t\ r/j

ECOLOGICALCHARACTERIZATION OFSURFACEWATERSIN THEPROVINCEOFOVERIJSSEL (THENETHERLANDS)

0NTVAM3EM

't\ r / j C8-KAR0EX

0000 0417 4310

l^jSßBOUWUNIVERSITEIJ 8EAGENINGEN

Promotoren:

dr.C.W.Stortenbeker hoogleraar inhetnatuurbehoudennatuurbeheer dr.W.J.Wolff hoogleraar indeaquatischeecologie

/JfJoS2o) \ZSS

ECOLOGICALCHARACTERIZATION OFSURFACEWATERS IN THEPROVINCEOFOVERIJSSEL (THENETHERLANDS)

PietF.M.Verdonschot

Proefschrift terverkrijgingvandegraadvan doctor indelandbouw- enmilieuwetenschappen, opgezagvanderectormagnificus, dr.H.C.vanderPlas, inhetopenbaar teverdedigen opwoensdag30mei1990 desnamiddags tevieruur indeaula vandeLandbouwuniversiteit teWageningen.

/5XA 5 - 0 ^ 7 /

CONTENTS

VOORWOORD (PREFACE)

7

SAMENVATTING

9

SUMMARY

13

1.INTRODUCTION 1.1Background 1.2Almsofthestudy 1.3 Presentationoftheresults 1.4 Designofthestudy 1.4.1 Considerations andbackground 1.4.2 Thepracticaldesign 2.ONTHEPRINCIPLESOFTYPOLOGY 2.1 Introduction 2.2 Community concept,somegeneralconsiderations 2.2.1 Theclassicalcommunity concept 2.2.2 Development ofthecommunity concept 2.3 Developments inlotieecology 2.3.1 Introduction 2.3.2 Scale 2.3.3 Communityboundaries 2.3.4Region 2.3.5 Concepts inlotieecology 2.4Developments inlenticecology 2.5 Typeasanecologicalentity 2.6Typologymethods 2.7Typology andwatermanagement 3.STUDYAREA,MATERIALSANDMETHODS 3.1 Studyarea 3.1.1 Climate 3.1.2 Geology,soilandmorphology 3.1.3 Hydrology 3.1.4Human impact 3.2Macrofauna samplingandsampleprocessing 3.3Abiotic samplingandsampleprocessing 3.4Dataprocessing 3.4.1 Preprocessing ofthedata 3.4.2 Multivariate analysis techniques 3.4.3 Applicationofmultivariate analysis 4.REPRODUCIBILITYOFAMACROFAUNA SAMPLE 4.1 Introduction 4.2Materialandmethods 4.3 Reproducibility ofpresence

17 17 18 19 20 20 23 28 28 28 28 31 33 33 34 36 37 37 38 39 41 43 46 46 46 46 51 51 57 59 59 59 61 63 67 67 68 69

r jNJo??oi,)^5'5

STELLINGEN

1. Eenecologische typologie iseenstukgereedschap datontworpenis om de werkelijkheid waaraan zij isontleend,verantwoord tekunnen beheren. (ditproefschrift) 2. Dehoofdfactorendiedeverspreidingvandebenthische macrofauna van zoete wateren in Nederland bepalen,zijnstroming,zuurgraad, zoutgehalte,dimensies van het water en frequentie en duur van droogvalling. Binnen dit patroonvanhoofdfactoren zijnvervolgens hetgehalteaanvoedingsstoffenendehabitatstructuur bepalend. (ditproefschrift) 3. Degeleidelijkeverdwijningvansoortenover de laatste honderd jaar uit de Nederlandse rivier-enbeekstelsels gaandevanmonding naarbronvindtbinnenkorthaar trieste dieptepunt als de typische bronbewoners verdwijnen tengevolgevanhetdoorslaanvanfosfaatin dezandgronden. 4. Het "NutrientSpirallingConcept" (Websteret al. 1983) is in natuurlijke laaglandbeken inNederlandvanondergeschiktbelangomdat hetmerendeelvanhunvoedingsstoffen niet wordt opgenomen in het systeemmaar slechtswordt getransporteerd. Webster,J.R., Gurtz,M.E.,Hains,J.J., Meyer,J.L.,Swank,W.T., Waide, J.B. & Wallace, J.B. 1983. Stability of stream ecosystems. In: J.R. Barnes& G.W. Minshall (eds.), Stream ecology. PlenumPress,NewYork: 99-136. 5. Het "RiverContinuumConcept" (Vannote et al. 1980) is niet bruikbaar voor toepassing inhetbeekbeheer inNederland,enerzijds vanwegehetuitgangspunt dat de fysische variabelen van bron tot monding binneneen riviersysteem een continue gradiëntvanfysische omstandigheden bieden en anderzijds vanwege het gehanteerde schaalniveau. Vannote,R.L.,Minshall,G.W., Cummins, K.W., Sedell, J.R. & Cushing, C E . 1980. The River Continuum Concept. Can. J. Fish. Aquat. Sei. 37: 130-137. 6. Hetroutinematig onderzoekenvandewaterkwaliteitkomtvaak neer ophetvragennaardebekendeweg. 7. Met het begrazen, bekalken en bevissen staat het moderne natuurbeheer inNederland terechttussenlandbouwenvisserijin. 8. Metdeaanlegvanooibossen inderivieruiterwaarden zal de kans ophetontstaanvanlokale steekmuggenplagensterktoenemen.

9. Hetverdwijnenvaneenenkeleogenschijnlijk onbelangrijke soort kaneendomino-effectindehelelevensgemeenschap teweegbrengen. Dit maakthetzoekennaareenno-effect-level voor toxische stoffen op ecosysteemniveauvrijweleindeloos. 10. Many ecological concepts are limited explanations of the biological reality around us. This limitation is broughtabout throughdimensionsofspaceand timeandthroughthe the limitations ofdefinitions (May 1984). May,R.M. 1984. Anoverview: Real and apparant patterns in community structure. In: D.R. Strongetal. (eds.), Ecological communities. PrincetonUniv. Press,Princeton : 3-18. 11. Het invoeren van bronvermelding zal beleidsstukken beter verifieerbaar maken waardoor ook de juistheidenkwaliteit zullen toenemen. 12. Integenstelling totwatdenaamsuggereert is de rol van een onderzoekbegeleidingscommissie vaakmeervolgenddanbegeleidend. 13. OverdeecologievandeAMOEBE (DerdeNota Waterhuishouding) is nogweinigbekend. 14. Stellingengaanheteneoor inenhetandereuit.

PietF.M. Verdonschot Ecological characterizationof surface waters in the province of Overijssel (TheNetherlands) Wageningen,30mei1990.

4.4Reproducibility ofabundance 4.5 Reproducibility intypology 4.6 Reproducibility intime 4.7 Conclusions

70 72 73 76

5.MACROFAUNALCOMMUNITYTYPES INHELOCRENESPRINGS

77

5.1 Introduction 5.2Results 5.2.1 Datacollection 5.2.2 Preprocessing ofthedata 5.2.3 Multivariate analysis 5.3 Discussionofdata

77 77 77 79 79 89

6.MACROFAUNALCOMMUNITYTYPESINSTREAMS

93

6.1 Introduction 6.2 Results 6.2.1 Datacollection 6.2.2 Preprocessing ofthedata 6.2.3 Multivariate analysis 6.3 Discussionofdata

93 93 93 94 94 104

7.MACROFAUNALCOMMUNITYTYPES INDITCHES

109

7.1 Introduction 7.2Results 7.2.1 Datacollection 7.2.2 Preprocessing ofthedata 7.2.3 Multivariate analysis 7.3 Discussionofdata

109 109 109 109 109 122

8.MACROFAUNALCOMMUNITYTYPES INRIVERS,CANALS ANDLARGELAKES

126

8.1 Introduction 8.2 Results 8.2.1 Datacollection 8.2.2 Preprocessing ofthedata 8.2.3 Multivariate analysis 8.3 Discussionofdata 9.MACROFAUNALCOMMUNITYTYPESINPONDSANDSMALLLAKES 9.1 Introduction 9.2 Results 9.2.1 Datacollection 9.2.2 Preprocessingofthedata 9.2.3 Multivariate analysis 9.3 Discussionofdata 10.AREGIONALTYPOLOGYOFSURFACEWATERS INTHEPROVINCEOF OVERIJSSEL

126 126 126 126 138 142 142 142 142 142 143 155

161

10.1Introduction 10.2Multivariate analysis 10.2.1DCCA -RUN1 10.2.2DCCA -RUN2 10.2.3DCCA -RUN3 10.2.4DCCA -RUN4 10.2.5DCCA -RUN5 10.2.6Theenvironmental characterizationofthe sitegroups 10.3Thebiological similaritybetweenthesitegroups 11.GENERALAPPLICATIONS INWATERMANAGEMENT 11.1Thewebofcenotypes 11.2Applicationofthewebofcenotypes 12.REFERENCES APPENDICES

161 161 168 175 179 183 187 191 192 198 198 201 206 225

VOORWOORD

Toenikinseptember 1980indienst tradvandeProvincialeWaterstaat in Overijssel,had ikdekeuzeuittweeopvattingenvanmijnfunctie, enerzijdsvragenenproblemenbehandelendieopkorte termijn moesten worden opgelost of anderzijds mijn aandacht richtten op de ontwikkelingvaneenbasisvoorhetoplossenvanvragenen problemen. Het onderwerp vanditproefschrift toontaandatikvoorhetlaatste heb gekozen. Desubgroep StandaardisatievandeWerkgroep Biologische Waterbeoordeling heefteenbelangrijk aandeelgehad indezekeus. De discussies rond de ontwikkeling van een indeling van oppervlaktewateren liepen vooruit ophetbeleid enstimuleerdenmij dezeweg inteslaan. Eenbelangrijkebijdrage aan deze discussies werd ondermeergeleverddoorJeanGardeniers. Sjeng,jebentsteeds bijhetonderzoekbetrokkengeweestenjewas een stimulans bij de gedachtenvorming rondtypologieën. Ookalwas ikindeeerstejarenbijdeafdeling Waterhuishouding een vreemde eend in debijt,tochleiddenmijn,somseigenzinnige, keuzentoteen ecologisering van de waterstaat. Mijn toenmalige collega's brachtten steeds meerhetgeduldopteluisteren. Alleen JanLaseurenEllyvanMourikbehoefdenniettewordenovertuigd, zij steunden het onderzoek vanafhetprillebeginenzorgdenervoordat mijnonderzoekersgeest involdoendemateopdepraktijk gerichtbleef. Deuitvoeringvanmonsterprogramma's, determinaties, bewerkingen en verslaggevingen als onderdelen van dit onderzoek waren niet mogelijk geweestzonderde inzet van een groot aantal studenten, stagiairs, gewetensbezwaarden en tijdelijkemedewerkers. Iedervan hendook, soms letterlijk, in een fysisch-geografisch watertype, Gerard Willemsen indelaagveensloten,JanJanseenDaanMonnikendam indevaarten,BertOude-EgbrinkenJosNotenboom inde bronnen, Wim Heyligers enCyrilLiebrand indevennen,RobGerritsen indemerenen meertjes,Gertie SchmidtenLouistenCate indebeekbovenlopen, Paul Latour in de kanaalbeken, Ton Ruigrok indeslootbeken,Dwightde Vries indeslotenenSteefvanHoof inderivierenriviertjes. Inseptember 1984kreeg ikeenaanstellingbijhet Rijksinstituut voor Natuurbeheer enverplaatste deonderzoekbasiszichnaarLeersum. Desamenwerkingmetdeprovinciebleefenmijnopvolger aldaar, Hans ten Cate, stimuleerde met groot enthousiasme devoortgangvanhet onderzoek. Vanafmijnkomsthebbende medewerkers van de afdeling Hydrobiologie steeds hun directe en indirecte bijdrage aan de voortgangvanhetprojectgeleverd. MetnameBertHigler vormde een positiefkritischekrachtachterdewetenschappelijke funderingvande optestellentypologie. MetdekomstvanJokeSchotwas ikverzekerd van eenmeergeordende laboratoriumorganisatie enkon ikeendeelvan debegeleidingvantijdelijkemedewerkers aan haar overdragen. De aanpak van fysisch-geografische watertypen ging door enweldoor toedoenvanLeonThijssenenIngridMeulemanvoordemiddenlopen,Bert Rademakers voor debenedenlopen,PetervanLeeuwenenGertie Schmidt voor de kanalen, Reinder Torenbeek en Gertie Schmidt voor de tijdreekspunten, Henk Ketelaars voor dedroogvallendewatergangen, HenkWegmanvoorderandmeren,MartijnHokken,PierreVerbraaken Ron Huls voor de zandwinplassen, Gert-Jan Gastvoor dekolkenenoude

rivierarmenenRobbertdeRiddervoordepoelenterwijlJan Buijs en HansBakkerassisteerdenbijdebemonsteringstochten. Danblevennogveelrestantenvanniet-afgemaakte deelonderzoeken enachtergeblevendeterminaties overdiekeurigwerdenweggewerktdoor Marie-JoséRuikenenWalterMommersteeg (watermijten) en Nico Rawee (oligochaeten). Eenaantalmoeilijkdetermineerbareorganismenwerden doorverschillende specialisten geverifieerd. Deverwerkingvandegegevenswaseengeheelaparthoofdstuk. In de eerste jaren droegen Onno van Tongeren,JosNotenboom enStan Tummers ingrotematebijaanhet inzicht in de mogelijkheden van automatisering enverwerkingstechnieken. Deuiteindelijkeverwerking wasnietmogelijk geweestzonderde inzetvanJokeSchotbijdein-en uitvoer van kilometers computerpapier, de gegevenscorrecties van Elbert Brethouwer en Ton van Eijden en de uiterst deskundige begeleiding doorCajoterBraak. Cajo,zonderCANOCOhadditprodukt nooittotstandkunnenkomen. DeeersteruwemanuscriptenzijndoorBertHiglervanzakelijk en helder commentaar voorzien. Cajo ter Braak waakteeroverdatde complexematerievandemultivariate analyseeenduidigoppapierkwam. Zijn persoonlijke inzet zorgde voor een verbetering van de manuscriptendiezekerniet ophield bij de statistische aspecten. Prof. dr. C.W. Stortenbekerwasmijnpromotorenzijncommentaren maaktenmijnmanuscripten toegankelijkvooranderen. Prof. dr. W.J. Wolff wasmijnco-promotor enzijnkritische inbrengheeftdelaatste hindernissenvoordetotstandkoming van het definitieve manuscript weggenomen. Jackie Senior corrigeerdeuiteindelijkmijnsteenkolen engelsenArjanGriffioenmaaktemetveelgedulddetekeningen. Zoalsduidelijkzalzijn,hebbenveleneenbijdrage geleverd aan de totstandkoming van dit proefschrift. Slechts sommige direct betrokkenenheb ikmetnaam kunnen noemen, andere medewerkers van zowel de Provinciale Waterstaat als het Rijksinstituut voor Natuurbeheerhebbeneveneens iederophun eigen wijze een steentje bijgedragen. RalfenJeroenjulliehebbennog geen andere vader gekend dan eentje die een proefschrift probeert teschrijven. Yolande,jouw steunvormdehetbelangrijkste element inhetwelslagenvandit werk. Je leefde mee met defrustraties ensuccessenenvroegnooitomde tijddievooronswasbedoeldmaaraanditwerkwerdbesteed. Mijnbijzonderedankgaatuitnaaraldegenendieditproefschrift hebbenlatenwordentotwathetnu is.

