalterthegeophysicalconditionofsediments,promotedetritusdecompositionandnutrientcycling,andfacilitateenergytransferamongtrophiclevels(Vanni2002,Covichetal.
2004).
Benthicmacro-invertebratesarethemostcommonlyusedbiologicalindicatorsinmostaquaticecosystemsduetotheirvariablesensitivitytoenvironmentalchange(Pignataetal.
2013)andeaseofsampling(i.
e.
lowcost).
Unfortunately,recentdiscoverieshaveshownextinc-tionratesoffreshwaterfaunatobeasmuchas5timesgreaterthanthatofterrestrialfauna(Ricciardi&Rasmussen1999).
Thedeclineofbenthicmacro-invertabratesinaquaticsystemsislargelyduetoanthropogenicimpactintheformofhabitatdeterio-rationandnutrientoverload.
Thediversityofthebenthiccommunityisprogressivelysimplifiedduetothedecreasednumberoftaxa(Yuan2010,Caietal.
2012b).
Habitatcomplexityisoneofthekeyenviron-mentalfactorsinfluencingmacroinvertebratecom-munities.
Complexhabitatsprovidemoreecologicalniches,whichmakemacroinvertebrateshighlyvul-nerabletothelossoftheirpreferredhabitat(McGoffetal.
2013).
Consequently,habitatdeteriorationwillseverelydepressthediversityandcompositionofbenthiccommunities.
Thus,identifyingthepossiblefactorsregulatingmacroinvertebratestructure,diver-sity,anddistributioncanaidthedevelopmentofmoreprescriptiveconservationandmanagementstrate-giesforfreshwaterecosystemsinhighlydevelopedregions.
TheLakeTaihuBasinisoneofthemostindustrial-izedareasinChina.
Withapopulationofmorethan37million(2.
9%ofthetotalpopulationofChina),andasasignificantindustrialcomplex,thisareacon-tributes11%toChina'sgrossdomesticproduct(Liuetal.
2013).
Cultivatedland,whichisheavilyfertil-izedeachyear,covers15100km2ofthedrainagearea,whichis>40%ofthetotalarea.
TheplainrivernetworkisthemainwastewaterdischargeregionofsouthernJiangsuProvince(Qinetal.
2007);con-sequently,thefreshwaterecosystemofthebasinissubjecttoseveredamage.
Studiestodatehavelargelyfocussedonwaterphysico-chemistry(e.
g.
eutrophicationandheavymetals)(Liuetal.
2013)andaquaticorganismshavereceivedlittleattentiondespitetheirimportanceinecosystemhealth(Xiaoetal.
2013).
StudiesofbenthicmacroinvertebratecommunitiesintheBasinhavebeenlargelyfocusedonLakeTaihu(Caietal.
2011,2012a),withveryfewstudiesex-aminingthemainstreamsandrivers(Wangetal.
2007,Gaoetal.
2011,Wuetal.
2011).
Therefore,3questionsneedtobeaddressed:(1)Whatistheconditionofthebenthicmacroinvertebratecommu-nitystructureintheLakeTaihuBasin(2)ArethemacroinvertebrateassemblagesintheLakeTaihuBasinsimilartootherhighlydevelopedregions(3)WhatarethepossiblevariablesregulatingtheirabundanceandcommunitystructureToelucidatethesequestions,themacroinvertebratecommunities,waterchemistryandhabitatcharacteristicsofstreamsandriversintheLakeTaihuBasin,aswellasthepos-siblefactorsregulatingtheirstructure,diversityanddistributionwereinvestigated.
MATERIALSANDMETHODSStudyareaTheLakeTaihuBasin(30°7'19''to32°14'56''Nand119°3'1''to121°54'26''E)issituatedintheYangtzeRiverDeltaineasternChinaandcoversanareaofapproximately36900km2(Fig.
1).
Ithasawatersurfaceareaof6134km2,ofwhichriversandlakeseachcomprisearound50%.
Thedenserivernetworkcomprises7%ofthetotaldrainagearea,withatotaltributarylengthofaround120000km(Gao&Gao2012).
TheLakeTaihuBasin,characterizedbyplainsandrivernetworksintheeastandhillsandstreamsinthewest,canbedividedinto2ecoregions:thewesternhillaquaticecoregion(S1)andtheeasternplainaquaticecoregion(S2)(Gao&Gao2012).
SignificantdifferencesinlanduseexistbetweenS1andS2.
Farmlandandwoodlandarethedominantland-scapesinS1,accountingfor44.
70and42.
00%,respectively.
Therelativelylowlevelsofagriculturaldevelopmentinthisecoregionhaveresultedinfewerpollutantsfromagriculturalsourcesandlesspressureontheaquaticenvironment.
Incontrast,cultivatedlandmakesup50%ofthetotalareaofS2,followedbybuilt-upareas(27.
36%)andwaterbodies(17.
00%).
HighlevelsofurbanizationandahighpopulationdensityinS2haveledtosignificantagriculturalandurbanrunoffandindustrialwaste,whicharedirectcausesofwaterqualitydeteriorationinthisarea.
In16Zhangetal.
:MacroinvertebrateassemblagesinLakeTaihuBasinriversaddition,rapiddevelopmentofaquacul-turehasalsoacceleratedwaterpollutionandeutrophication.
Forthisreason,spa-tialdifferentiationinecology,vegeta-tion,andnutrientlevelswithineachecoregionareevident,andeachecore-gion(S1orS2)hasbeendividedinto2(S11andS12)or3(S21,S22andS23)sub-ecoregions(Fig.
1).
Astheobjectsofthisstudywereriversandstreams,S22(LakeTaihu)wasexcludedinthisstudy.
MacroinvertebratesamplingAtotalof93samplingsiteswereselectedacrosstheentireLakeTaihuBasin.
Macroinvertebratesampleswerecollectedwithina100mreachforeachsiteinOctober2012.
Samplingwascon-ductedwitha30cmdiameterD-framenetwitha500mmmeshusingthemulti-habitatapproachdescribedbyBarbouretal.
(1999).
Ateachsite,10samplingunits(30*50cm)werecollectedtoincludethemajormicrohabitats(e.
g.
riffles,pools,banks,submergedandoverhangingmacrophytes,depositionalareas).
Determinationofmajorhabitattypeswasmadepriortosamplingbyqualitativelyevaluatingthesamplereach.
The10samplingunitsweredividedpro-portionallybasedonavailablehabitats.
Allmaterialscollectedfromasitewerepooledandrinsedinthefieldtoremovefinesediments,andallremainingmate-rialswerefixedwitha7%bufferedformaldehydesolution.
Inthelaboratory,samplesweresortedbyhandinwhiteenamelpanswiththeaidofadissectingmicroscope.
Allorganismswerepre-servedin70%ethanolandidentifiedtothelowestfeasibletaxonomiclevelunderadissectingandanuprightmicroscope,usingkeysbyLiuetal.
(1979),Morseetal.
(1994),Wang(2002),andTang(2006).
Macroinvertebrateabundancewasob-tainedbycountingallindividualsandexpressingtheresultsasind.
m2.
Invertebratetaxawereassignedtofunctional-feedinggroups(FFGs)basedoninvertebratemorphologicalandbe-havioraladaptationsforacquiringtheir17Fig.
1.
Locationofthesamplingsites,aquatic(sub)ecoregions,andlanduse.
S1(westernhillaquaticecoregion)iscomprisedofsub-ecoregionsS11andS12,andS2(easternplainaquaticecoregion)ismadeupofsub-ecoregionsS21,S22,andS23.
Astheobjectsofthisstudywereriversandstreams,S22(LakeTaihu)wasexcludedAquatBiol23:15–28,2014food(Cummins&Klug1979,Merritt&Cummins2008).