SAMENVATTING

ECOLOGISCHE OVERIJSSEL

KARAKTERISERING

VAN

OPPERVLAKTEWATEREN

IN

Veeloppervlaktewateren InNederland ondergaan een voortschrijdende nivellering vandewaterkwaliteit. Omditprocesvanverarming tegen tegaaniseengedifferentieerde aanpak inhetwaterbeheernodig. De regionale waterbeheerders vormeneengeschiktbestuursniveauomdeze aanpakterealiseren. Mededaarom isdeprovincieOverijssel in 1981 begonnen met het project 'Ecologische karakterisering van oppervlaktewateren inOverijssel (EKOO)'. HetdoelvanhetEKOO-projectishetontwikkelenvaneenregionaal ecologisch referentiekader vanoppervlaktewateren inOverijssel,dat gebaseerd isopdemacrofaunasamenstelling endatalsbasisdientvoor de ontwikkelingvaneenbeleid tenbehoevevanhetwaterbeheer enhet natuurbeheer. Ditreferentiekader dientmetnamealsbasis voor het opstellen van waterbeoordelingsmethoden en van normstellingen in relatiemetmenselijke gebruiksfuncties en typen oppervlaktewateren. Metderesultatenwordt tevensbeoogdmeer inzichtteverkrijgen inde structuurvanmacrofaunagemeenschappen. Eenbijkomend doel van het project is het verderontwikkelenvandetypologischebenadering in hetwaterbeheer. De gepresenteerde typologische benadering is een combinatie van elementen uitdezonerings-endecontinuumbenadering. Dezoneringsencontinuumbenadering zijnbeide afkomstig uit het onderzoek aan stromende wateren. Bij de zoneringsbenaderingwordthet stromende wateronderverdeeld inzones terwijlbij de continuumbenadering het stromende water wordt gezienalseengradient. Bijde typologische benaderingwordensoortencombinatiesbeschrevenalstypen. Dezetypen representeren binnen bepaalde waarden van milieuomstandigheden gemeenschappen,die inelkaarovervloeien. Om verscheidene redenen kan een typeslechtsopregionaalniveau (eenregiometonderandere vergelijkbare meteorologische, geomorfologische en hydrologische kenmerken)wordeningevuld. Destudie isopgezetalseen grofschalig beschrijvend veldonderzoek waarbij de temetenparameterszoveelmogelijk zijngekwantificeerd. De macrofaunasamenstelling (taxonsamenstelling en abundantie) is gekozen als belangrijkste parameter. Daarbij zijn op elke monsterplaatsongeveer 70fysisch,chemisch en biologisch relevante parametersbepaald. Intotaalzijn664monsterplaatsen,verdeeldover ongeveertwintigfysisch-geografische watertypen, bezocht. Hiermee zijn alle regionaalvanbelangzijndehoofdfactoren, inverscheidene combinaties en overgangsvormen, in de studie betrokken. De fysisch-geografische watertypologie isslechtsgebruiktalspraktisch handvattenbehoevevandeuitvoeringvanhet veldwerk maar speelde geen rol bij het opstellen van de ecologische typologie. De ecologische typologie isontleendaan alle verzamelde biotische en

abiotischegegevens. Dereproduceerbaarheidvanhetmacrofaunamonsterdatgenomenwordtmet behulp vanhetstandaardmacrofaunanet, isnaderonderzocht. Dezein Nederlandveelvuldig toegepastemonsternametechniekgeeftslechts een semi-kwantitatief beeld van de meerabundante soortendieaanwezig zijnineenoppervlaktewater. Ongeveer 55%vande taxa aanwezig op het moment vandemonsternamewordtmetdeze techniekverzameld. De toepassingvanhetstandaardmacrofaunanetvoorregionaal typologisch onderzoek is echter gerechtvaardigd, omdatdeverschillenalsgevolg vanbemonstering overseizoenendanwelverschillenalsgevolgvan de bemonsteringstechniek zelf veelgeringerzijndanverschillen tussen de beschreven watertypen. Er kan worden geconcludeerd dat de reproduceerbaarheid van eenstandaardmacrofaunanetmonstervoldoende isvoordegekozentypologischebenadering. Allereerst zijndegegevensafkomstigvan de beschrijvingen van de twintig fysisch-geografische watertypen gecombineerd in vijf hoofdgroepen. Decombinatievanbiotische enabiotische gegevens is voor elkvandezevijfhoofdgroepenstatistischbewerkt. Daarnazijn allebiotische enabiotische gegevensgecombineerd enals een geheel bewerkt om een uiteindelijke ecologische typologie te kunnen opstellen. Voor dit doel zijn multivariate analysetechnieken geschikt. Verschillende multivariate analysetechnieken (clusteranalyse encanonische ordinatietechnieken)zijn toegepast om groepen monsterplaatsen te beschrijven in termen van taxonsamenstelling engemiddeldemilieuomstandigheden. De indelingin groepenmonsterplaatsenopbasisvanmultivariate analyseresultatenis incombinatiemet informatie overdeecologievan de taxa handmatig bijgesteld en verfijnd. Deresulterende groepenmonsterplaatsenmet bijbehorendemacrofaunagemeenschappen wordencenotypengenoemd. De helocrene bronnen zijn verdeeld over zes cenotypen. De belangrijkste verschillen tussen deze cenotypen kunnen worden gerelateerdaandehydrologie (duurvandroogvalling)endezuurgraad. Daarbinnenblijkendehoeveelheidvoedingsstoffen en/ofdehoeveelheid organischmateriaal differentiërend te zijn. Elk cenotype bevat verschillende microhabitats. Deze microhabitats kunnen, op twee uitzonderingenna,wordengerelateerd aande duur van droogvalling. De beide uitzonderingen zijn gerelateerd aan respectievelijk de bronkop en de bronbeek. Sommige onderzochte bronnen lijken waarschijnlijk nog sterk op denatuurlijke referentiesituatievoor helocrenebronnen. Debelangrijkstebedreigingvoorhelocrenebronnen is de invloed van de mens op dechemische samenstellingvanhet grondwater. Inhetwaterbeheer dientdeaandachtdan ook allereerst hiernaaruittegaan. De beken zijn verdeeld over elf cenotypen. De belangrijkste verschillen tussen deze cenotypen kunnenwordengerelateerdaande dimensies,het 'beekkarakter',deduurvandroogvalling endematevan belasting met organisch materiaal. Alle beken wordeninmeerof minderematedoordemensbeïnvloed. Vooralde regulatie van beken heeft grotegevolgengehadvoordebeeklevensgemeenschappen. Slechts 2%vandetotale lengteaanbekenbezitnogeenminofmeernatuurlijk

10

'beekkarakter' en eendaarbijbehorendemacrofaunagemeenschap. Deze 2%zijnslechtsbewaard gebleven als gevolg van hun geografische ligging op steile hellingen. Inhetalgemeendienthetwaterbeheer zichterichtenop het herstel van de fysische en hydrologische eigenschappenvanbeken. De sloten zijn verdeeld over elf cenotypen. De belangrijkste verschillen tussen deze cenotypen kunnenwordengerelateerd aande dimensies, de duur van droogvalling, de zuurgraad en de stroomsnelheid. Daarbinnen blijken de hoeveelheid voedingsstoffen en/ofdehoeveelheid organischmateriaaldifferentiërend tezijn. De taxonsamenstelling vaneenslootblijkteengecombineerdeweergavete zijnvanhetsuccessiestadium inruimte (vormvanhetprofiel) en in tijd (rijpheid). Hierdoor bestaat een overlap in de taxonsamenstelling tussende verschillende cenotypen. Sloten zijn kunstmatige ecosystemendievooralzijngelegeninagrarische gebieden (ditbetekenteentoevoervanvoedingsstoffen)enafhankelijk zijnvan menselijke ingrepen (schonen, baggeren). Bijhetecologischbeheer vanslotendientmenrekening tehoudenmetdevormvan het profiel, hetsuccessiestadium endemenselijkeingrepen. Derivieren,kanalenenrandmerenzijnverdeeld over elf cenotypen. Deze cenotypen vertonen grote overeenkomsten in taxa (meestal opportunisten)enmilieuomstandigheden. De grote dimensies van de stromende, lijnvormige wateren hangen samen met een groot afwateringsgebied. Grotewaterenfungerendanookalsverzamelbakvan voedingsstoffen, organisch materiaalentoxicanten. Ditveroorzaakt eenvoortdurende stress op het ecosysteem en waarschijnlijk het verdwijnen van veeloorspronkelijke soorten. Devoortdurende stress overheerstookdewerkingvannatuurlijkehoofdfactoren zoals stroming en dimensies. Slechts enkeletaxazijnnogkarakteristiekvooreen deelvaneenrivierofeendimensionele gradient ineenkanaal. Depoelenenkleine meren zijn verdeeld over elf cenotypen. De belangrijkste verschillen tussen deze cenotypen kunnen worden gerelateerd aandeduurvandroogvalling,dezuurgraad,devorm en de hoeveelheid voedingsstoffen. Vooral de viercenotypenwaaroverde stilstaande,neutralepoelenenkleinemerenzijn verdeeld, vertonen onderling groteovereenkomsten intaxonsamenstelling. Dezecenotypen vertegenwoordigen eenwebvormigcontinuumwaarindedimensies (relatie tussen breedteendiepte),devoedingsstoffenendesamenstellingvan debodem (vooralhetlaagveen)domineren. Debelangrijkstemenselijke beïnvloedingen zijn wijzigingen in de hydrologie, verzuring en eutrofiëring. Cenotypen beschreven voor de vijf hoofdgroepen van fysisch-geografische watertypen kunnen elkaar sterk overlappen. Daaromdienenallebiotische enabiotische gegevens ook in samenhang te worden bewerkt. Opnieuw ishiervoor clusteranalyse encanonische ordinatie toegepast. Naelkeordinatie zijnhiervoor de eerste twee resulterende assengebruikt. Vervolgens zijndeherkenbare cenotypen afgesplitst enzijnderesterende dataopnieuwbewerkt. Opdezewijze zijn uiteindelijk 42 cenotypen onderscheiden. Dezecenotypenzijn doormiddelvandeecologievandetyperende taxaendewaardenvande

11

belangrijkstemilieuvariabelenbeschreven. Deonderlinge relaties tussendecenotypenworden weergegeven in een hierarchisch dendrogram eneenwebvancenotypen (Figuur 11.1). Hethierarchischdendrogramvancenotypentoontaandat onder andere decenotypen inmidden- enbenedenlopenvangereguleerdebeken,kleine rivieren,sloten en enkele middelgrote, min of meer stilstaande wateren, alle met hypertrofe,mesosaprobeomstandigheden,eengroot aantalovereenkomende taxabezitten. Blijkbaarvermindert menselijke beïnvloeding (zoals regulatie vanbeken,lozingvanafvalstoffenen intensieve agrarische activiteit inhet stroomgebied) het relatieve belangvandenatuurlijkehoofdfactor stroming instromendewaterenen van de natuurlijke hoofdfactoren dimensies (vorm en diepte) en bodemtype instilstaandewateren,watinbeidegevalleneveneens leidt toteenverarming indemacrofauna. Hetwebvancenotypengeeftdeonderlingepositie, de overgangen tussen cenotypen (hetcontinuum)endebelangrijkste milieugradiënten weer. Devierhoofdfactoren indeze typologie zijn 'beekkarakter', zuurgraad,duurvandroogvallingendimensies. Omeenwater,zelfssubjectief,tebeoordelen ishetnoodzakelijk dit water te vergelijken met andere wateren in verschillende omstandigheden. Voordezebeoordelingsschaal iseenreferentienodig. Dezereferentiehoeftgeenweergave tezijnvaneeneindstadium ineen successie ofvaneenoorspronkelijke toestandmaardienteen toestand weer tegevenwaarmeeeenrichtingvoorverbeteringwordtaangegeven. Degepresenteerde regionale ecologische typologiebiedteenhandvatom de mogelijke ontwikkelingsrichting naardereferentievaneenwater aantegeven. Voordekeuzevan de oorspronkelijke referentie kan gebruikwordengemaaktvaninformatieverkregenuitstudiesaanminder beïnvloede wateren zoals natuurlijke beken, oligotrofe vennen, mesotrofe ouderivierarmen,endergelijke. Er worden een aantal voorbeelden gegeven van mogelijke toepassingen van de regionale ecologische typologie in het waterbeheer. Omdatdefactoren in het web van cenotypen op een beschrijvende studiezijngebaseerd, isvoorzichtigheid geboden indien hetweb invoorspellende zin wordt gehanteerd. De typologie moet gezien worden als hulpmiddel dat samen metdejuiste ecologische principesdienttewordentoegepast tenbehoevevan het waterbeheer. Het is een basis die gebruikt kan worden vooronderanderehet monitorenen beoordelen van de waterkwaliteit, het aangeven van potenties van wateren en hetopstellenvanrandvoorwaardenbijhet beheerendeherinrichtingvanwateren.

12

SUMMARY

Nowadaysmany surface waters in The Netherlands tend to become ecologicallyuniformwiththesamemediocre quality. A differentiated approach towatermanagement isnecessary to stop this process of impoverishment of aquatic ecosystems. In The Netherlands the provincial authorities represent theappropriate levelat which this differentiated approach towatermanagementcanbeputintopractice. Suchan approach has been realized by the Department of Water Management oftheProvince ofOverijssel. Asapartofthisapproach theproject 'Ecologicalcharacterization of surface waters in the province ofOverijssel (EKOO)'was formulated in1981. TheaimsoftheEKOO-projectaretodeveloparegional ecological characterizationofsurfacewatersbasedonmacrofaunacompositionand toreachabetterunderstandingofthevariety and the structure of the macrofauna communities present inthewatersoftheprovinceof Overijssel. Theprojectthusprovidesknowledge ofaquatic ecosystems on a regional scale and a basis for the development ofwater managementpolicies. Anadditionalaimofthisstudy is to develop thetypological approachused inwatermanagement. Thetypological approachusedcanbe interpretedasanintegration of the zonalconceptand thecontinuumconcept. Taxoncombinationswill bedescribed as types. Within a limited range of environmental conditions, thesetypesarerepresentative ofcommunities,butatthe sametime they together form a continuum. Due to hierarchical relations between majorenvironmentalvariables (master factors)and evolutionary andhistorical factors,a type can only be described withinabiogeographicalregion. Thestudywasdesignedasaqualitative surveybutorganized as much as possible alongquantitative lines. Macrofaunacomposition (taxon compositionandabundance)waschosenasabasicparameter. About 70 variables thatwereconsideredphysically,chemically orbiologically relevant,weremeasuredateachsampling site. In total 664 sites were sampled,distributed overabouttwentyphysico-geographicalwater types. These twenty types include all the major environmental variables relevant tothisregion;theyoverlap inabioticfeatures. Thephysico-geographicalwater typologywasusedasa practical tool for carrying out the survey, but was not used inobtainingthe ecological typology. Theecological typology was derived from the collectedbioticandabioticdataalone. A studyonthereproducibility ofthestandardmacrofauna sample was carriedout. Thestandardpondnetmacrofauna sampling techniqueused inTheNetherlands appearstopresentonlyasemi-quantitative picture of the more commontaxa. Withthistechnique onlyabout55%ofall thetaxapresentatasiteatthemomentofsampling were collected. However, when the standard pond net was used for a regional typological study itappeared thatseasonal differences as well as inconsistencies due to sample techniquewereoflittle significance comparedwithdifferencesbetweentypes. Itwas concluded that the

13

reproducibility of amacrofaunasample issufficient for typological purposes. Fordataanalysis,thetwentyphysico-geographical water types were combined intofivemaincategories. Theabiotic andbiotic datawere processedforthesefivemaincategories. Later,all the data were processed together to obtain an ecologicalwatertypology forthe province of Overijssel. Multivariate analysis techniques are appropriate in data analysis for typologicalpurposes. Different multivariate analysis techniques (cluster analysis and canonical ordination) were usedtoderiveanddescribe sitegroups intermsof taxon composition and mean environmental conditions. Sites intermediate between groups were manually relocated by using informationfromothersources (e.g. literature)abouttheecologyof the constituent taxa. The resulting site groups were termed cenotypes. Sixcenotypesweredistinguished amonghelocrene springs. The main differences between the cenotypeswererelated tohydrology (mainly durationofdrought)andacidity. Furthermore,the nutrient content and/or the load of organic material differed between related cenotypes. Eachcenotype contained its own microhabitat group(s). The microhabitat groups, except for two,wereassociatedwiththe durationofthedroughtperiod. Theother twogroupswere associated with the spring source and the springstream,respectively. The naturalreference situation forhelocrene springs probably resembles someoftheactualhelocrene springs investigated. Themost important humanactivities causingdisturbancearethosewhichcause changes in the chemical composition of the groundwater. The managementof springsshouldbedirectedatthisfactor. Elevencenotypeswere distinguished among the streams. The main differences between the cenotypes were related to dimensions, "stream-character",duration of drought and the load of organic material. All the streams were more orlessinfluencedbyhuman activities. Streamregulation, especially, has caused a dramatic change inthetaxoncomposition. Onlyabout2%ofthetotallengthof streams isstillmoreorlessnatural in 'stream-character' and its corresponding community. These 2%areonlypreservedbecause oftheir geographicalpositiononthesteepest slopes. Ingeneral,efforts at improvement of the ecological character shouldbedirectedatthe physicalandhydraulicconditions. Elevencenotypeswere distinguished among the ditches. The main differencesbetweenthecenotypeswererelated todimensions,duration ofdrought,acidityandcurrent. Furthermore, the nutrient content and/or load of organicmaterialdifferedbetweenrelatedcenotypes. Itisillustrated thatthetaxoncombinationfound inaditchreflects a stage of succession in space (profile structure) and time (maturity). Therefore, an overlap in taxon combination between cenotypes occurred (continuum). Ditches areartificial ecosystems whichmainlyoccur incultivated areas (which implies eutrophication) and depend on regularhumaninterference (cleaning,dredging). The ecologicalmanagement ofditches shouldbe based upon the relation

14

betweenprofile structure,successionstageandhumaninterference. Elevencenotypesweredistinguished amongtherivers,canalsandlarge lakes. They showedgreatoverlap intaxa (mostlyopportunists)andin environmental circumstances. Increasingdimensions oftheline-shaped more or less runningwatersgotogetherwithanincreasing drainage area. These large-sized water bodies function as collectors of nutrients, organic material andtoxicants. Thisresults inchronic stress and probably the disappearance of most taxa occurring originally. The chronic stress also overrules thenaturalmaster factorsof current and dimensions. Only a few taxa are still characteristic for the reach ofariverorthegradient insizeof canals. Ninecenotypesweredistinguished amongthe ponds and small lakes. Themaindifferencesbetweenthecenotypeswererelated todurationof drought,acidity,morphology andnutrient load. In particular, the four cenotypes within thegroupofstagnant,pH-neutralponds/lakes showedanoverlap intaxoncompostlon. These cenotypes represent a web-shaped continuum dominated by dimensions (relationofwidthto depth),nutrientload,andbottomcomposition (especially mesotrophic peat). The most important processess induced by men are acidification,eutrophication,andchanges intheoriginalhydrology. Cenotypes ofdifferentmain physico-geographical watertypes can be very similar. Therefore theyshouldbecombined orrearranged. This isdonebyprocessing allabiotic andbiotic data together. Again, cluster analysis and canonicalordinationwereapplied. Aftereach ordinationalongtwoaxes,thedistinctive cenotypeswereremoved and the remaining sites were reordinated. Through this progressive removalofgroupsofsites,finally,42cenotypeswere distinguished. Somenotesontheecologyofthetypifying taxaand themost important environmentalvariablesweremade foreachcenotype. The mutual relations between the cenotypes are shown in a hierarchical dendrogrambasedonbiological similarity aswellasina webofcenotypes (Figure 11.1). Thehierarchical dendrogram shows,amongothers,thata group of cenotypes related to middleandlowerreachesofregulatedstreams, smallrivers,ditchesandsome of the medium-sized, more or less stagnant waters,-allhypertrophic,mesosaprobic environments -,has afairnumberofthemacrofauna incommon. Apparantly,humanactivity (e.g. byregulationofstreams,dischargeofwastesandagricultural activity inthewatershed)leads toadecreasingrole of the factor current in running watersandthefactorsdimensions (infactshape anddepth)andbottom type inthestagnant waters and leads to an impoverishment ofthemacrofauna. Thewebof cenotypes illustrates the mutual position of the cenotypes, the transitionsbetweenthecenotypes (thecontinuum)and themajorenvironmental gradients. Thefourmostdominant factorsare 'stream-character',acidity,durationofdrought,anddimensions. Toevaluateagivenwaterbody,evensubjectively,itisnecessary to compare itwithotherwaterbodies indifferentstates. Thisscaleof evaluationneedsa reference water. The reference water is not

15

necessarilyanendpointofsuccessionnorapristine situation,butit shouldatleastrepresentasituationthatcanbeusedtoindicatethe desired direction of improvement. In choosing the reference communitiesforwatermanagementonemustfocusoncommunitiespresent in waters where major environmentalconditions arelessdisturbed. Suchwatersaretheundisturbed streams, the oligotrophic moorland pools,and themesotrophicoldmeanders cutofffromstreams. Theregionalecological typologypresented (Figure 11.1)offers a tool for establishing the developmentaldirectionfromacommunity observed inthefield towards thereferencecommunity. Itis also a tooltodetermine thereference community foraparticularwaterbody. Becausethefactors in the web of cenotypes are a result of a descriptive study, the web cannot be used without cautionasa predictive tool; thefactorsarepurely indicative. Thetypology and associated ecological concepts are a tool to helpsolvingwater managementproblems. Itisabasis,amongothers,formonitoring and assessment of waterquality, it indicates potential capacitiesof surface waters and it provides guidelines for management and restoration of surface waters. May itbe serve totheadvantageof theenvironmentaroundus.

16

INTRODUCTION

1.1

Background

Until recently man has more or less taken for granted the 'self-maintaining' character ofhisenvironment. Amajorreasonwill havebeenthathisenvironmental manipulations only affected local balances. Hence, in thefirsthalfofthiscenturybiologists only dealtwith localorpartialenvironmentalproblems. During the last decennia, however, ithasbecomeevidentthatnotonlylocalbutalso regional,nationalandevenglobalbalancesarebeing affected (Odum 1971). Environmental problems have become more and more obvious, especially in a densely populated and highly cultivated and industrialized country likeTheNetherlands. Around 1970,thegrowing awareness of the environmentalproblemsresulted inthefirststeps towards improvement. Inthatyear,theDutchparliamentpassed anact which aimedatareductionoftheorganicpollutionofsurfacewaters (MinisterievanVerkeer enWaterstaat 1975). Withinten years, many purificationplantswere improvedorestablishedalloverthecountry. Since then,effluentsandwastewaterare subject to water quality standards, based on oxygen and ammonium concentration and the biological oxygendemand. Theorganicallypollutedwaters definitely improved in that period. Ontheotherhand inthesameperiod,the stillundisturbed oranthropogenically onlyslightlydisturbed aquatic ecosystems became increasingly impoverished through (secondary) eutrophication,diffusesourcesofpollution,regulation of streams, acidatmospheric deposition,changes in(ground-)waterlevel,inputof water from thepollutedriverRhine,etc. (e.g. Gardeniers&Tolkamp 1985, van Dam 1987,Klapwijk 1988). Mostofthesehumaninfluences affectednotonly the variables related to the load of organic materialbutotherabioticvariables intheaquatic ecosystemaswell. Despitealltheeffortstoimprove thewaterquality intheperiod 1970-1980, theoverallwaterquality,andthereby theoverallquality ofaquatic ecosystems,remained low. Due to the increased demands made by human activities, differentwater types tended toresemble eachothermoreandmore. Hence,thereductionof organic pollution from point-sources didnotresult intheexpected improvement. This wasbecause theapproachchosenonlytackledasmallpart (theorganic wastes) of the problem and the solutionchosenonlydealtwitha technologicalaspect. Furthermore, an impoverishment of more or less undisturbed ecosystems may be assumed to have takenplacethrough thegrowing isolationand/or thedecrease ofareaoftheseecosystems. The area and degree of isolation of a refuge (likeanaturereserveoran undisturbed pool) together allow a certain equilibrium between immigration and extinction ofspecies. An increaseofthedistance between twohabitatsofaspeciesresults in a reduced immigration rate. A decrease of the habitat area results an increase of extinctionandthus in a reduced number of species. Increasing isolation anddecreasing areabothcause taxatodisappear (MacArthur

17

&Wilson1963,denBoer 1983). Waterqualitymanagement should not only be directed to the component H2O and dissolved substances but tothequality ofthe complete aquatic environment,whichonits turn is related to the conditions in the entire drainage or catchmentarea (Hynes1970, Lothspeich1980,Cumminsetal. 1984). This approach is gradually becoming accepted atanational levelaswellasatan international level (Persoone 1979, Higler 1988). Thus in 1980, the Dutch government adjusted itspolicyof1970andstressed the importanceof notonly takingdirectcareofhumanhealth but also of protecting material and ecological interests against the effects of water pollution. "Therelationbetweenhuman impact and water pollution will be seen from anecologicalpointofview. Theknowledge that surfacewatersareecosystems -thismeans systems in which living organismsplayanimportantpart -evenwhenthesewatersareman-made ordisturbed,makes itpossible todescribe thisrelation in a more logical and coherentway" (translated fromMinisterievanVerkeeren Waterstaat 1981). Tostop thethenstillcontinuing process of impoverishment of aquatic systems and to prevent all surface waters becoming ecologicallyuniformwiththesameaverage 'grey' quality, a water management policy differentiated according to the ecological properties of the water bodies, had to be put into practice. Therefore, in 1985, the national government furtheradjustedits policyandformulated, -fromanecological point of view -, aims which werebasedonthe interactionbetweenmajor functions forhuman societyandwater types;a differentiated water management policy. Thisdevelopment fromthelimitedscopeoftheorganicwastes approach towards one in which several functions for human society are considered together can only beachievedbymeansofanecological approach (Golterman 1976). Sofar,theseaimshavenotyet been put into operation. Thus, themainquestionnowishowtotranslate this ecologicalapproach intopracticalmanagement. In The Netherlands, the national government formulates the national water qualitypolicybutprovincial authoritieshave toput thispolicy intopractice. The provincial authorities, therefore, represent the appropriate level at which tointegrate ecological knowledge andfunctions for human society into a coherent water managementpolicy. In1981,thedepartmentofwatermanagement ofthe Dutchprovince ofOverijssel startedsuchanapproach and formulated the project 'Ecological characterization of surfacewaters inthe province ofOverijssel (EKOO)'.