TheFFGsusedwerecategorizedasfollows:(1)shredders(SH,feedoncoarseparticulateorganicmatter>1mminsize,eitherliveaquaticmacrophytetissueorcoarseterrestrialplantlitter);(2)scrapers(SC,scrapeoffandconsumetheorganicmatterattachedtostonesandothersubstratesurfaces,pri-marilyliveplantstems);(3)filteringcollectors(FC,siftfineparticulates1000to0.
45mfromtheflowingcolumnofwater);(4)gatheringcollectors(GC,gatherfineparticulatesoforganicmatterfromthedebrisandsedimentsonthestreambeds);and(5)predators(PR,feedonotheranimals,i.
e.
liveprey).
Therela-tiveabundance(%)ofeachFFGwascalculatedfromtheirtotalabundanceinsamplesfromeachsite.
MeasurementofenvironmentalparametersThepH,conductivity(cond),dissolvedoxygen(DO),chlorophylla(chla),salinity,andturbidityweremeasuredinunder-surfacewater(collected0.
5mbelowthewatersurface)usingawaterqualityana-lyzer(YSI6600V2);sampleswerekeptat4°Cforfurtherchemicalanalysis.
Whenthewaterdepthofthestreamwaslessthan0.
5m,watersampleswerecollectedfromanintermediatedepth.
Thetotalnitrogen(TN),ammonium(NH4+-N),nitrate(NO3-N),totalphosphorus(TP),orthophosphate(PO43-P),totalsuspendedsolids(TSS),chemicaloxygendemand(CODMn),andsulfate(SO42)weremeasuredinthelaboratorybasedonstandardmethods(APHA2012).
Todescribethephysicalpropertiesofthehabitat,6parameterswerechosen:habitatdiversity(habitat),channelmorphology(channel),aquaticvegetationcoverage(vegetation),roaddensity(road),riparianzonelanduse(landuse),andsubstrateindex(SI).
Thevaluesforhabitat,channel,road,andlandusewerescoredinthefieldandrangedfrom0to20(Barbouretal.
1999).
Thevalueforvegetationreflectedthepercentagecoverofaquaticvegetation.
SI(Weatherhead&James2001)isacompositeindexreflectingtheheterogeneityofthesubstrate,anditcanbecalculatedas:SI=0.
07*(%largeboulder)+0.
06*(%boulder)+0.
05*(%cobble)+0.
04*(%pebble)+0.
03*(%gravel)+0.
02*(%sand)+0.
01*(%mud/silt)Thisgivesavaluerangingfrom1formud/siltto7forlargeboulders.
PriortotheSIcalculation,thepro-portionofeachparticlesizeclasswasestimatedforeachsamplingsitebasedonmethodsdescribedbyKondolf(1997)andJowett&Richardson(1990).
Sub-strateswereclassifiedaccordingtothefollowingcri-teria:512mm=largeboulder.
DataanalysisThebiologicalindicesofShannon-Wiener(H'),Simpson(1D),MargalefandPielouwerecalcu-latedintermsofabundanceusingthePASTsoftwarepackage(PaleontologicalStatisticsv.
2.
17)(Hammeretal.
2001).
Aprincipalcomponentanalysis(PCA)basedonacorrelationmatrixamongsampleswasusedtoanalyzethephysico-chemicaldata.
PCAwasperformedusingCANOCOv.
4.
5software(terBraak&Smilauer2002).
Differencesinenvironmentalvari-ableswereexaminedamongthe4sub-ecoregionsusing1-wayANOVAatasignificancelevelofp3)orRDA(<3)wouldbeappropriate.
Manualforwardselectionwasusedtodeterminewhichenvironmentalvariablesweresignificantlyrelatedtothebenthiccommunity(MonteCarlotestwith9999permutations,p<0.
1).
Wese-lectedasignificancelevelofp<0.
1giventhetypicallyhighvariabilityininvertebratedata(Yoccoz1991).
Thestatisticalsignificanceofspeciesenvironmentcorrelationsfortheordinationaxeswerealsodeter-minedbasedon9999MonteCarlopermutationtests.
Datawerelogarithmicallytransformedtoapproximatenormality,andtaxaoccurringatlessthan5siteswereexcludedfortheANOSIM,CCA,andRDAanalyses.
RESULTSEnvironmentalcharacterizationANOVAanalysesindicatedthatthe4sub-ecoregionsdifferedsignificantly(p<0.
05)inmostphysical,chemical,andbiologicalvariables,exceptfor18Zhangetal.
:MacroinvertebrateassemblagesinLakeTaihuBasinriversinpH,TSS,andturbidity(Table1).
Variationofenvi-ronmentalvariablesbetweenecoregionswasgreaterthanwithinecoregions.
Ingeneral,S11andS12hadlowernutrientconcentrations(e.
g.
nitrogenandphos-phorus)andhigherDO,whereasS21presentedthehighestnutrientconcentrationsandpollutionlevels(e.
g.
condandCODMn).
S1valuesforhabitat,channel,vegetation,road,riparianlanduse,andSIwerehigherthanthecorrespondingS2values.
Concentra-tionsofthestudiednutrients(TNandTP)andchlashowedconsistentspatialpatternsincreasingfromwesttoeast,buthabitatdiversityshowedtheoppositetrend,indicatingintensiveanthropogenicdisturbanceinthedownstreamsections(S21andS23).
ResultsofPCAindicatedthatPC1andPC2accountedfor33.
1and12.
2%ofthetotalvarianceofenvironmentalvariables,respectively(Fig.
2).
ThePC1primarilydescribedthenutrientloadswithneg-ativeloadingsforTN,TP,PO4-P,NH4-N,NO3-N,F,Cl,SO42,CaCO3,CODMn,salinity,andcond,whilethephysicalhabitatsharedpositiveloadings(e.
g.
landuse,channelmorphology,roadwidth)onPC1.
ThePC2hadastrongpositiverelationshipwithTSSandturbidity.
Takentogether,resultsofthePCAsug-gestedthatnutrientloadsandhabitatdegradationincreasedfromwesttoeast.
CommunitycompositionanddiversityAtotalof104macroinvertebratetaxawerere-corded(TableS1intheSupplementatwww.
int-res.
com/articles/suppl/b023p015_supp.
pdf),including10Gastropoda,8Bivalvia,3Oligochaeta,2Poly-chaeta,8Hirudinea,8Crustacea,30Chironomidae,20Odonata,and14otherinsecttaxa.
Thenumberoftaxaateachstationvariedbetween1and24,withanaveragenumberof7.
6,andshowedadecreasingtrendfromwesttoeast(Fig.
3).
Bellamyaaeruginosawasthetaxonwiththehighestfrequencyofoccur-rence(71outofthetotal93sites),andtogetherwithParafossaruluseximius(41)andSemisulcospiracan-celata(27),the3taxawereimportantcontributorstothetotalabundanceofGastropoda.
LimnodrilushoffmeisteriandBranchiurasowerbyibestrepre-sentedOligochaetawithinthebasin,withfrequen-19S11(n=9)S12(n=13)S21(n=21)S23(n=50)FppH7.
88(7.
148.
73)7.
93(7.
468.
72)7.
89(7.
548.
83)7.
88(7.
308.
41)0.
2650.
850DO(mgl1)6.
0(2.
312.
0)ab7.
1(2.
811.
9)a4.
8(0.
77.
3)b4.
8(0.
49.
4)ab4.
5050.
005Cond(scm1)0.
40(0.
231.
12)ab0.
27(0.
150.
63)a0.
64(0.
341.
08)c0.
60(0.
340.
90)bc12.
448<0.
001TN(mgl1)2.
11(0.
984.
51)b2.
15(1.
344.
11)b4.
01(1.
608.
13)a3.
55(1.
086.
87)a8.
462<0.
001NH4+-N(mgl1)0.
02(0.
010.
60)b0.
08(00.
99)ab1.
03(04.
51)a0.
65(0.
024.
36)a4.
7970.
004NO3-N(mgl1)1.
15(0.
061.
83)0.
88(0.
023.
24)1.
55(0.
083.
04)1.
48(0.
024.
21)2.
3400.
079TP(mgl1)0.
05(0.
030.
16)ab0.
03(0.
020.
26)a0.
20(0.
040.
94)c0.
16(0.
040.
60)bc5.
1960.
002PO43-P(gl1)22.
10(7.
8436.
95)8.
59(5.
01129.
17)57.
79(3.
33252.
11)45.
50(2.
08247.
62)2.
3020.
083CODMn(mgl1)3.
44(1.
716.
91)2.
75(1.
198.
64)4.
50(2.
926.
04)4.
14(1.
716.
91)1.
6960.
174TSS(mgl1)37(14428)29(6278)75(34120)55(5652)0.
1740.
913Chla(gl1)5.
6(2.
424.
0)7.
0(1.
425.
6)9.
5(2.
827.
4)7.
6(1.
923.
3)2.
6300.
055SO42(mgl1)35.
52(16.
1663.
33)a30.
59(14.
61137.
37)ab94.
75(25.
61246.
60)b83.
03(6.
54358.
62)b5.
3620.
002CaCO3(mgl1)73.
84(57.
4386.
15)b65.
64(41.
0298.
46)b94.
65(73.
84123.
07)a90.
25(53.
33141.
54)a11.
516<0.
001Salinity(%)0.
19(0.
110.
87)ab0.
13(0.
070.
31)b0.
32(0.
160.
53)a0.
30(0.
160.
57)a7.
841<0.
001Turbidity(NTU)29.
6(0.
2667.
5)16.
9(0.
11217.
6)94.
4(19.
9236.
4)63.
6(10.
01139.
1)0.
8910.
449Habitat(20')10(115)a10(217)a3.
29(111)b4(112)b13.
315<0.
001Channel(20')10(218)a10(217)a5.
71(114)b3(113)b10.
987<0.
001Vegetation(%)10(090)a20(070)ab6(050)b0(070)b6.
2460.
001Road(20')9(613)a9(212)a6(312)b5(110)b11.
627<0.
001Land(20')9(411)a8(314)a5(29)b4(19)b14.
381<0.
001SI(17)1.
1(1.
03.
0)a1.
5(1.
03.
9)b1.
1(1.
01.
9)a1.
0(1.
02.
0)a11.
427<0.
001Simpson0.
62(0.
180.
88)0.
53(00.
83)0.
43(00.
82)0.
41(00.
87)2.
2370.
089Shannon-Wiener1.
44(0.
402.
52)a1.
31(02.
05)ab0.
84(02.
07)b0.
87(02.
31)b4.
2080.
008Margalef2.
55(0.
634.
94)a2.
30(04.
25)a1.
11(02.
51)b1.
28(03.
88)b8.
945<0.
001Pielou0.
69(0.
251)0.
51(00.
82)0.
53(00.
93)0.
48(00.
99)1.
6140.
192Table1.
Comparisonofenvironmentalvariablesamongthe4sub-ecoregions(westernhillaquaticecoregionsS11andS12,andeasternplainaquaticecoregionsS21andS23).
When1-wayANOVAindicatedsignificantdifferences(p<0.
05),eachsub-ecoregiondifferingintheposthocTukeytestswasgivenadifferentletter(a,b,orc).
DO=dissolvedoxygen;Cond=conductivity;TN=totalnitrogen;NH4+-N=ammonium;NO3-N=nitrate;TP=totalphosphorus;PO43-P=orthophosphate;CODMn=chemicaloxygendemand;TSS=totalsuspendedsolids;SO42=sulfate;SI=substrateindex.
20'referstoascalewithamaximumscoreof20.
Valuesaremedian(range)AquatBiol23:15–28,2014ciesofoccurrenceof49and41,respectively.
Com-paredwiththehighoccurrencesoftheformer2taxa,RhyacodrilussinicuswasfoundonlyoncewithinS23.
ThefrequencyofoccurrenceforCorbiculafluminea(42)wasthehighestoftheBivalvia.
OtherBivalvia(Anodontawoodianaelliptica,A.
w.
pacifica,Acuti-costachinensis,Uniodouglasiae,andLimnopernafortunei)alsohadfrequenciesofoccurrencegreaterthan10.
ExopalaemonmodestusandNeocaridinadenticulatawerethe2maintaxaofCrustaceaandwerefoundin37and26stations,respectively.
TheChironomidaePolypedilumscalaenumandChirono-musplumosusarepollution-toleranttaxaandhadfrequenciesof12and11,respectively.
Totalabundanceofmacroinvertebratevariedgreatlybetween1.
33and39080ind.
m2,withanaverageof1159ind.
m2.
Arelativelyevenabundancedistribu-tionwasobservedandanextremelyhighabundancewasrecordedatsomestations(mainlylocatedinurbanrivers)inthenortheasternregionsofS2(Fig.
3)duetothehighabundanceofOligochaeta(Fig.
S1intheSupplement).
ThemostabundanttaxonomicgroupwasOligochaeta,accountingfor89.
42%ofthetotalmacroinvertebrateabundance(Fig.
S1)followedbyGastropoda(7.
57%),Crustacea(1.
17%),Bivalvia(0.
74%),andChironomidae(0.
65%).
Ofallthe104taxa,L.
hoffmeistericontributed86.
3%ofthetotalmacroinvertebrateabundanceand96.
56%ofthetotalOligochaetaabundance.
Additionally,B.
sower-byiaccountedfor3.
4%ofthetotalOligochaetaabun-dance,andthese2taxatogetheraccountedformostoftheOligochaetaabundance.
Gastropodawasthemostwidelydistributedtaxonomicgroup,withB.
aeruginosarepresenting70.
8%ofthetotalGastro-podaabundance.
P.
eximius(8.
8%),S.
cancelata(6.
4%),Parafossarulusstriatulus(5.
5%),andAlo-cinmalongicornis(3.
6%)werealsoimportanttaxaofGastropoda.
BivalviaweremainlydistributedatthestationaroundLakeTaihu,butcouldnotbefoundinmanystationsinthewestofS1andintheeastofS2.
C.
flumineaandL.
fortuneiwerethemostabun-dantBivalvia,representing51.
8and21.
7%oftotalBivalviaabundance,respectively.
CrustaceawerecommonatthesitesaroundLakeTaihu,andE.
mod-estusandN.
denticulataeachaccountedforhalfoftheabundance.
Odonatamainlyappearedintheup-streampartsofthebasin,andPolychaetaweremorecommondownstream(Fig.
S2intheSupplement).
20Fig.
3.
Spatialpatternsoftaxanumber(taxasamplesite–1)andtotalabundanceofbenthicmacroinvertebrates(ind.
m2)Fig.
2.
Principalcomponentanalysis(PCA)plotofthefirst2principalcomponentsof23environmentalfactors.
Valuesontheaxesindicatethepercentagesoftotalvariationac-countedforbyeachaxisZhangetal.
:MacroinvertebrateassemblagesinLakeTaihuBasinriversThe4diversityindicesshowedsimilarspatialpatterns,withhighervaluesinthewest(S1)andlowervaluesintheeast(S2)(Fig.
S3intheSupple-ment).
ThemeanShannon-WienervaluesforS11andS12were1.
44and1.
31,whereasforS21andS23theywere0.
84and0.
87.
TheMargalefrichnessindexforS11(2.
55)andS12(2.
30)werealsomuchhigherthanthoseforS21(1.
11)andS23(1.
28).
One-wayANOVAanalyses(significanceatp<0.
05)indicatedthatthe4sub-ecoregionsdif-feredsignificantlyinShannon-WienerandMargalefindices,butnotforSimpsonandPielouindices(Table1).
TheSimpsonvaluesforS11(0.
62)andS12(0.
53)weremuchhigherthanthoseforS21(0.