1.2 Aims of the study TheaimsoftheEKOO-project are to develop a regional ecological characterization ofsurfacewatersbasedonmacrofaunacompositionand togatheragreaterunderstanding ofthevariety andthestructure of the macrofauna communities present inthewaters oftheprovinceof Overijssel. Inthisway theproject provides knowledge of aquatic ecosystems on a regional scale (with anemphasisontherelation betweenwater typesandtheimpactofhumanactivities) and a basis

18

for the development ofpolicies,particularly forwatermanagement, naturemanagementandnatureconservation. Anadditional aimofthisstudy is to explain and develop the typological approach inwatermanagementandtomakeacontributionto itsmethodologicalbasis.

1.3 Presentation

of the

results

Theaim (Section1.2) andthedesignofthestudy (Section 1.4) are defined first because these are important for deciding on an appropriate sampling technique (Chapter4 ) . InChapter2aconsistent terminology andsomedefinitions within aframe thatfollowsfrombroadtheoreticalconsiderations ontypology willbegiven. Oneofthemostimportantaspectsoftypologyor,more generally, of ecology isthecommunity concept. Thecommunity isan arbitrary entity. Therefore, conditions and criteria should be formulated which are not influenced by theobjectionsagainstthe artificial recognition of communities or, more generally, of ecological entities. InChapter 2theseconditionsandcriteriaare formulated tomake theentity 'type'functionalwithintheaimofthis study. Chapter3describes theareato be studied and introduces the materials andmethodsused. Oneofthemajoraspectsofthisstudy isthe reproducibility of the standardpondnetsample. Hence,Chapter4dealswithaspectsof presence,abundance,spaceandtimeofastandardpondnetsample. Thelargeamountofdatamade itlessefficient toprocessallthe data at the same time. Therefore,apriori,physico-geographical water typesweredistinguished (seeSection1.4.2)and thefield work was executed on each type separately. Closely related physico-geographical water types were, at a higher hierarchical typological level,combined intofivemainphysico-geographicalwater types. Dataprocessing tookplace forthese five main water types (Table 1.1) as described inChapters 5to9. Thesechaptershavea comparablestructure: -ageneral introductiontothemainphysico-geographical water type described, -adescriptionof the results of multivariate analysis and the validation oftheseresultsbasedupontheknowledge oftheecology ofthetypifyingtaxa, -adiscussionontheecologicalvalidity oftheresultingprovisional site groups and the presentation ofprovisional cenotypological relationsbetweenthesesitegroups. After themainphysico-geographical water typeshavebeenanalysed and translated into provisional ecologicalwater types,Chapter10 presents thefinalecological water typology for the province of Overijssel. The relation betweenthewater typespresentedwillbe discussedwithrespecttotheenvironmental factors that cause the distinctionbetweenthedifferenttypes. Chapter 11will indicate which environmental variables can be managed inorder toinfluence thesefactors.

19

Table1.1 Themainphysico-geographicalwater types intheprovince ofOverijssel,TheNetherlands.

Helocrene springsandspringstreams Smallstreams Upperandmiddlereachesofstreams Regulated streams Temporarywatercourses Rivers andcanals Lowerreachesofstreams Smallrivers Rivers Regulatedrivers Canals LakeIJssel Ditches :Ditchesonclay,sand,peatand infenland Poolsandsmalllakes:Moorlandpools Pools Dikeponds Sand-,peat-andclay-pits Ox-bowlakes Smalllakes Ponds Helocrene springs Streams

1.4 Design of the study 1.4.1 Considerations andbackground Warrenetal. (1979)introduced 'a conceptual framework on living systems theory' and statedfiveconceptualandmethodological rules togetherwith thisframework. These ruleswere takenintoaccount in thedesignofthisstudy. Rule 1. Only thestructure ofabiological community canbemeasured, its capacity and functioning canonlyberepresented indirectlyand incompletely. Capacity, inthesense of Warren et al. (1979), is a strictly theoretical concept. It canneverbemeasured directly;atbestit canbedescribed asaset of structures, each determined under a different set ofenvironmentalconditions. IntheviewofWarrenet al. the surrounding environment has two phases: an 'effective environment' composed of conditions acting immediately anddirectly uponthebiological community atanytimeduring itsexistence,and a 'generative environment' composedofsequencesofconditions leading tothose inthe 'effective environment'. As no natural ecosystem exists in a void, the biological community existsalso inathird phase. This is a surrounding 'extra-environment' composed of conditions that for theoretical reasons (e.g. therelation isnot

20

known) or methodological reasons (e.g. the variable is not measurable) have been excluded from the'generativeandeffective environment'butwhichmayultimatelybeimportanttothe persistence ofthebiological community. The'potentialcapacity'ofabiological community is the predetermination of all possible states and, therefore, of all possible structures which can evolve fromthe presentbiological community, under all possible changes in the surrounding environment. In other words, 'potential capacity' determinestheformsandamountsthatwillpermitacertain structure and thus determinesthepossible environmentswithinwhichacertain biological communitycanpersist. The'realizedcapacity'depends on the actualsurrounding environment throughtime. Thus,the'realized capacity'isdeterminedbytheformsand amounts actually provided, i.e. the actual environment determines the actual biological community thatpersists (Figure 1.1).

(

\

Realized capacities that could red at state2 t environmental nditions existed

structure ^ü functioning

Figure 1.1 The capacity of a biological community. Each biological community possesses a potential capacity to behave in certain ways. The interaction of system capacity and the state of the environment determine the system structure realized at any moment. If the environment at any time had been different, a different sequence of capacities actually realized would have been the result (adapted from Warren et al. 1979).

21

Macrofaunacomposition (taxoncompositionandabundance)waschosenas the biotic structure parameter. It meets thefirstrulegivenby Warrenetal. (1979),namely thatonly thestructure of the system can be measured. Macrofauna was chosen because the many,well describedandeasily identifiable speciesareusefulfor thedetection of changes intheexternalenvironmentonthescaledesired forthis study. This isaconsequence ofthedurationoftheirlifecycles (on average 1-12 months) and the diversity of habitats theyoccupy (Hellawell 1978),especiallybecause anumber of representatives is more or lesssessile. Furthermore,macrofauna isrelatively easyto collect inthefieldandtoprocess in the laboratory although the identification is time consuming. Hellawell (1978)concludedthat macrofaunaislikelytobe thefirstchoicewhenselectingagroupfor such a study. Macrofauna is defined asall invertebrate animals retainedbyapondnetwithameshsizeof0.5 mm. This mesh size boundary is used forpracticalreasons,sincemostofthesediments pass throughsuchanetandthemajorpartofthefauna is retained. Large representatives of meio- and microfaunagroups (e.g. large nematodes,ostracods anddaphnids)retainedby thenet were excluded fromtheanalysis. Rule2. Measurements of the structure of a biological community without relevant measurements of its surrounding, level-specific environment areoflittleexplanatoryvalue. Anecological characterizationrequires themeasurement of biological parameters as well as the measurement of relevant environmental conditions tosatisfythesecondrule. In this study, potentially relevant variables were chosen from a setofabout135physical, chemicalandbiologicalvariables identified. Persiteabout seventy variables were actually measured, dependentonlocal circumstances (e.g. water type,sitestructure,presence ofvegetation,etc.). Rule3. Explanationofthe functioning of a biological community should take into account thestructuresand thefunctioning ofits partson successively lower levels of organization (e.g. taxa, populations,individuals)andcannotbebasedonlyonknowledge ofits partsonthelowestleveloforganization. Theregionalecological characterization shouldbe a basis for the development of water quality standards taking into account the ecologicalwater typesand the impactofimportant human activities. It should also be a basis forthedevelopmentofawaterquality assessment system. Therefore,ideallythisstudyshouldaimat: a. thecharacterizationofmacrofaunacommunities representative for natural,undisturbed ecologicalwatertypes, b. thecharacterizationofmacrofauna communities representative of ecological water types which are under disturbance/stress, dependentonthecharacter and intensityofmajorhumanactivities. Thisdivisionstronglydependsonasubjective human evaluation and description of the natural conditions. It isartificial froman ecologicalpointofview (eachcombinationofenvironmental conditions implies a biological communitywhich isnomoreorlessnatural than anyother),but isnecessary forpolicypurposes. The management of

22

water requires a selection of those abiotic variableswhichare manageable and related to the different macrofauna communities. Therefore it is useful to separate natural and anthropogenically-inducedprocesses. Themacrofaunacommunitiesofdisturbedwater typesarerelatedto thecommunitiesofnatural,undisturbed ecological types. Information abouttheprocesseswhich influence the functioning of both these types of communities canbeextracted fromthisrelation. Withthis knowledge itis possible to purposefully influence the variables related to these processes. Therefore,information isneededabout theecologyofthecomposing taxa,asthethirdrule states. It is then possible to relatethetypifyingtaxaofnaturalanddisturbed ecosystems intermsofwaterqualitystandardsandtouse them in an assessment system. Rule4. Explanatory generalizationspertaining directly toany level of organizationofabiological community shouldcontainatleastone concept specific to that level and should subsume conceptual, methodological, or other sortsofindeterminacy thatmayexistwith respect tolowerlevelsoforganization. Thechoiceofthetypological approach,which isfurtherexplained in Chapter 2,satisfies thefourthrule. Typology isanapproachuseful atthecommunity levelandthepresentedmethodology,which is based on classification and ordination techniques, uses thehypervolume model (Hutchinson1957)asatheoretical concept. Hutchinson (1957) designates theentire setofconditionsunderwhichaspeciescanlive and reproduce as the fundamental niche (comparable to the above-mentioned potential capacity, but nowforaspecies)andthe actualsetofconditions (abioticandbiotic)under which a species exists astherealizedniche (comparable torealizedcapacity;Figure 1.2). Rule5. Perceptionand explanation of biological communities and environments are alwaysrelated toitsspaceandtimedimensionsand thecomponentsofthesystems. Rule five is important in stressing the relativity of collected results in space and timereferring tothegeographical limitations andindicatingthe importance,butalsothelimitations of,thechosen scale level. This is ageneralrulewhichprobably applies toall ecologicalstudies. Becauseofpracticalandfinancial reasons the EKOO-project is limited mainly (besidesthenaturalvariationpresent)toaspectsof organicpollution,eutrophicationandstreamregulation. Furthermore, the projectwasdesignedasasemi-quantitativestudy (Section1.4.2) andmostsampling siteswerevisitedonlyonce (Chapter4 ) . 1.4.2 Thepracticaldesign Ideally inastudylike this one, detailed information should be gathered on the macrofauna andtheimportantphysicalandchemical variables ofeachwaterbody. Butitwaspractically and financially impossible toinvestigateeachwaterbody: itwasnecessary totake

23

©

•

preference

+[

•^%-optimum ] O)

c

'E

o u c

D LL

\ JEnvironmental \ variable 1

• sublethal

•

lethal

*

tolerance

^j

©

optimum Environmental variable 1

^

Environmental variable 1

\J potential niche £23 realized niche

type A n

1I I I

type B 1

1

n

1—n—n

Environmental variable 1

Figure 1.2 The functioning of a species with respect to environmental variables and other species. The top figure (a) illustrates the functioning of one species along one environmental gradient. Two figures showing the functioning of two communities, A and B, versus their positions along two environmental gradients, variables 1 and 2; (b) a three-dimensional figure with the functioning axis, and (c) the corresponding contour plot with the contours indicating equal functioning levels. (d) A theoretical model of the fundamental niche (dotted and cross-hatched) of a species and its realized niche (cross-hatched), which is a subset of the fundamental niche, after competition and complete competitive displacement due to six superior competitors, represented by the circles with dotted parts (adapted from Pianka 1978).