43)andS23(0.
41).
Theevennessofthemacro-invertebrateassemblageshowedasimilarpattern.
Specifically,thevaluesforthePielouindexinS11(0.
69)andS12(0.
51)werehigherthanthoseforS21(0.
53)andS23(0.
48).
Furthermore,valuesofthe4diversityindiceswererelativelylow,show-inglowbiodiversityofbenthicmacroinvertebratesinthebasin.
DistributionofmacroinvertebrateFFGsThe104taxawerecategorizedasfollows:10SH,13SC,18FC,23GC,and40PR.
SCdominatedthebenthiccommunitiesandaccountedfor50%ofthetotalabundance,followedbyGC(25%),FC(13%),andSH(11%)(Fig.
S4intheSupplement).
AlthoughPRhadthemosttaxa,itrepresentedlessthan2%oftheabundanceoftheinvertebratecommunity.
Specifically,B.
aeruginosaalsocontributed70%tothetotalabundanceofSC.
Thus,SCwerethemostwidelydistributedandmadeupthelargestpropor-tion.
L.
hoffmeistericontributed96%tothetotalabundanceofGC,thedistributionofwhichwasinaccordwiththatofL.
hoffmeisteri.
FFGhaddifferentdistributionpatternsindifferentsub-ecoregions.
Specifically,SCrepresented45.
1%ofthetotalabundanceinS11.
ThepercentagesofGCandFCweresmaller(24.
2and20.
4%,respectively),andSHandPRaccountedfor7.
0and3.
3%oftheabundance.
SCwasalsothemainfunctionalgroupinS12andmadeup60.
0%oftheabundance,followedbyFC(15.
1%)andSH(14.
5%),butGC(6.
93%)andPR(3.
53%)wererarelyfoundinthissub-ecoregion.
Onthecontrary,GCwerethemostdominantFFGinS21andS23,accountingfor82.
0and82.
4%abun-dance,respectively.
Overall,thehighestpercentagesofSCwereinuplandstreamsectionswhileGCdom-inatedfurtherdownstream.
MultivariateanalysesOne-wayANOSIMindicatedthatthereweresig-nificantdifferencesinmacroinvertebratecommunitystructurebetweenS11andS23(p=0.
040),andbe-tweenS12andS21(p=0.
026)(Table2).
AlthoughtherewasnosignificantdifferencebetweenS11andS12inmacroinvertebratecommunitystructure,thedominanttaxaofthese2sub-ecoregionswerequitedifferent.
S21andS23weremainlycharacterizedbyGastropodaandOligochaeta,andtheircontributionsinthese2areaswereverysimilar(Table3).
DCArevealedthataunimodalmodel(i.
e.
CCA)wouldbemoreappropriateforthewholedataset(DCAaxis1length=12.
12).
TheCCAidentified5environmentalvariableshighlycorrelatedwiththemacroinvertebratecommunity.
Thefirstandsecondcanonicalaxesexplained8.
2%(eigenvalueof0.
261)and3.
4%(0.
111)ofthevariationinthetaxadata,respectively(Table4).
MonteCarlopermutationtestsrevealedsignificantspeciesenvironmentcorrela-tionsforthefirst2axes(p<0.
05).
ThefirstaxiswaspositivelycorrelatedwithhabitatdiversityandSI(intra-setcorrelationsof0.
69and0.
66,respectively),andnegativelycorrelatedwithTN(r=0.
53).
Thisaxismainlyreflectedhabitatdiversity,substrate,andnutrientgradient.
Axis2showedapositivecorrela-tionwithturbidity(r=0.
43).
OntheCCAplot,stationsinS11andS12wereplottedmainlyonthepositiveside,whereasstationsbinS21andS23wereclusteredonthenegativesideoftheplot.
TheseplacementsindicateddifferentenvironmentalconditionsalongAxis1fortheecore-gions(Fig.
4a).
Moreover,S11andS12hadhighervaluesofSIandhabitatdiversitythanS21andS23,whereasS21andS23hadhigherTNconcentrations.
TheCCAalsorevealedrelationshipsamong32ben-thictaxaandenvironmentvariables(Fig.
4b).
TwoOligochaetataxa(L.
hoffmeisteriandB.
sowerbyi)and2Polychaetataxa(NephtysoligobranchiaandNereisjaponica)occurredatsiteswithhighTN21S11S12S21S23S1179.
782.
380.
42S120.
05581.
3378.
19S210.
1230.
132*75.
31S230.
194*0.
0970.
017Table2.
One-wayANOSIMshowingsignificancelevelsinmacroinvertebratecommunitystructureamongthe4sub-ecoregions(seeFig.
1).
Uppertriangularmatrixshowsthedissimilarity(%),andlowertriangularmatrixshowstheRstatistic;*p<0.
05AquatBiol23:15–28,2014andlowhabitatdiversity,whereasAciagrionsp.
(Odonata)wasfoundatsiteswithlowTNandhighhabitatdiversity.
L.
fortunei(Bivalvia)occurredatsiteswithhighturbidity.
OneChironomidaetaxa(Procla-diussp.
)and2Gastropodataxa(RadixswinhoeiandStenothyraglabra)werefoundatsiteswithhighSI.
Inconsiderationofthesignificantspatialdifference,themacroinvertebratecommunitiesofthe2ecore-gionswereanalyzedseparately(Fig.
5).
Whenthe2ecoregionswereanalyzedseparately,DCAindicatedthatCCAandRDAwouldbemoreappropriateforS1(DCAaxis1length=6.
05)andS2(DCAaxis1length=2.
9),respectiv-ely.
TheCCAidentified3environmentalvari-ables(vegetation,chan-nel,andSI)thatweresignificantlycorrelatedwiththebenthiccom-munityoftheS1ecore-gion.
Thefirst3CCAaxesexplained10.
2%(eigenvalueof0.
368),6.
9%(0.
249)and5.
5%(0.
180)ofspeciesdatavariance,respectively.
AsfortheS2ecoregion,RDAindicatedthat5environmentalvariables(vegetation,landuse,TP,turbidity,andCODMn)werehighlycorrelatedwithmacroinvertebratecommunities.
Thefirst2CCAaxesexplained14.
5and5.
5%ofspeciesdatavariance,witheigen-valuesof0.
145and0.
055,respectively.
DISCUSSIONBenthicmacroinvertebratecharacteristicsAtotalof104macrozoobenthictaxawererecordedinthisstudy.
Comparedwithpreviousstudies,113macroinvertebratetaxaweredetectedbyWangetal.
(2007)inthisregion.
ItmustbepointedoutthatwhenWangetal.
(2007)sampledupstreamintheChangzhouarea,thesamplingsitesweremainlylo-catedinstreams,somoreaquaticinsectswerefound.
ThisstudyfoundmoretaxathanGaoetal.
(2011)andWuetal.
(2011)duetoinclusionmoreriversandstreamsinthebasinandhighersamplingefforts.
MacroinvertebrateFFGpropor-tionswerelikelyaffectedbyan-thropogenicalterationofriparianvegetation(Compin&Céréghino2007).
Inthisstudy,thepercent-agesofGCincreasedfromup-streamtodownstream,reflectingthatGCwereabletofindsufficient22Axis1Axis2Axis3Axis4TotalinertiaEigenvalues0.
2610.
1110.
0670.
0333.
202Speciesenvironmentcorrelations0.
8220.
6270.
5120.
468Cumulativepercentagevarianceofspeciesdata8.
211.
613.
714.
7ofspeciesenvironmentrelationship52.
474.
788.
294.
8Sumofallcanonicaleigenvalues0.
498Intra-setcorrelationsTotalnitrogen0.
530.
140.
290.
10Chla0.
160.
040.
060.
37Turbidity0.
040.
430.
280.
19Habitat0.
690.
240.
020.
07SI0.
660.
230.
140.
16Table4.
Summarizedresultsofcanonicalcorrespondenceanalysisandmacro-invertebrateabundancedataof93stations(taxaoccurringatlessthan5siteswereexcluded).
Intra-setcorrelationsbetweenthefirst4canonicalaxesandtheenvironmentalvariablesarepresentedS11S12S21S23TaxaAbun.
Contr.
Abun.
Contr.
Abun.
Contr.
Abun.
Contr.
OligochaetaBranchiurasowerbyi1.
56.
199.
011.
1Limnodrilushoffmeisteri3.
07.
31289.
320.
01318.
418.
2CrustaceaNeocaridinadenticulata9.
612.
417.
86.
9sinensisExopalaemonmodestus1.
84.
35.
05.
9ChironomidaeCricotopussylvestris0.
93.
7Ploypedilumscalaenum5.
84.
1GastropodaBellamyaaeruginosa42.
139.
545.
631.
970.
943.
267.
947.
4Parafossaruluseximius3.
68.
114.
17.
87.
08.
9Radixswinhoei7.
38.
0Semisulcospiracancelata14.
013.
2Stenothyraglabra2.
15.
1BivalviaCorbiculafluminea1.
56.
213.
57.
2Total65.
080.
0105.
183.
91473.
282.
11398.
380.
4Table3.
Characteristicspeciesforeachsub-ecoregion(seeFig.
1)identifiedbySIMPERpro-cedure,theircontributions(contr.
,%)towithin-groupsimilarity(uptoacumulativepercent-ageof80%),andaverageabundance(abun.
,ind.
m2)ineachsub-ecoregionwerecalculatedZhangetal.
:MacroinvertebrateassemblagesinLakeTaihuBasinriversfoodinurbanstreams(Suren&McMurtrie2005,Compin&Céréghino2007).
Wangetal.
(2012)reportedthattherelativeabundanceofGCwassig-nificantlyhigherindisturbedstreams.
Similarpat-ternshavealsobeenfoundinotherregionsofChina(Jiangetal.
2011),consistentwiththerivercontin-uumconcept(RCC)(Vannoteetal.
1980)thatlongi-tudinaldistributionsofFFGsfollowlongitudinalpatternsinbasalresources.
Inaddition,LimnodrilushoffmeisteriwasthemostdominantGC,anditsdis-23Fig.
4.
Canonicalcorrespondenceanalysisofmacroinvertebratetaxaandenvironmentalvariablesshowing(a)stationscoresand(b)speciesscores.
TN=totalnitrogen,SI=substrateindex,X1=Bellamyaaeruginosa,X2=Limnodrilushoffmeisteri,X3=Corbiculafluminea,X4=Branchiurasowerbyi,X5=Parafossaruluseximius,X6=Exopalaemonmodestus,X7=Semisul-cospiracancelata,X8=Neocaridinadenticulatasinensis,X9=Anodontawoodianaelliptica,X10=Acuticostachinensis,X11=Parafossarulusstriatulus,X12=Radixswinhoei,X13=Uniodouglasiae,X14=Alocinmalongicornis,X15=Anodontawood-ianapacifica,X16=Glossiphoniacomplanata,X17=Limnopernafortunei,X18=Stenothyraglabra,X19=Ploypedilumscalaenum,X20=Chironomusplumosus,X21=Glossiphoniasp.
,X22=Dicrotendipuslobifer,X23=Procladiussp.
,X24=Macrobrachiumnipponense,X25=Aciagrionsp.
,X26=Nephtysoligobranchia,X27=Nereisjaponica,X28=Helobdellafusca,X29=Cricotopussylvestris,X30=Glyptotendipestokunagai,X31=Hippeutiscantori,X32=Physasp.
Fig.
5.
(a)CanonicalcorrespondenceanalysisbiplotofwesternhillaquaticecoregionS1and(b)redundancyanalysisbiplotofeasternplainaquaticecoregionS2.
SI=substrateindex,TP=totalphosphate,CODMn=chemicaloxygendemandAquatBiol23:15–28,2014tributionalsosupportstheideathattaxainhigh-orderriversaremoretoleranttodisturbanceandorganicpollutionthanthoseinlow-orderrivers(Jiangetal.
2011).
Incontrast,SCweremoreabun-dantinheadwaters,anddecreasedgraduallywithincreasingstreamsize,thedistributionofwhichalsofollowedthepredictionsoftheRCC.
Multivariateanalysesindicateddistinctspatialpat-ternsinbenthiccommunitystructure,coincidingwithlocationsfromwesttoeastandreflectingassoci-atedenvironmentalconditions.
MacroinvertebrateassemblagesinthebasinweremainlydominatedbyOligochaeta(e.
g.
L.
hoffmeisteriandBranchiurasowerbyi),Gastropoda(e.
g.
Bellamyaaeruginosa)andChironomidae(e.
g.
ChironomusplumosusandPolypedilumscalaenum).
Generally,thesetaxaareabletoliveinstressfulenvironmentsandareknowntooccurinhighlydegradedsystems(Wangetal.
2012).
Accordingly,L.
hoffmeisteriandB.
sowerbyi,themostrepresentativeOligochaetawithinthebasin,werewidelyspreadandoccurredabundantlyinsomesites.
These2Oligochaetataxaarereportedtobeabletotolerateextremelyoxygen-deficientenvi-ronments(Takamuraetal.
2009).
FiveGastropodataxawerealsocharacteristictaxa,amongwhichB.
aeruginosawasthemostfrequent(Table3).
Bel-lamyaspp.
havebeenreportedtobeinsensitivetopollutionandabletoinhabitmoderatelytohighlypollutedwaterbodies(Cao&Jiang1998).
Forexam-ple,Bellamyapurificataisnotonlyabletolivenor-mallyinconditionswithextremelyhightotalnitro-gencontent(2.
77%),butalsoinareaswithextremelyhighbiomass(428gm2)(Cao&Jiang1998).
Chi-ronomusspp.
larvaearewidelydistributedinthebasin,andthehighdensitiesofthesetaxahasbeenregardedasanexcellentbio-indicatoroffreshwaterpollution(Hooperetal.
2003).
Thelowtaxonrichnessanddominanceofpollution-toleranttaxaindicatesseriousenvironmentalpollutioninthebasin.
Com-paredwiththehighlydevelopedLakeTaihuBasin,theQiantangRiverBasin,withhighforestcoverandlessanthropogenicimpact,containsmoreOdonataandEPT(Ephemeroptera,Plecoptera,andTricho-ptera)(Wangetal.
2012).
Thesetaxaareknowntobesensitivetoexternalinfluence(Jowettetal.
1991)andaremainlyfoundintheupstreampartsofthebasin.
Thismaybeascribedtomoreanthropogenicdisturbanceintheeasternplainareasthanthewest-ernhillareas.
Corbiculaflumineawasfoundmainlyinsiteswithhighturbidityandlownutrientloads.
Thistaxonisreportedtobesensitivetoenvironmen-talchanges,andlowDOlevelsmightstronglyinflu-enceitssurvival(Saloom&Duncan2005).
Themacroinvertebratediversityindicesshowedapatternofhighvaluesinthewestandlowvaluesintheeast.
Inparticular,theShannon-WienerandMar-galefindicesforthewesternhillaquaticecoregion(S1)weremuchhigherthanthatfortheeasternplainaquaticecoregion(S2).
Thismaybeascribedtothedifferenturbanizationandanthropogenicdisturbancelevelsinthe2ecoregions.
PrimaryfactorsgoverningthebenthicmacroinvertebrateassemblagesOurresultssuggestthatthemostimportantfactorsregulatingassemblagestructurearehabitathetero-geneity,organicenrichment,andaquaticvegetationcoverage.
Habitatheterogeneityisknowntobeoneofthedeterminatefactorsofbenthicmacroinverte-bratecommunitystructure(Shostell&Williams2007).