24

representativesamples. Ingeneral,two kinds of faunal survey can be defined; (1) quantitative surveys, and (2) qualitative surveys (Elliott1971, Winterbourn1985). Aquantitative surveyaimstoestimatenumbers of organisms per unit area, whereas a qualitative survey aims at discoveringwhichspeciesarepresentand trying to estimate their relative abundance. Quantitative samplingrequiresrelatively large amountsofmanpower and finance (Elliott 1971) and meets problems because of thelimitations ofsamplingdevices (Harper&Hynes 1972) andbecause ofbiological featuresoftheorganisms themselves (Resh 1979). In this study we wanted to coveralargeareawithmany samplesandalargevarietyofsampling sites. Because of financial restrictions, this necessitated simple field and laboratory procedures. Thismadeaquantitative surveyunfeasible,and hence a more qualitative surveywasdesigned. Tomeetthedesiredaccuracy, thisqualitative survey was organized as much as possible along quantitative lines,andwasthussemi-quantitative (Hellawell 1978). Asnotall waters could be sampled, a reasonable number of representative samples over the area studied should be included (Elliott 1971). Inrandomsampling,everysampling sitehasan equal chance of selection. Butrandom selectionofsites isdifficultto achieveandcanbeextremely laborious (Elliott 1971). Therefore, sampling siteshavebeencarefully selected (selective sampling). To thisendthesurfacewatersoftheregionstudiedhavebeendivided a priori intoabouttwentyphysico-geographicalwatertypes (Table1.2), basedonthemajorenvironmentalvariables andon general ecological knowledge. To design the sampling program for each of these physico-geographical water types, theexactlocations,themajorof pollutionanddisturbance (physicaland chemical) influencing them, andthemajorhumanactivitieshad tobeestablishedbeforehand (often based on literature, information from water boards and field examination). Thereafter it was possible toselectsamplingsites (Figure 1.3)whereby representative sitesofmoreorlessnatural and disturbed (related to themainhumanactivities)circumstanceswere chosen. Thenthesamplingprogram was executed, the physical and chemical variables wereanalysed,theorganismswere identified,and thedatawereprepared forinterpretation (Chapter3). Thedivisioninphysico-geographicalwatertypesdoes not affect the final typological outcomeforthefollowingreasons. Thetwenty typeschosenincludeallthemajorenvironmental variables effective in this regionandshowmutually strongoverlap inabioticfeatures. Thelargenumberofsamplesensuresthattheeffectsof less obvious abiotic variables and biotic processes are also included. Furthermore,thescaleofthesamplingweb isfinerthan the pursued scale of ecological types. Theseargumentswillbeverifiedbyan elaborationofalldatacollected inChapter10. Therefore,sincethe chosen physico-geographicalwater typology issimplyapractical tool toexecute the survey, the resulting macrofauna community types provide theecologicaltypology.

25

Table 1.2 Characteristics ofphysico-geographicalwater types inthe province ofOverijssel,TheNetherlands.

A.Linear shapedwaters

Environmental streamvelocity streamdirection bottom profileshape profile origin dimension/ w.0.5-1.5m d.30m d.>1m

variable/ >20cm/s one-way sandwithgravel,silt irregular natural

SPRING(STREAM) SMALL STREAM UPPERCOURSE MIDDLECOURSE LOWER COURSE SMALLRIVER

0-20 cm/s one-way mixedsand, silt regular normalized regulated

0(5)cm/s none/bothways independent

independent independent independent

regular regulated or dug

(ir)regular independent

REGULATED STREAM

DITCH

TEMPORARY WATER COURSE

REGULATEDRIVER

CANAL

RIVER

-w.=width,d.=depth -Upper,middleandlowercoursearedividedbydimension (respectively w. 1.5-3,3-5,5-10mandd. 15-50, 15-50, 50-100cm) -ditchesaresubdivided inditchesonsand,clay,peatandinfenland -canalsaresubdivided incanalsandcanals infenland (w.>8m,d.>50cm) B.Round orirregular shapedwaters

Environmental hydraulic regime profile origin bottom/ sand peat clay fenland independent

variable/ isolated

isolated

natural

dug

isolated communicating natural dug

communicating diking

MOORLAND POOLSAND-PIT MOORLAND POOLMOORLAND POOL CLAY-PIT

POND

POND

OX-BOWLAKE DIKEPOND

PEAT-PIT (SMALL)LAKE POND LAKE IJSSEL

(small)lakesaredugaspeat-pitsbutwereenlargedbywindand/orsea LakeIJssel (IJsselmeer)originated throughdikingabranchofthesea andreclamationofsomepolders ifdimensionsaresmallespeciallypoolsandpondscanbe temporary dikepondsareformedafterburstingofadike

26

! «üt.o

3

z

/

X

+ — + D/nkel

^ . ^ 4 0 ^ -~

*•

o

-'\/^ "^ïïf^r^

•a•s/-S V -c

1 rJ( 0

I

!__

^ /\

a

1*

\* +

Y ^

-^ s -— i^ \ W \'—-~ "^

\

E

,'oA

L
ST < z

i /(

i

OJ

> X

/P?

>

\ >

x %;_ ^Aj^\./>f^A.

^V'

•'• \

V' V;: •••

\

\

1/ 0

••. \ R

"A/ '•••£ / ( 7 N . K-"'/ / j V

'.••-.•A'-''*

y

•-.._ N j i ^ - j

.

ÂIB

: \ : .x''.-x'-- / x x ..••*••-"'.'?/v'~•'*' ••':-'\ ssyi?

\ M i i 4 < N ^ """{fi

\\

/:W;

i

• -'^^^J

\

I"

'S'1

r'7. V ^j\ iy MB/'-' 9 r *

^ S ^ ^ - . "'••J>!^ ^»*S:-.

VV,;.fiwatc^rr^

j/4>-

"/v ••'c" •:•• ..a

p^v ' "'" \

v

/; \

r^f

;

/';': } M.-V'.

''•.'':>c •...;•a)

i > O

•/

\ 3

/ •*''• \)

^7 ••.X .\-\;i^f//

/ \

(

J

J jf\ '''••''•?:-\ //••' '*"" -•''.

s: a

( p ^ ^

\

•••''/ to

S

0 ic")

i3 l/l

S

^ V '

/ \ \f

Figure 3.7 The main discharge patterns of groundwater in the of Overijssel, The Netherlands (Bruinsma 1982). 55

province

Afsluitdijk,adikeacross the sea-inlet of the Zuiderzee, which formed lake IJssel. After thereclamationofthe 'Noordoostpolder' thegroundwater flow,whichuntil thenhadfedthepeat area in the northwestern part of the province, changed direction towardsthe polder. Theseepageareaof the northwest became an infiltration area. During thelastcentury, in general, streams and rivers have changed inthreechronologicalphases: la. Around 1900themajorbendswereremoved which resulted in an increase offall (andcurrentvelocity). Thisreduced thenumber of summer floods. Some weirs were constructed to prevent summer-drought 2a. Between1930-50moreweirswereconstructed, transverse profiles wereenlarged,morebendswerecutoutandbranchesofstreamsand riverswerediverted. 3a. Inthe1960stheactualchannel formswere made permanent; the bankswere strengthenedbystonesandgroynes,dikeswereenlarged whereby riparianvegetationwasremoved, weirs or sluices were improved, and with weirs andpumping-enginesmuchoftheriver systemwasextensively diverted. Thesemajor alterations ofstreamsandrivers induced the following changes inhabitats: lb. The increaseoffallleadstoanincreaseofcurrentvelocity and coarse substrates, resulting inmorediversehabitats. Preventionofsummer-droughtispositive for the macrofauna assemblages. 2b. Weirs leadtoareducedcurrentand thetotalcombination of physical alterations leadstoamoreconstant seasonalflow, bothresulting insiltation. Theintersticesbecome filledup and a more homogeneous substrate evolves, aprocessmore generally describedbyWard&Stanford (1979). The bank and bedbecomemorestable,andthecurrentandtemperature regime becomemoreuniform,generally resulting ina change from a small-river character tolarge-streamorrivercharacterwith the corresponding macrofauna assemblages. The macrofauna often comprises a more lentic,collector-dominated species assemblage, with bottom feeders and planktivorous taxa (Cummins1979). 3b. The processes described under 2. become definite and widespread. Removal of the riparian vegetation strongly altersthefunctionalrelations (asdescribed in general by Cumminsetal. (1984))present inthesystemandbed fixation markedly impoverishes thediversity ofhabitats. Apartfromphysicalchanges,the macrofauna have suffered severely from an increase ofpollutionstressduring thelastcentury. Both physicalandchemicalchangeshaveresulted in the extinction of a numberofrheophilic species (Mol1986b). Thenaturalstate isnolongerpresent inthestreams and rivers of the province ofOverijssel. However,studiesoftheactualbiota canimproveourknowledge ofrunning waters which is necessary to predict the consequences ofourpresentmanagement ofwaterquality andquantity. Besides all hydrological changes brought about by human activities, some other important figureswillbegiventoillustrate

56

man's impactonthequalityoftheenvironment inthisprovince. The totalpopulationdensity isabout270inhabitantsperkm 2 . About 50% lives inthemajormunicipalitieswithmorethan 30,000 inhabitants. Mostoftheprovince isused foragriculture (Table3.2).

Table3.2Thetotalareaandlanduse intheprovinceofOverijssel, TheNetherlands.

Area (km2)=3925 Landuse (%) urbanwoodsnaturereserverecreationwateragriculture others 8.4 8.8 3.3 1.3 2.9 74.8 0.7

Agriculturalactivity implies a continuous but diffuse source of contaminants for the surface waters. Italsopromotesatransport intotheenvironmentofespeciallynitrogen (959 kg/ha), phosphorous (326 kg/ha),potassium (713kg/ha)andpesticides. Anothernon-point sourceofcontaminants isprecipitation (Table3.3).

Table 3.3 Contaminantspresent inprecipitation (calculatedwithan averageof2.3mmprecipitation/day)inTheNetherlands.

Precipitation pH* (mg/m2/day)4.4

CI

S04

6.7

12.9

F 0.11

N03 6.9

NH4 3.7

P04

Zn

0.01

0.07

* naturalunpollutedpH= 5.6

Themainpointsourcesofcontaminationdischargemostly waste water on surface waters to atotalof715,000 inhabitantequivalences (1 inhabitantequivalence= 125 litre water with BOD (biodegradable organicmatter)54g,Kjehldahl-N10g,andtotal-P4g).

3.2 Macrofauna

sampling

and sample

processing

Intotal664sampling siteswere sampled. Mostsites (609)were only visited once and some sites (55)werevisited twice. The sampling dateswere spreadoverthefourseasonsaswellasover several years (1981-1986) (forfurther informationseeSections5.2.1,6.2.1,7.2.1,

57

8.2.1 and 9.2.1). Beforesampleswere taken,25-50mstretches oflinearwaters and along the banksoflargelakes,orthetotalcircumferencesofsmall lakesandpondswere searched forthepartitions ofvarious habitats. This search resulted in a schematicpicture ofthemajorhabitats present,e.g. standsofmacrophytes,leafpackets,bare sandybottoms etc. Ateachsampling siteitwasattempted tocompose thesampleby combining subsampleswhichweretakenin proportion to the various habitats present as estimated from theschematicpicture. So,the totalsample represented the observed environment. This careful presamplingprocedure isnecessary forobtainingareliablesample. Abottom subsamplewasobtainedbyplacing thepondnet (meshsize 500 jim, frame height 200mm,andframewidth300mm)onthebottom and,facing upstream in running waters, sampling the substratum (sometimes including some of the standing lower parts of the vegetation)directinfront. Thepond net was pushed, with short quick movements, throughtheuppercentimetres ofthesubstratumand thensweptback immediatelyabovethesampledarea. Subsamples from bank, emergent,floatingand/orsubmergedvegetationwereobtainedby sweeping the pond net several times through that part of the vegetation. Ateachsampling siteallthemajorhabitatswere sampled inthisway. A trajectofabout0.5-1 mwassampled in every major habitat,sothetotalcombined samplecomprisedatleastabout5m (or 1.2m2ofvegetationand0.3m^ofbottom subsample). Insmall water bodies, the sample couldnotbecomposedofbothavegetationanda bottom subsample,soonlyone combined sample was taken. In the deeper parts (depths of more than 1 to 2 m ) oflargewatersan Ekman-Birgesamplerwasused,withwhichfivegrabsampleswere taken at eachsite. Indeepwaters thegrabssubstitute one0.5mpondnet bottomsubsample. Inhelocrene springs thepondnetcouldnotbeusedand here the micro-macrofaunashovelwasused (Tolkamp 1980). This shovel is10cm wideand15cmlongandsubsamplesweretakentoasedimentdepthof3 cm,whichpermits samplingofsmall-scalemosaic substratepatterns. All sampleswerewashed inabucket,takentothe laboratory and stored in a refrigerator (at6°C)while theywereaerated. Inthe nextfewdays thesamplewascarefully processed in the laboratory whilemostoftheanimalswerestillalive,bysievingthesampleover threesieves (4.0,1.0,and0.2 mmmeshsizes)andplacing the sample in white, flat-bottomed traysfromwhich theanimalswere sortedby eye. Ifataxonwaspresent inlargenumbers,a representative part (large and small individuals)wasremovedandtheremainingpartwas estimated. Thecollected individualswereconserved inethanol (70%) except for oligochaetes andwatermiteswhichwereconserved informalin (4%) andKoenike-fluid,respectively. Organismswere identifiedbyusing the keys listed by Tolkamp (1984). Mostmacrofaunaorganisms,exceptforsomedipteranfamilies, couldbe identified togenusorspecies level. Closely-related taxa which could not be distinguished consistently arepresentedasa 'group'. Because identificationwascarriedoutbyseveralpeople, a substantial control procedurewaspartoftheidentificationprocess toensureaccuracyandconsistency in identification. In cases of doubt, specialistswereconsulted. Areference collectionofallthe

58

taxaiskeptattheResearchInstituteforNatureManagement. An eight-letter code was developed for all macrofauna taxa occurring in The Netherlands (Verdonschot&Torenbeek 1988),which facilitated thetransferof the macrofauna data to the computer. Physicalandchemicalvariableswerealsocodedforthispurpose.