Inthisstudy,thehabitatheterogeneityinS1,amostlyhillyandmountainousregion,wasrelativelyhigherthanthatinS2,wherethelandscapeconsistsmainlyofplainsanddenserivernetworks.
ThemacroinvertebratecommunityinS1wasalsomoreabundantanddiverse.
Macroinvertebratebiodiver-sityismainlydeterminedbythenumbersoftaxaandindividuals,andhigherdiversitycanbedetectedincomplexhabitatsbecauseofmorelivingspaceorsurfacearea(Shostell&Williams2007).
Therefore,themorestructurallycomplexthehabitat,themorediversemacroinvertebratecommunitycanbe(McGoffetal.
2013).
Substrategrainsizeandheterogeneityalsoaffectedthediversityofbenthicassemblages(Allen&Vaughn2010).
Inthisstudy,RadixswinhoeiandStenothyraglabraweremainlyfoundinsiteswithhighSIvalues,becauseoftheirhabitofclimb-ingonhardsubstrate.
Moreover,turbidityisgener-allyconsideredasurrogatemeasureforsedimentloading,andhasbeenfoundtobecloselyrelatedtotheabundanceanddiversityofbenthicmacroinverte-brates(Hall&Killen2005).
Ourresultsalsoindicatedthatthecompositionofthemacroinvertebratecommunitywasinverselyrelatedtothelevelofnutrientenrichment(asindi-catedbymeasurementsofthewatercolumnTP,CODMn,andchla),whichiscongruentwithotherstudies(Donohueetal.
2009,Yuan2010,Pokornyetal.
2012).
Moreover,greatimpactsofnutrientloadsonbenthiccommunitystructurewerepreviouslyre-portedinthisregion(Gaoetal.
2011,Wuetal.
2011).
Nutrientloadscouldinfluencemacroinvertebratecommunitystructureasaconsequenceofincreasedfoodavailabilitywherenutrientsstimulateprimary24Zhangetal.
:MacroinvertebrateassemblagesinLakeTaihuBasinriversproduction.
Theintermediateproductivityhypothe-sis(IPH)predictsthatspeciesdiversityismaximizedatsomeintermediatelevelofproductivity,atwhichcompetitionforfoodisreducedandthecoexistenceofpotentiallycompetingspeciesispromoted(Widdi-combe&Austen2001).
Themostresistanttaxawouldtendtoincreasesignificantly,excludingthemostsensitiveones,andthecommunitywouldeventuallyhavefewertaxabutagreaternumberofindividuals.
Itwouldalsoincreaseeutrophicprocessesandor-ganicenrichmentresultinginthereductionofDO,whichmaybecomelimitingtothesurvivalofsomesensitivetaxa.
Aquaticvegetationalsohasanimportantroleinstructuringmacroinvertebrateassemblagesviapro-vidinglivingspace(Angradietal.
2001)orselectingspeciestraitsrelatedtopopulationdynamicsandfeedinghabits(Céréghinoetal.
2008).
Inaddition,therootsofsomemacrophytesmightprovidebenthicmacroinvertebrateswithDO(Takamuraetal.
2009).
TheriverchannelsinS2haveundergonehighlevelsofregulationandchannelization,whichmightalsoexplainthelowbiodiversityinthisecoregion.
ThisresultisinagreementwiththefindingsofKennedy&Turner(2011),whoreportedthatriverregulationandchannelizationcanreducemacroinvertebratediversityanddensity.
Regulatedandchannelizedriverchannelsisolateriversfromsurroundingripar-ianareasbyseveringlinkagesbetweentheaquaticandripariancommunitiesandbreakingtheintegrityofecologicalstructuretoacertainextent.
Thus,riverchannelizationprocessesandstructuralheterogene-itywereidentifiedasimportantfactorsaffectingtheabundanceandspeciesrichnessofresidentmacro-invertebrateassemblages(Leporietal.
2005).
Therewasadistinctgeographicaldifferencebe-tweentheadjacentS1andS2ecoregions,andtheirbenthicmacroinvertebratecommunitiesreflectedsig-nificantdifferencesbetweenthem.
Specifically,themostsensitivefactorsformacroinvertebrateassem-blagesinS1werehabitatconditionsandaquaticvegetationcoverage,whereasthebenthicmacro-invertebrateassemblagesinS2weremutuallyinflu-encedbynutrientenrichment,habitatheterogeneity,andturbidity.
Previousstudieshaveindicatedthatnaturalgeologicandgeographicfeaturescouldbeimportantfactorsinfluencingmacroinvertebratecommunitystructureacrosslarge-scaleregionaldis-tances(Lietal.
2001,Weigeletal.
2003).
Geographicdifferentiationbetweenthe2ecoregionsmightaffecthumanactivityandlandusepatterns,which,inturn,magnifyregionaldistinctions(Allan2004).
Withlessimpactsofurbanization,S1hadrelativelybetterenvironmentalconditions,lowernutrientconcen-trationsandmorediversehabitat.
MoretaxawerefoundinS1,includingseveralsensitivetaxa.
TheurbanizationlevelinS2showedanincreasingtrendfromwesttoeast,andtherewasalsoincreasingtrendsfornutrientloadsandhabitatdegradationfromwesttoeast.
Themajorcities,withdenserhumanpopulations,aremainlydistributedinthenortheasternareas(Fig.
1).
Urbanriversaremoreregulatedandchannelized,andresultinlessaquaticvegetation.
Accordingly,macroinvertebrateassem-blagesaresubjecttosevereanthropogenicpressuresandwaterpollution,asillustratedbythelowerdiversityanddominanceofpollution-toleranttaxa.
ThenumberoftaxafoundateachstationandthemacroinvertebratediversityindicesinS2decreasedgraduallyfromwesttoeast,whiletheabundanceofOligochaetashowedtheoppositetrend.
ExtremelyhighabundanceofOligochaetawasfoundinseveralurbanriverssites.
HighabundanceofOligochaeta,especiallyL.
hoffmeisteriandB.
sowerbyi,arere-gardedasexcellentbio-indicatorsofhighlypollutedecosystems(Takamuraetal.
2009).
ThepercentagevariationofbenthicfaunaexplainedbyCCAwaslow,whichisoftenthecasewitheco-logicaldata,andthecumulativepercentagevarianceofspeciesdatawas14.
7%ofall4axes(kland1999).
Moreover,theenvironmentalvariablesusedinthispaperaremainlyphysico-chemicalpara-metersofwaterandlimitedhabitatdata.
Wedidnotincludesedimentdatasuchassedimentorganicmattercontent,andconnectivityofriversandlakes.
Allthesefactorswouldsignificantlyaffectthecom-positionofbenthicmacroinvertebrates.
Forinstance,hydrologicalconnectivitycanstronglyinfluencemacroinvertebrateassemblages(Gallardoetal.
2008,Leigh&Sheldon2009).
Inthisstudy,Nephtysoligo-branchiaandNereisjaponicawereonlyfoundinthoseriversconnectedwiththeYangtzeRiverorcoastalrivers,whichmayberelatedtothedispersalprocessandsalinity.
SynthesisandimplicationsforbenthicbiodiversityconservationOurresultsindicatedthattheLakeTaihuBasinmacroinvertebrateassemblagewasmainlycharac-terizedbypollution-toleranttaxa,andthemacroin-vertebratediversitywasrelativelylow,indicatingsevereanthropogenicdisturbanceandhabitatdegra-dation.
TheShannon-WienerandMargalefindicesshowedasignificantdifferencebetweenthewestern25AquatBiol23:15–28,2014hillaquaticecoregionandtheeasternplainaquaticecoregion.
Theresultsalsorevealedthedetrimentalimpactsofanthropogenicdisturbance(suchasin-creasednutrientconcentration,degradedhabitatandaquaticvegetation)onmacroinvertebrateassem-blages.