3.3 Abiotic

sampling

and sample

processing

Adatasheetwasused tonoteanumberofabiotic (and some biotic) variables directly in the field. Some of those variableswere measured directly (width, depth, surface area, temperature, transparancy etc.) or classified (substrate, colour etc.). Field recorders were used to measure oxygen, electrical conductivity, currentvelocity andpH. Anumberofchemicalvariableswere analysed atthemomentofsampling fromallwater surface samples. The water sampleswerefixedimmediately inthefieldbyaddingHgCl2 (Mackereth etal. 1978)andwere laterfrozen (Harmsenetal. 1981). Analyses were conducted bygenerally following theprescriptions oftheDutch NormalizationInstitute.

3.4 Data

processing

3.4.1 Preprocessing ofthedata Theconclusionspresented inChapter4ledtothefollowing techniques beingused inthisstudy. A flowschemeofthedataanalysis isgiven inFigure3.8. Theperformance ofmostmultivariate analysis techniques is best whentaxahaveresponse surfacesofasimpleform (linearorunimodal) withrespect toenvironmental gradients. Thiscondition is best met at species level. Duetotaxonomicalshortcomings individualscould sometimesonlybe identified to generic or higher levels. Higher taxonomical units oftenhaveresponsecurveswithawideramplitude. Whetherornotataxonisincluded intheanalysis isbasedoncareful individual weighing. Thisweighing isbasedonecologicalknowledge andthefrequency ofoccurrence ofthetaxon. Furthermore, a taxon cannot belong tomore thanonetaxonomicalunit. Becauseofitsaims thisstudywasrestricted tothemacro-invertebrates (e.g. fishesand amphibianswereexcluded) (Verdonschot 1983). Data transformation isanimportantstep inecological ordination (Noy-Meir 1973, Noy-Meiretal. 1975,vanderMaarel 1979). Dueto the semi-quantitative character of the sampling technique, the mathematical background of themultivariate techniques andthefact thatthe difference between 1 and 2 individuals found is more significant than the difference between 101 and 102, the taxa abundancesneedtobe transformedwhenusingquantitative data. The higher ranges of abundance lead to an overweighting ofabundant speciesduetotheeffectofdominance (Section 2.2.1). Therefore,

59

Sites

Sites

Macrofauna taxa

Environmental variables PREPROCESSING OF DATA Taxonomical reduction Elimination of i n d i f f e r e n t variables Transformation of abundance Logarithmic transformation of into logarithmic classes q u a n t i t a t i v e variables Standardization M U L T I V A R I A T E ANALYSIS Classification sites (FLEXCLUS) taxa (NODES)

Correspondence analysis (CANOCO)

Principal component analysis (ORDIFLEX)

Review of literature (independent data) INTEGRATION

Figure 3.8 Scheme of data

processing.

thedatawere transformed into the following logarithmic abundance classes,asindicatedby Preston (1962): NumberofindividualsLogclass 1 1 3 2 7 3 15 4 31 5 63 6 127 7 255 8 511 9 1023 10 2047 11 4095 12 etc. Toavoidoveremphasis oftheroleof single samples with a highly divergent taxon composition (so-called 'outliers'), theoptionof 'downweightingofrarespecies'(Hill1979)isinvoked ineachrun of

60

themultivariate techniquesused. Quantitative environmentalvariables,exceptpH,were transformed into logarithmsbecauseofskeweddistributions. Theothervariables suchas 'substrate type', 'permanence', 'shadow' and 'season', are nominal andweredealtwithbydefining socalleddummyvariables. A dummyvariable takesthevalueofonewhenit concerns the observed stateofthevariable andthevaluezerootherwise. 3.4.2 Multivariate analysis techniques

Cluster

analysis

Siteswereclusteredbymeansoftheprogram FLEXCLUS (van Tongeren 1986). The clustering strategy of FLEXCLUS canbe summarizedas follows. The strategy is based on an initial, non-hierarchical clustering, following the algorithm ofS^rensenforasite-by-site matrixbasedonthesimilarity ratio. In this initial clustering, sites are fused according tosinglelinkagebutafusionisskipped whentwositeswithalowerresemblance toeachother thanaspecified threshold wouldbecomemembers ofthesamecluster. Thevalueofthe thresholddependedonthenumberofsitesclustered and the cluster homogeneity. The homogeneityofacluster isdefinedastheaverage resemblance (basedonthesimilarity ratio) of the sites of this cluster to its centroid. The initialclustering isoptimalisedby relocative centroid sorting. Largeand/orheterogeneous clusters are divided, small and/or comparableclusters (withahigh resemblance) arefused,andthensitesarerelocated. Inrelocationeach site is compared to each cluster (asitwasbeforerelocationofanysite) and,ifnecessary,moved tothecluster towhich its resemblance is highest. Beforeasite iscomparedtoitsowncluster,therespective siteisremoved fromthatclusterand the new cluster centroid is computed.

Ordination Thedataareordinatedbydetrended correspondence analysis (DCA) and detrended canonicalcorrespondence analysis (DCCA),using theprogram CANOCO (terBraak1986,1987). Both DCA and DCCA are ordination (reciprocal averaging) techniques. DCCA isanextensionofDCA,in whichtheordinationaxesarechosenas linear combinations of the environmental variables. DCCA is anintegrationofregressionand ordination (Jongmanetal. 1987, ter Braak 1987) and shows the response of taxa or groups of taxa to environmentalvariables. Detrendingbyfourth-orderpolynomials is used. DCCA leads to an ordination diagram in whichtaxa,sitesandenvironmentalvariables (arrows)canberepresented. Anexample isgiveninFigure 3.9. The arrow of an environmental variable points approximately in the directionofsteepest increaseofthatvariableacross the ordination diagram and the rate of change inthatdirection isequaltothe lengthofthearrow. This means that the value of an important environmental variable inasite (orsitegroup)isvisualizedbyits perpendicular projectionontheenvironmental arrowor its imaginary extension (inbothdirections). Note thatthisrelation isoptimized forspecies,notforsites. Withinthediagram the sites (or site

61

groups) and environmental arrows should be seen as a relative projectionuponeachother. Note thatsiteswithataxon composition of only common taxalieinthecentreofthediagram. Thesecommon taxahave theiroptima inthecentreorare independentof the axes. Sites with a poorandaberrant taxoncompositionoftenlienearthe peripheryofthediagram. Whenthetaxon composition is due to a specific environmental situation itisconsidered typifying otherwise itisduetochance. Althoughsiteswereclassified intositegroups, the distributionpatternofthe individual taxaisnotdiscontinuous. Therefore,intheordinationdiagram eachsitegroupisrepresentedby

rsi

'5

» environmental v a r i a b l e

Çiiïl? 90% ellipse 1

| samplp u

HABITAT GROUP

T Y P I F.

1 1 2 2 2 2 2 2 2 2 2 2 2,4 2 2 2 3 3 3 3 3 3 3 3 4 4 4,6 4,5 4 42 6 6 6 6 6 6 6

I N D I F.

MICROHABITAT GROUP T I Y N P D I I F. F. D D G F

R E M A R K

P t t

F 3

C F F A,C F F F

G F

P a,t t

E,F

P s

E E E,C E C 2 ,3 5

C E,G A P A

B

5 5 6

t t

2

G a t E E,F

3 1 3 1 1 ,3

s,a

E D D

E E C,F

Ptychoptera sp. VI Dicranota sp. III Pisidiidae Ceratopogonidae Pedicia sp. II II,IV Brillia modesta Limnophila sp. VII Macropelopia sp. III Micropsectra sp. II III Plectrocnemia consp ïrsa III Beraea maurus II,IV II Krenopelopia sp. Enchytraeidae VII Nemurella picteti III Polypedilum breviantennatum III VI Radix peregra Thaumasoptera sp. IV Asellus aquaticus V Diptera Scirtes sp. 1. VI VI Chironomus gr. annularius Psectrotanypus varius VI Adelphomyia sp. VII Dixa maculata III,IV Dicranomyia sp. IV Ochtebius sp. 1. III VI Ormosia sp. Prodiamesa olivacea III Psychoda sp. III Muscidae V Tanytarsus sp. Sialis fuliginosa II VI Glossiphonia complanata VI Proasellus meridianus Legend: GROUP I = groudwater II = spring III = lotie IV = hygropetric V = lentic VI = ubiquitous VII = terrestrial

i io r is is i i is s bc,wv s um i is s,u is r io us s,u be s s us i is s,u be s,m us us i s,m us i us i us,bc i io i io i io s,u be,um s,m us i be,us i us i io s,m us i us s,u is us i i us

HABITAT s = spring u = upper-reach m = middle-re ach i = indifferent b = bank r = running watei

1 1,3 2,3 2,3 2,3 3 3 3,4 3 3,5 3 3 5 5 1 1 1 2,3 2,3 2 2 2 3 3 3 3 3 3 3 4,5 4,6 5

C E E,F E F,G,H

81

E,H

E E C E E I

C B

E,F

E,B

C F G

E E

E

C D C

MICROHABITAT be - between coarse material us = upon se iiment um = upon mineral sediment is = in sediment im = in mine rai se diment io = in organic se diment wv = between dense vegetation and mosses REMARK s = semi-aquatic a = acidophilous t = temporary p = polysaprobic

(MICRO)HABITAT GROUP (figures and characters are explained in the text) TYPIF. = typifying taxon INDIF. = indifferent taxon

E,F

E

Table 5.2 Helocrene springs; number of sites,means, and standard deviations of the quantitative environmental variables and percentages of the nominal environmental variables per site group. Abbreviations and dimensions or units are explained in Table 10.5.

Quantitative environmental variables Habitat group 1 Number of sites 13

T sd PH sd EC sd 02 sd 02% sd

4.7 1.7 7.0 0.2 261 91 8.2 3.7 67 28 1.3 0.6 .24 .16 .04 .04 10. 8.0 .21 .10 .22 .09 36 19 10 17 .06 .10 30 18 13 6 7 1 43 18 18. 12. 1.0 .43 7.4

kj-N

sd NH4 sd N02 sd NO3

sd 0-P sd T-P sd CL sd S04

sd FE sd NA sd K sd MG sd CA sd BOD sd COD sd TOC

82

2

3

4

5

6

6

14

3

4

3

2.9 3.0 6.0 0.8 207 49 7.0 2.8 51 28 2.0 0.5 .24 .15 .03 .03 9.0 8.4 .11 .12 .13 .10 26 7 30 24 .75 .91 14 6 12 8 9 4 42 21 23. 16. 1.3 .10 9.0

4.5 1.3 6.3 0.4 240 60 7.2 2.5 56 20 0.8 0.4 .09 .10 .02 .02 14. 8.7 .11 .07 .14 .06 28 10 28 19 .04 .07 22 19 14 8 9 3 51 18 9.3 7.0 0.9 .29 3.6

4.3 1.2 4.4 0.3 327 101 9.0 9.0 69 25 0.7 .04 .08 .09 .01 .00 26. 15. .01 .01 .04 .00 22 10 31 4 .40 .35 10 5 11 6 7 3 56 36 8.0 0.0 0.7 .17 1.7

3.5 0.6 4.9 0.3 175 6 8.7 8.7 65 31 0.7 0.4 .19 .01 .00 .00 11. 5.3 .05 .01 .06 .02 25 1 36 12 .30 .35 11 1 10 2 7 2 41 7 7.0 0.0 0.6 .00 1.0

3.7 0.6 4.2 0.2 120 0 18. 18. 135 32 0.8 .02 .13 .05 .02 .01 6.5 2.1 .04 .02 .07 .02 14 3 33 6 .00 .00 5 1 4 0 3 0.1 12 2 10. 1.7 0.5 .00 1.3

8.6 .47 .49 .07 .08 .15 .18 73 32 .07 .17 11 17 0 0 13 26 13 26 6 10 0 0 0 0 9 15 26 16 11 14 48 22

sd W sd D sd S sd FALL sd SOURDIST sd S-T sd %-S sd %-B sd %-T sd %MM-B sd %MM-E sd %MM-S sd %MMSILT sd %MMSAND sd %MMGRAVE sd %MMDETR sd

9.6 .39 .63 .06 .08 .02 .02 70 46 .03 .08 18 15 0 0 41 37 41 37 23 21 0 0 0 0 23 23 17 24 0 0 37 36

4.3 .74 1.1 .14 .18 .11 .11 74 38 .00 .00 19 21 3 11 24 30 27 33 10 20 4 12 1 5 15 22 21 23 4 12 44 22

.58 .34 .41 .09 .10 .14 .15 83 29 .00 .00 13 23 0 0 18 28 18 28 23 25 0 0 0 0 7 12 3 6 0 0 67 15

.00 .21 .23 .06 .07 .10 .11 63 25 .00 .00 25 30 0 0 34 36 34 36 8 15 0 0 0 0 0 0 15 19 0 0 78 17

.29 .38 .12 .10 .09 .07 .05 83 29 .00 .00 12 16 0 0 27 46 27 46 33 58 0 0 0 0 0 0 7 12 0 0 60 53

Nominalenvi ronmentalvariables Habitatgroup1 SPRINGSOURCE SPRINGSTREAM TEMPOR STCMDLPE STSAND STCDLE STFDPE STSILT SHADOW SURFOR SURGRA SURWOB SURHEA

83

39 62 0 31 69 69 15 8 92 62 15 23 0

2

3

83 17 83 33 33 33 0 50 67 67 50 0 0

43 50 29 36 50 57 7 21 86 86 14 0 0

4

5

6

33 50 67 50 0 33 33 50 33 25 33 50 0 67 0 25 67 100 67 100 0 0 0 0 33 0

67 33 33 67 33 33 33 0 67 67 0 0 0

Nominalenvironmentalvariables Microhabitat group Numberof sites leaves leaves/sand leaves/vegetation pool baresand sand/detritus sand/coarse detritus vegetation/mud vegetation leaves/irondeposit springsource spring stream

A

B

C

D

E

F

G

H

I

6

12

10

24

23

13

7

2

8

0 0 0 0 0

0 13 13 0 13

33 58 50 13 48 23 14 8 20 25 13 17 0 0 0 17 10 4 0 0 0 0 0 10 0 0 0 0 0 10 4 0 4 0 0 50 17 0 17 9 31 0 0 25 0 0 0 33 4 0 0 0 0 0 0 4 17 46 0 0 0 0 0 0 4 0 14 50 13 0 0 0 0 0 0 71 50 25 50 42 10 46 52 85 57 50 63 50 58 90 54 48 15 43 50 38

84

0 0

t/1 3

(U

4-»

4-»

+-» L. C

ra

'C 4-* ü 5 T3 O

*-*c

n3

j

\

>

1

\

'

1l

/

LU

' 4

+

\

i

a

°\

z. .

i»

— - ^

M

O

10

l

•

ui 3

'Z. 4-*

c o

01 •o tv

ra

ai i 4-» 3-

CU

c

•

•

-r

tu

r

>

czvp

• ^ — « - J - S

4C

•

i

+

«"8-8

m» 0) JL

5

S| ^

4-»

. *

CL

£

E

g.2

•2 CQ

dl

u> ^ + -

tu •o

•

•

tu

3

1 OL 4-"

X

i/>.c

>•? £2 \ s-

. u

•

•

•

i_ ^ \ ra o \ a \ E \ D •x-J +-*

y1 «£

C 4) CL O

>-

\

H

i o\° • .

'

s ' / y

y

/I

o\o

Figure 5.2 Helocrene springs; ordination (PCA) diagram for axes 1 (horizontally) and 2 (vertically) with only environmental variables indicated. Dotted lines indicate the environmental parameter complexes. For further explanation see Sections 3.4.2 and 5.2.3, and Figure 3.10. Abbreviations are explained in Table 10.5. 85

Figure 5.3 Helocrene springs; ordination (DCCA) diagram for axes 1 (horizontally) and 2 (vertically) with the habitat groups (contour lines) and the most important environmental variables (with an interset correlation greater than 0.3 (arrows)) indicated. For further explanation see Sections 3.4.2 and 5.2.3, and Figure 3.9. Abbreviations are explained in Table 10.5. 86

Figure 5.4 Helocrene springs; ordination (DCCA) diagram for axes 1 (horizontally) and 2 (vertically) with the microhabitat groups (contour lines) and the most important environmental variables (with an interset correlation greater than 0.3 (arrows)) indicated. For further explanation see Sections 3.4.2 and 5.2.3, and Figure 3.9. Abbreviations are explained in Table 10.5. 87

indicated. Forsixteensites,therespectivehabitatsample from the spring source and the adjacent springstreamaggregatewithinone group. This indicates thedominance ofcommon environmental factors in these seemingly different habitats, which is of coursealso reflected inthemicrohabitat groups. Therefore,the habitat groups are seen as site groups. Aberrant groupsconsisting ofonlyone sample inbothordinationdiagramswillnotbediscussed. Becauseofthedominanceofcommonenvironmental factors on both site (habitat) and microhabitat groups,thesegroupsarediscussed together. Onforehand,somesimilaritiesbetweenthedistribution of the site groups in Figure 5.3 andthedistributionofmicrohabitat groups inFigure 5.4 canbeseen. Forexample,thedominantvariables related tositegroup 2 (Kj-N,temporary,Fe2+,etc.)correspondwith thevariables related tomicrohabitat groupsF,GandH (leaves, iron deposits, vegetation and mud). The relation between site and microhabitatwillbedealtwithfurther inthenextsection. Numbers between parentheses in the textindicate therelative frequencyof occurrenceofthatenvironmentalvariable intherelevantgroup. Sitegroup1 Within thisdatasetgroup 1is associated with a relatively high degree of eutrophication (high conductivity, high phosphate and ammonium concentrations,and high pH) (Figure 5.3). The sampled habitats consist mainly of coarse detritusandleavesonsandand gravel. Theyarepredominantly (62%)situated inspring streams and vegetation cover is sparse. Mostofthesitesaresituated onthe northeasternhill-ridge. Typifyingand indifferent speciesaremostly rheophilous (upper reaches and spring streams),andtypicalspring species are scarce. These rheophilous species also typify microhabitat group D, a grouprelated tosand,gravelanddetritus (Figure5.4). Sitegroup2 Siteswithingroup 2containlarge amounts of organic material in solution and the ironconcentration ishigh. Fiveofthesixsites aretemporary springsourcesandfourofthemarevegetated. Themain species are ubiquitous freshwater species, indifferent tohabitat type,withapolysaprobic character,and most of them are drought resistant. Typifying species inhabitingsiltappear inmicrohabitat groupF,thepolysaprobic andsomesemi-aquatic species appear in G and H. Microhabitat groupFandpartsofgroupHareassociatedwith vegetationandmud in spring sources. Microhabitat groups H and especially Gareassociatedwithaccumulationsofleavescoveredwith irondeposits (71%). Sitegroup3 Site group 3 consists of sites mostly (72%) situated on the southeastern hill-ridge. Potassiumandmagnesium concentrationsare slightlyelevated. Thisistheonlygroupwithanumberof typifying and indifferent species from real springs, spring streams and hygropetric environments. The microhabitats are divided into two

groups. Microhabitat group Econtains therealspring species,and speciesofwaterfilmsandspringstreams. Thisgroup takesacentral position inFigure 5.4,indicating theabsenceofanyassociationwith theextremes ofany specific parameter, although the presence of leaves has some importance (48%). Microhabitat groupCincludes ubiquitous freshwater speciesandmorecommonrunning water species. This group represents samples composedofleaveswhetherornotin combinationwithvegetationorsand (80%)andtakeninspring streams (90%). Sitegroup 4 Atsiteswithingroup4sulphateconcentrations arehighand ammonium and phosphate concentrations are low. Typifying and indifferent speciesareubiquitous freshwater species. Someofthe microhabitats belong to group C, the others to microhabitat groupA,whichis composedofindifferent lotie species resistent to drought. This group isassociatedwiththevariables sandanddetritus. Sitegroup5 Allsitesof group 5 are situated within one watershed on the northeastern hill-ridge (called Springendal). Althoughmostofthe sampleswere takenfromdetrituspackages,concentrations of organic material in solution were low. The water isrelativelypoor in nutrients andweakly acid. There is only one typifying species, Nemurella picteti. The presence of many speciesfromspringsand springstreams isinteresting,althoughalloccur inlow frequencies. Microhabitat group B, associatedwithleaves (81%),consists ofthe species combination of Nemurella picteti, Macropelopia sp. and Plectrocnemia conspersa. This microhabitat group together with microhabitat groupI composes group 5. Microhabitat group I is typifiedby semi-aquatic speciesoccurring inseveralsubstrates. Sitegroup6 Thisgroupoftemporary, acid and oligotrophic sites is situated entirely on thewesternhill-ridge. Thechemicalcompositionofthe wateratthese sites indicates a direct input of rainwater. This group istypifiedbyseveralColeopteraandDiptera. Nomicrohabitats weresampled atsiteswithinthisgroup.

5.3 Discussion

of data

lilies (1952)reported thatspecieslivingin the lower reaches of mountain streamsalsooccurred intheupperreachesofstreamsofthe westernEuropeanplain. Withintheregionwestudied, the ubiquitous freshwater species of mountain streams indeedhavebecome typical spring species. According tolilies (1961)and lilies & Botosaneanu (1963), this transition dependsonthelatitudeandaltitude ofthe sourceand isduemainly totemperature and zoogeographical factors. Thisprinciple isgenerally accepted (Hawkes 1975), although therehas

89

beencriticism ofsomeaspects (Thorup 1966). Comparedwith thefaunistic compositionof springs in a nearby region at a higheraltitudeandlowerlatitude,i.e. theregionof Baumberge inGermany (Beyer 1932), thefauna in our region consists definitely oflesstypicalspringinhabitants. However,thefaunaof Baumbergeconsists inturnoflesstypical spring inhabitants compared to alpineand subalpine springregions (lilies 1952). Theresultsof thiscomparisonareconsistent with the above-mentioned principle. However, as Thorup (1966) argued pollution effectsshouldnotbe disregarded. All of the springs under study are influenced by eutrophication tosomedegree. Maas (1959)concludedthat the southeastern hill-ridge in the study areawasricher innutrients (macro-ions)thanthenortheastern ridge. Thisdifference,whichwasdueto the nature of the local soil,seemedtohavedisappeared in1982 (Table5.3). Especiallythe

Table5.3 Comparisonofsomeaveragechemicalparametersmeasured in helocrene springsonthesoutheasternhill-ridgeandthe northeasternhill-ridge in1959andin1982.Aberrantnumberofsites aregivenbetweenbrackets.

Southeasternhill-ridge 1959 numberofsites PH EC /uS Ca mg/1 Mg mg/1 NO3 mg/1 SO4 mg/1 NO2 mg/1

4 6.9 278 22.7 8.3 1.7 39.6(2) 0

1982 9 6.5 260 47.2 9.2 19.3 41.9(5) 0.04

Northeasternhill-ridge 1959 3 5.8 95 9.5 2.3 .8 17.1(3) 0

1982 11 6.2 218 42.8 7.1 10.7 34.8(7) 0.02

calciumandnitrateconcentrations inthespringsofthe southeastern hill-ridge have risen during the last twentyyears. Inthesame periodthevalues ofallthemeasuredvariables on the northeastern ridge rose strongly, which abolished thelocaldifferencesbetween them. Duetoastrongincrease intheintensityof the agricultural use of large partsoftheridge,the inputofnutrients (especially nitrate)hasapparentlymasked thenaturaldifferences ofthesoils. Detailedknowledgeontheautecologyof typifying taxa supports the ecological validity of thegroups. Many literature references gaveonlyroughindications. Stilldifferencesbetweensomegroupsof typifying taxaappeared,according totheliterature andaccordingto ourobservations,tobeclearlyassociated with differences in the majorenvironmental factors. Inbothsiteandmicrohabitat groupsthe master factorsarethedurationofdroughtand the pH. Within the

90

occurring combinations of these factors the nutrient loadhasa diversifyingfunction. Figure 5.5 showsthe typological relations within the springs under

drought ^acidity cenotype: corresponding with | habitatgroup(number)I and microhabitatgroupj (letter) i impactofenvironmental! variable L

,permanence

Figure 5.5 Helocrene springs; relations between site (figures) and microhabitat (characters) groups (circles) and cenotypological relations. The arrows indicate the direction of increase of the respective variables. study. Sincemicrohabitatgroupsmostlyoccurwithinone site group and since spring sources and springstreams (ouroriginalhabitat samples) often group together, only the site groups, will be considered ascenotypes. Acenotypedevelopsorbecomes impoverished under the impact of environmental changes. Series of cenotypes, therefore, can occur inparticulardevelopmental stages (succession stages). Itisconcluded thatsitegroups 1,3and5arerelated cenotypes of helocrene springs. The group sequence (5, 3,1)reflectsan increase ineutrophicationand/or (secondary)organic enrichment due to humanactivityaswellasduetonaturalprocesses (e.g. inputof leaves). But therearealsodifferences ingroundwater supplyas can be seen fromthelowerpH ingroupH5. Sitegroup2isdescribedas thecenotypeoftemporary,neutralseepagemarshes. Sitegroups4and 6 are related cenotypes of temporary, acidseepagemarshes. The

91

droughtperiod islonger in6thanin4. Mostmicrohabitat groupsoccurwithinonlyonecenotype. For the helocrene springsstudied,eachcenotype isaspecificcombinationof microhabitats. Furthermore,themicrohabitat groupsare related not only to 'substrate' (inabroad sense likeleaves,mud,vegetation) differencesbutalsotootherdominantenvironmental variables (e.g. constancy inwater flow,permanence,acidity). Thiscanbeexplained bythefactthat 'substrate'isdeterminedonitsturnby thedominant surrounding environmental variables, its character results from conditionspertaining toclimatology,geology,history and hydrology. The last factor is themoststronglydifferentiated onewithinthe regionstudied. Only microhabitat groups C and E can also be indicatedascenotypesofhelocrenespringstreamandhelocrene spring source,respectively. Othermicrohabitat groups are more strongly associated with permanence, probably reflectinggradientswithina marshorspring fromseepagesource todrybank. Thesegroupscan be seenaschoriotypes (Section 2.3.2). lilies&Botosaneanu (1963)divided a spring (crenon) into an eucrenon (spring source) and a hypocrenon (spring stream). This divisionwasnotsupportedbyThorup (1966). Our results show that such adivisiondoesnotreflectafundamental differencebutdepends stronglyonlocalenvironmentalmaster factors. Within the region studiedhydrology isthislocalenvironmentalmasterfactor.