Atpresent,nutrientenrichmentandhabitatdegradationappeartobethemostpervasiveanthro-pogenicstressesonfreshwaterecosystemsinChina,resultingfromindustrialandagriculturaldevelop-ment(Fangetal.
2006).
Understandingtherela-tionshipsbetweennutrientconcentrationsandthemacroinvertebrateassemblagesisimportantforbio-diversityrestoration,andcontrollingtheinflowofnutrientswouldhelpacceleratetherestorationpro-cess.
Therestorationofaquaticmacrophytescouldbeachievedthroughincreasinghabitatheterogene-ity.
Macrophytescanalsoremovenutrientsfromthewater,andcanbeharvestedtoeliminatethemfromtheecosystem(Takamuraetal.
2003).
Itisofvitalimportancetoreducechannelizationandmaintainconnectivitybetweentheaquaticandterrestrialenvironments(Kennedy&Turner2011).
Ecologicalrestorationwillhavelittlebeneficialeffectonmacro-invertebratebiodiversityiftherestorationschemesdonotaccountforriverregulationandchanneliza-tion,whichaffectstructuralheterogeneityrelevanttothebenthicmacroinvertebrates(Leporietal.
2005).
Ourresultsprovidevaluableinformationforthecon-servationandmanagementofbiodiversityindevel-opedanddevelopingareas.
Weproposethatpar-ticularattentionshouldbepaidtonutrientreductionandrestorationofhabitatheterogeneity.
Acknowledgements.
ThisworkwasfinanciallysupportedbytheMajorScienceandTechnologyProgramforWaterPollu-tionControlandTreatment(Grants2012ZX07501001-03,2014ZX07101011)andNationalNaturalScienceFoundationofChina(Grant51279060).
LITERATURECITEDAllanJD(2004)Landscapesandriverscapes:theinfluenceoflanduseonstreamecosystems.
AnnuRevEcolSyst35:257284AllenDC,VaughnCC(2010)Complexhydraulicandsub-stratevariableslimitfreshwatermusselspeciesrichnessandabundance.
JNAmBentholSoc29:383394AnW,HuJ(2006)EffectsofendocrinedisruptingchemicalsonChina'sriversandcoastalwaters.
FrontEcolEnviron4:378386AngradiTR,HaganSM,AbleKW(2001)Vegetationtypeandtheintertidalmacroinvertebratefaunaofabrackishmarsh:Phragmitesvs.
Spartina.
Wetlands21:7592APHA(2012)Standardmethodsfortheexaminationofwaterandwastewater.
AmericanPublicHealthAssocia-tion,Washington,DCBarbourMT,GerritsenJ,SnyderB,StriblingJ(1999)Rapidbioassessmentprotocolsforuseinstreamsandwadeablerivers:periphyton,benthicmacroinvertebrates,andfish,2ndedn.
USEnvironmentalProtectionAgencyOfficeofWater,Washington,DCBirkS,BonneW,BorjaA,BrucetSandothers(2012)ThreehundredwaystoassessEurope'ssurfacewaters:analmostcompleteoverviewofbiologicalmethodstoimplementtheWaterFrameworkDirective.
EcolIndic18:3141CaiY,GongZ,QinB(2011)Influencesofhabitattypeandenvironmentalvariablesonbenthicmacroinvertebratecommunitiesinalargeshallowsubtropicallake(LakeTaihu,China).
IntJLimnol47:8595CaiY,GongZ,QinB(2012a)BenthicmacroinvertebratecommunitystructureinLakeTaihu,China:effectsoftrophicstatus,wind-induceddisturbanceandhabitatcomplexity.
JGtLakesRes38:3948CaiY,JiangJ,ZhangL,ChenY,GongZ(2012b)Simplifica-tionofmacrozoobenthicassemblagesrelatedtoanthro-pogeniceutrophicationandcyanobacterialbloomsin2largeshallowsubtropicallakesinChina.
AquatEcosystHealthManag15:8191CaoZG,JiangXP(1998)TheinfluenceofenvironmentalfactorsonBellamyapurificata.
JShanghaiOceanUniv7:200205(inChinesewithEnglishabstract)CéréghinoR,RuggieroA,MartyP,AngélibertS(2008)Influenceofvegetationcoveronthebiologicaltraitsofpondinvertebratecommunities.
IntJLimnol44:267274ClarkeKR(1993)Non-parametricmultivariateanalysesofchangesincommunitystructure.
AustJEcol18:117143ClarkeKR,WarwickRM(2001)Changesinmarinecom-munities:anapproachtostatisticalanalysisandinter-pretation,2ndedn.
PRIMER-E,PlymouthCompinA,CéréghinoR(2007)Spatialpatternsofmacro-invertebratefunctionalfeedinggroupsinstreamsinrelationtophysicalvariablesandland-coverinsouth-westernFrance.
LandscapeEcol22:12151225CovichA,AustenM,BaerlocherF,ChauvetEandothers(2004)Theroleofbiodiversityinthefunctioningoffreshwaterandmarinebenthicecosystems.
Bioscience54:767775CumminsKW,KlugMJ(1979)Feedingecologyofstreaminvertebrates.
AnnuRevEcolSyst10:147172DonohueI,JacksonAL,PuschMT,IrvineK(2009)Nutrientenrichmenthomogenizeslakebenthicassemblagesatlocalandregionalscales.
Ecology90:34703477FangJ,WangZ,ZhaoSQ,LiYKandothers(2006)Biodiver-sitychangesinthelakesoftheCentralYangtze.
FrontEcolEnviron4:369377FuBJ,ZhuangXL,JiangGB,ShiJB,LüYH(2007)Environ-mentalproblemsandchallengesinChina.
EnvironSciTechnol41:75977602GallardoB,GarcíaM,Cabezas,GonzálezE,GonzálezM,CiancarelliC,ComínFA(2008)Macroinvertebratepat-ternsalongenvironmentalgradientsandhydrologicalconnectivitywithinaregulatedriver-floodplain.
AquatSci70:248258GaoJF,GaoYN(2012)DelineationofaquaticecoregionsinTaihuLakeBasin.
ChinaEnvironmentalSciencePress,Beijing(inChinese)GaoX,NiuCJ,HuZJ(2011)MacrobenthoscommunitystructureanditsrelationswithenvironmentalfactorsinTaihuRiverbasin.
ChinJApplEcol22:33293336(inChinesewithEnglishAbstract)26Zhangetal.
:MacroinvertebrateassemblagesinLakeTaihuBasinriversHallLW,KillenWD(2005)Temporalandspatialassessmentofwaterquality,physicalhabitat,andbenthiccommuni-tiesinanimpairedagriculturalstreaminCalifornia'sSanJoaquinValley.
JEnvironSciHealthPartAToxic-HazardSubstEnvironEng40:959989Hammer,HarperD,RyanP(2001)PAST:paleontologicalstatisticssoftwarepackageforeducationanddataanalysis.
PalaeontolElectronica4:19HooperHL,SiblyRM,HutchinsonTH,MaundSJ(2003)Theinfluenceoflarvaldensity,foodavailabilityandhabitatlongevityonthelifehistoryandpopulationgrowthrateofthemidgeChironomusriparius.
Oikos102:515524JiangX,XiongJ,XieZ,ChenY(2011)Longitudinalpat-ternsofmacroinvertebratefunctionalfeedinggroupsinaChineseriversystem:atestforrivercontinuumconcept(RCC).
QuatInt244:289295JowettIG,RichardsonJ(1990)MicrohabitatpreferencesofbenthicinvertebratesinaNewZealandriverandthedevelopmentofin-streamflow-habitatmodelsforDelea-tidiumspp.
NZJMarFreshwRes24:1930JowettIG,RichardsonJ,BiggsBJF,HickeyCW,QuinnJM(1991)MicrohabitatpreferencesofbenthicinvertebratesandthedevelopmentofgeneralisedDeleatidiumspp.
habitatsuitabilitycurves,appliedtofourNewZealandrivers.
NZJMarFreshwRes25:187199KennedyTL,TurnerTF(2011)Riverchannelizationreducesnutrientflowandmacroinvertebratediversityattheaquaticterrestrialtransitionzone.
Ecosphere2:art35KondolfGM(1997)Applicationofthepebblecountnotesonpurpose,method,andvariants.
JAmWaterResourAssoc33:7987LeighC,SheldonF(2009)HydrologicalconnectivitydrivespatternsofmacroinvertebratebiodiversityinfloodplainriversoftheAustralianwet/drytropics.
FreshwBiol54:549571LeporiF,PalmD,BrnnsE,MalmqvistB(2005)Doesresto-rationofstructuralheterogeneityinstreamsenhancefishandmacroinvertebratediversityEcolAppl15:20602071LiJ,HerlihyA,GerthW,KaufmannP,GregoryS,UrquhartS,LarsenDP(2001)Variabilityinstreammacroinverte-bratesatmultiplespatialscales.
FreshwBiol46:8797LiT,ShillingF,ThorneJ,LiF,SchottH,BoyntonR,BerryAM(2010)FragmentationofChina'slandscapebyroadsandurbanareas.
LandscapeEcol25:839853LiuY,ZhangW,WangY,WangE(1979)EconomicfaunaofChina:freshwatermollusks.
SciencePress,Beijing(inChinese)LiuB,LiuH,ZhangB,BiJ(2013)ModelingnutrientreleaseintheTaiLakebasinofChina:sourceidentificationandpolicyimplications.
EnvironManag51:724737McGoffE,AroviitaJ,PilottoF,MilerOandothers(2013)AssessingtherelationshipbetweentheLakeHabitatSurveyandlittoralmacroinvertebratecommunitiesinEuropeanlakes.
EcolIndic25:205214MerrittRW,CumminsKW(2008)AnintroductiontotheaquaticinsectsofNorthAmerica,4thedn.
KendallHunt,Dubuque,IAMorseJC,YangL,TianL(1994)AquaticinsectsofChinausefulformonitoringwaterquality.
HohaiUniversityPress,NanjingMorseJC,BaeYJ,MunkhjargalG,SangpradubNandothers(2007)Freshwaterbiomonitoringwithmacroinverte-bratesinEastAsia.
FrontEcolEnviron5:3342klandRH(1999)Onthevariationexplainedbyordinationandconstrainedordinationaxes.
JVegSci10:131136PignataC,MorinS,ScharlA,TraversiDandothers(2013)ApplicationofEuropeanbiomonitoringtechniquesinChina:AretheyausefultoolEcolIndic29:489500PokornyN,FeslC,SchagerlM,WaringerJ(2012)Factorsstructuringmacroinvertebratecommunitiesinahead-waterstreamanditsspringsintheUpperAustrianLime-stoneAlps.
RiverResAppl28:637–646RicciardiA,RasmussenJB(1999)ExtinctionratesofNorthAmericanfreshwaterfauna.
ConservBiol13:12201222SaloomME,DuncanSR(2005)Lowdissolvedoxygenlevelsreduceanti-predationbehavioursofthefreshwaterclamCorbiculafluminea.
FreshwBiol50:12331238ShostellJM,WilliamsBS(2007)Habitatcomplexityasadeterminateofbenthicmacroinvertebratecommunitystructureincypresstreereservoirs.
Hydrobiologia575:389399SmithVH(2003)Eutrophicationoffreshwaterandcoastalmarineecosystemsaglobalproblem.
EnvironSciPollutRes10:126139SurenAM,McMurtrieS(2005)Assessingtheeffectivenessofenhancementactivitiesinurbanstreams.
II.
Responsesofinvertebratecommunities.
RiverResAppl21:439453TakamuraN,KadonoY,FukushimaM,NakagawaM,KimBH(2003)Effectsofaquaticmacrophytesonwaterqual-ityandphytoplanktoncommunitiesinshallowlakes.
EcolRes18:381395TakamuraN,ItoT,UenoR,OhtakaAandothers(2009)EnvironmentalgradientsdeterminingthedistributionofbenthicmacroinvertebratesinLakeTakkobu,Kushirowetland,northernJapan.
EcolRes24:371381TangHQ(2006)BiosystematicstudyonthechironomidlarvaeinChina(Diptera,Chironomidae).
PhDthesis,NankaiUniversity(inChinesewithEnglishabstract)terBraakC,SmilauerP(2002)CANOCOreferencemanualandCanoDrawforWindowsuser'sguide:softwareforcanonicalcommunityordination(version4.
5).
Micro-computerPower,Ithaca,NYVanniMJ(2002)Nutrientcyclingbyanimalsinfreshwaterecosystems.
AnnuRevEcolSyst33:341370VannoteRL,MinshallGW,CumminsKW,SedellJR,Cush-ingCE(1980)Therivercontinuumconcept.
CanJFishAquatSci37:130137WallaceJB,WebsterJR(1996)Theroleofmacroinverte-bratesinstreamecosystemfunction.
AnnuRevEntomol41:115139WangHZ(2002)Studiesontaxonomy,distributionandeco-logyofmicrodrileoligochaetesofChina,withdescrip-tionsoftwonewspeciesfromthevicinityoftheGreatWallStationofChina,Antarctica.
HigherEducationPress,Beijing(inChinese)WangB,XuD,YangL,ShenL,YuH(2007)CharacteristicsofbenthicmacroinvertebratescommunitiesinrelationtoenvironmentinupperreachesoftheTaihuLakeWater-shedinChangzhouArea.
JEcolRuralEnviron23:4751(inChinesewithEnglishabstract)WangB,LiuD,LiuS,ZhangY,LuD,WangL(2012)Impactsofurbanizationonstreamhabitatsandmacroinverte-bratecommunitiesinthetributariesofQiangtangRiver,China.
Hydrobiologia680:3951WeatherheadMA,JamesMR(2001)Distributionofmacro-invertebratesinrelationtophysicalandbiologicalvari-ablesinthelittoralzoneofnineNewZealandlakes.
Hydrobiologia462:11512927AquatBiol23:15–28,2014WeigelBM,WangL,RasmussenPW,ButcherJT,StewartPM,SimonTP,WileyMJ(2003)Relativeinfluenceofvariablesatmultiplespatialscalesonstreammacro-invertebratesintheNorthernLakesandForesteco-region,USA.
FreshwBiol48:14401461WiddicombeS,AustenMC(2001)Theinteractionbetweenphysicaldisturbanceandorganicenrichment:animpor-tantelementinstructuringbenthiccommunities.
LimnolOceanogr46:17201733WuZ,CaiY,ChenY,ShaoX,GaoJ(2011)AssemblagestructureinvestigationofmacrozoobenthosandwaterqualitybioassessmentofthemainriversystemsinTaihuBasin.
JLakeSci23:686694(inChinesewithEnglishAbstract)XiaoR,SuS,GhadouaniA,WuJ(2013)Spatialanalysisofphytoplanktonpatternsinrelationtoenvironmentalfac-torsacrossthesouthernTaihubasin,China.
StochasticEnvironResRiskAssess27:13471357YoccozNG(1991)Use,overuse,andmisuseofsignificancetestsinevolutionarybiologyandecology.
BullEcolSocAm72:106111YuanLL(2010)Estimatingtheeffectsofexcessnutrientsonstreaminvertebratesfromobservationaldata.
EcolAppl20:110125ZhangW,FengH,ChangJ,QuJ,XieH,YuL(2009)HeavymetalcontaminationinsurfacesedimentsofYangtzeRiverintertidalzone:anassessmentfromdifferentindexes.
EnvironPollut157:1533154328Editorialresponsibility:VictorMeyer-Rochow,Bremen,GermanySubmitted:March24,2014;Accepted:September2,2014Proofsreceivedfromauthor(s):October31,2014