92

6. MACROFAUNALCOMMUNITYTYPESINSTREAMS

6.1

Introduction

Duetothegentle slope (fallaveragesbetweenalmost zero and five pro mille) of theterrainthestreams inthestudyareaarelowland streams likemostrunningwaters in The Netherlands (Tolkamp 1980, Higler&Mol 1985). Moststreamsarefedbyrainwater,sotheylacka welldefined source. Theirdischargeshowsasmoothed relation with the amount and frequency of precipitation inthevariousseasons. Currentvelocityvaries from5to30cm/s insummerandearly autumn, andfrom 30to60cm/s inlateautumn,winterandspring. Incidental, currentvelocitiescanreachup to100cm/sormore,forexampleafter heavy rainfall or thunderstorms (Gardeniers & Tolkamp 1985). Especially,therainwater fedupperreaches can dry up in summer. Some streams are fedbyahelocrene springandhaveamoreconstant dischargepattern. Thestudiedlowlandstreamsareup to 10 m wide and1.5m deep. Thenatural lowlandstream issituatedmainlyonasandy soil; if the sand is mixed with gravel and/or clay, amosaicpatternof substrate typesoccurs (Tolkamp 1980). The dynamic flow pattern further diversifies this mosaic (silt, detritus, leaves, dams, riffles, etc.). Thenatural lowlandstream isoftenshadedby shrubs and trees, and its longitudinal profilemeanders. Its transverse profile is irregular, differing from sandy shallows to strongly overhanging (hollow) banks. This complex of morphological and hydraulic characteristicswillbecalledthe 'stream-character'andit isbestdeveloped innaturalstreams. Regulationoflowland streamsimplies: -straightening ofthebends, -cuttingofthe trees/shrubs often followed by removal of the aquaticvegetationtwotoeighttimesperyear, -wideninganddeepeningofthe transverse profile to a standard 1:2/1:3profilewhichresults inareducedcurrentvelocity (mostly lower than5cm/s)andasilty-muddybottom, -consolidationofthebankswithconcreteblocks,wooden frames or nylonmats, -constructionofmovableweirstoadjustthewaterlevel. The 'stream-character' inregulated streams ispoorlydeveloped.

6.2

Results

6.2.1 Datacollection Biotic andabiotic datawerecollected from156sites;136sites were visited only onceand22siteswerevisited twice. Thesampleswere collected from1982to 1984 inclusive, during spring, summer and autumn. A totalof56abioticvariableswasmeasuredateachsite.

93

6.2.2 Preprocessing ofthedata Ofthemacrofaunaldata (atotalof 647 taxa) were, after careful individual weighing, 510 taxa included in theanalysis. Further informationonthepreprocessingprocedures isgiveninSection3.4.1. 6.2.3 Multivariate analysis The interpretationoftheresultsofclustering (programFLEXCLUS)and ordination (programCANOCO)ofthedataleadstothedescriptionof14 sitegroups (seebelowforfurtherexplanation). Numbers of sites, averages and standard deviations ofthequantitative environmental variables,andrelative frequency ofthenominal variables per site group are giveninTable6.1. A listofalltaxapresentwiththeir typifyingweightpersitegroup indicated (programNODES)isavailable from the author on request. Theautecology ofthemost important typifying taxawillbediscussedlater. The PCA-diagram (Figure 6.1) illustrates the relations between variables. Variables thatareprojectedbothclose toeachotherand farfromthecenterofthe diagram follow similar patterns across sites. PCA also revealed site scores. Centroidsofsitegroups, basedonthesitescores,are also indicated in Figure 6.1. The centroids of thesitegroups1,2,3and4relate tovariableswhich characterizenaturalstreams. Sitegroups5and 14 are related to variables which indicate the presenceoforganicpollutants. Site groups 7,8 and 11 are related to variables which characterize regulated streams. Site groups 9 and 13arerelated tovariables characteristic ofwaterswithatemporarycharacter (oftensampled in summer). Finally sitegroups6and12takeanintermediateposition betweennaturalandregulatedstreams. Thesitegroupsenumerated so far are orderedalongagradient (indicatedby thelineABwhichwas fittedbyeye)withanaberrantgroupof sites at the top of the figure. The positionalongthelineAB isrelated tothetransition fromnaturaltoregulated streams. Thetransitionreflects the main patterninthisset. Theresultsofordinationofsitesandenvironmentalvariables (DCCA; program CANOCO) are illustrated in Figures 6.2 and 6.3. The eigenvalues (respectively 0.41,0.26,0.18 and0.17 forthefirstfour DCCA ordination axes) indicate thatthere isasteep environmental gradientalongthefirstaxisandalesssteepone along the second axis. Thethirdandfourthaxesareless important. Thesitegroups will first be described with respect to the important related environmental variables (Table 6.1) and totheordinationresults (Figures6.2 and 6.3). Afterwardsavalidationofthegroupswill be basedonknowledgepresent oftheautecologyoftypifyingtaxa. Thefirstandmostimportantenvironmental gradientinFigure6.2 runs fromsitegroup 1(right-hand lowercorner)tositegroup 8 (left-hand lower corner). Thesesitegroupsrepresenttheextremesofthe first axis showninFigure 6.1. Insitegroup1,thefallishigh,andthe related variables, like irregular longitudinal and transverse profiles,occurrenceofmeanders andasubstrateconsisting ofstones,

94

Table 6.1 Streams; number of sites,means, and standard deviations of the quantitative environmental variables and percentages of the nominal environmental variables per site group. Abbreviations are explained in Table 10.5.

Quant itative env .ronmental variables Site group 1 Number o E

2

3

4

5

6

7

8

9

10

11

12

13

14

si tes

19

5

5

21

12

16

21

16

9

15

4

3

11

1

7. 3. 7.1 0.4 288 90 11. 1. 87. 8. 0.4 0.9 0.1 0.1 12. 7.2 33. 13. 40. 15. 18. 8. 10. 4. 9. 4. 27. 8. 37. 18. 0.9 0.4 0.1 0.1 25. 10. 11. 8.7 1.0 1.2 .00

11. 3. 7.7 0.4 403 127 11. 1. 98. 15. 0.1 0.1 0.4 0.2 8.0 1.7 34. 12. 60. 26. 15. 9. 12. 2. 8. 2. 41. 18. 47. 10. 1.9 0.6 0.4 0.3 28. 14. 3.0 3.1 4.3 1.6 .00

7. 3. 7.4 0.5 292 54 9. 2. 73. 19. 0.4 0.1 1.5 1.6 8.4 3.6 36. 12. 44. 13. 22. 4. 12. 5. 11. 3. 25. 12. 50. 42. 1.0 0.4 0.2 0.1 17. 11. 4.4 1.3 0.9 1.3 .00

8. 3. 7.5 0.4 370 49 10. 2. 86. 18. 0.3 0.3 0.5 0.4 9.3 5.6 43. 13. 49. 16. 23. 8. 14. 5. 9. 3. 38. 9.

9. 2. 7.4 0.3 392 189 8. 1. 69. 10. 2.7 1.9 6.3 5.1 5.6 2.4 124 94. 76. 27. 76. 68. 14. 5. 8. 2. 50. 21.

9. 3. 7.1 0.4 454 209 11. 2. 88. 18. 0.7 1.3 2.7 4.5 6.0 3.5 42. 15. 63. 18. 32. 31. 11. 3. 8. 2. 41. 13. 61. 34. 3.1 1.3 0.4 0.3 24. 14. 1.2 1.4 6.4 5.1 .01

9. 12. 16. 3. 3. 4. 7.3 7.3 7.5 0.3 0.6 0.5 508 510 465 168 164 179 10. 8. 6. 2. 2. 5. 84. 70. 60. 24. 19. 40. 0.5 0.4 1.2 0.8 1.0 1.6 1.6 0.9 4.8 1.8 0.7 9.0 4.9 2.0 4.2 2.7 1.4 3.8 65. 44. 35. 112 24. 10. 78. 62. 64. 20. 23. 26. 38. 34. 23. 59. 34. 24. 11. 6. 21. 3. 2. 6. 8. 6. 9. 2. 3. 4. 64. 75. 38. 21. 19. 11. 9 7 /L25."L14. 54. 93. 57. 3.7 6.2 1.6 1.8 2.3 0.6 0.4 0.9 0.3 0.2 0.5 0.1 18. 5. 4. 13. 3. 8. 0.8 0.1 2.1 1.1 0.2 1.1 6.9 7.9 2.9 3.8 5.9 1.4 .14 .11 .07

8. 3. 7.0 0.3 494 189 9. 4. 77. 27. 0.4 0.5 2.5 2.1 5.6 3.6 39. 17. 76. 20. 27. 21. 10. 3. 9. 2. 70. 34. 90. 57. 1.8 0.7 0.3 0.1 40. 130 1.4 1.6 1.5 1.0 .10

7. 2. 7.3 0.6 360 59 11. 3. 89. 28. .02 .01 0.2 0.1 9.3 7.6 18. 4. 77. 15. 12. 13. 5. 1. 6. 1. 60. 12. 43. 20. 2.0 0.6 0.4 0.1 2. 2. 0.9 0.2 2.4 1.5 .01

7. 2. 6.0 0.6 292 108 14. 2. 112 13. .03 .04 0.8 0.3 19. 16. 18. 4. 81. 42. 4. 4. 10. 2. 7. 2. 49. 19. 7. 5. 0.9 0.1 0.3 .05 0. 0. 1.6 0.1 0.3 0.2 .02

19. 1. 7.0 0.4 462 77 4. 1. 40. 16. 1.1 0.6 2.4 2.3 8.8 4.9 47. 4. 54. 16. 17. 2. 42. 16. 8. 1. 27. 4. 65. 14. 1.1 0.8 0.2 0.1 0. 0. 3.1 0.5 3.8 1.0 .00

12. 0. 0. 0. 0. 65. 82. 0.8 0.2 2 .01 .01 .10

T sd PH sd EC sd 02 sd

02% sd

0-P sd

NH4 sd NO3

sd CL sd

S04 sd NA sd

K sd MG sd CA sd HC03

sd

W sd

D sd

S sd FALL

sd SOURDIST

sd

S-T

79.'L20.

33. 2.2 1.2 0.3 0.2 33. 18. 2.8 1.5 3.8 1.4 .04

54. 3.2 1.7 0.3 0.2 27. 27. 1.9 1.7 4.0 2.4 .10

95

sd .00 .00 .01 .09 .16 .03 0 0 0 0 4 6 sd 0 0 0 1 14 %-F 0 0 0 0 1 0 3 sd 0 0 0 %-S 0 0 18 3 14 sd 1 0 31 9 20 %-E 4 1 2 2 1 sd 10 0 4 5 2 %-B 15 3 2 2 2 sd 27 4 2 2 2 %-T 20 3 22 7 17 sd 29 4 31 12 22 %MM-B 5 6 6 6 17 sd 6 8 5 9 16 %MM-E 7 11 4 8 0 sd 14 16 8 13 0 %MM-F 0 0 0 0 0 sd 0 0 0 0 0 %MM-S 3 4 8 6 5 sd 11 8 16 10 12 %MMSILT 6 14 12 11 43 sd 12 13 10 12 29 %MMSAND 30 18 24 30 19 sd 15 18 13 15 22 %MMGRAVE 11 11 16 10 3 sd 9 9 16 10 6 %MMSTONE 6 1 0 2 3 sd 8 2 0 5 7 %MMDETR 33 36 29 29 12 sd 21 15 11 13 16 %-A

Nominalenvironmental Sitegroup1 SPRING SUMMER AUTUMN SHAPLSRE SHAPLSIR SHAPIRR TEMPOR WATLF SEE COLORLES YELLOW GREEN BROWN GREYBLAC SMELL CLEAR SLTURB TURBID

.21 8 22 4 10 11 23 1 1 2 3 16 26 7 10 5 8 0 0 6 15 35 20 31 19 3 6 2 5 12 11

.16 16 18 3 6 14 27 3 8 4 9 21 32 26 20 5 12 1 4 9 13 30 26 18 19 1 4 0 0 10 14

.17 .12 .00 .02 .00 0 9 60 23 0 20 26 0 24 39 33 0 14 0 6 1 2 0 0 20 2 0 18 1 34 0 0 7 32 20 29 0 17 40 28 0 1 2 0 0 7 5 3 5 9 0 0 7 0 0 0 0 0 2 2 0 1 0 0 0 49 1 23 95 33 0 4 38 0 30 5 47 0 22 7 18 15 13 18 13 13 17 19 0 11 0 12 18 33 0 9 0 11 0 22 18 6 8 0 2 0 0 11 0 9 13 0 0 24 0 21 45 0 7 18 0 18 15 9 0 30 60 31 13 23 50 0 15 17 25 13 17 5 33 8 0 0 50 11 24 11 0 0 0 0 0 1 0 0 0 0 0 4 0 0 0 0 0 1 0 0 0 0 0 2 0 0 0 2 0 5 3 23 0 8 0 11 4 17 0

0 0 0 0 0 10 0 0 0 10 0 0 0 0 0 0 0 0 0 50 0 0 0 0 0 0 0 50 0

variables

2

3

4

5

6

53 40 0 0 47 60 11 40 89 60 0 0 0 0 0 40 0 11 0 47 42 100 0 0 11 0 0 0 0 0 89 60 11 40 0 0

40 0 60 20 80 0 60 0 0 0 60 0 40 0 20 80 20 0

52 0 48 24 81 0 57 05 05 10 81 0 10 0 10 76 19 05

67 0 33 75 25 0 17 25 25 08 25 0 25 42 42 42 25 33

56 0 44 69 31 0 0 38 31 19 56 0 13 13 13 50 38 13

96

8

9

10

11

12

13

14

43 50 0 38 57 13 76 100 24 0 0 0 14 0 48 56 29 06 05 13 62 81 0 0 24 06 10 0 19 13 38 06 38 63 24 31

22 67 11 33 0 67 78 11 0 11 0 11 0 11 22 11 67 22

87 0 13 87 13 0 87 27 40 20 73 0 0 07 07 67 33 0

100 0 0 100 0 0 100 0 50 25 75 0 0 0 0 100 0 0

100 0 0 100 0 0 100 0 33 0 100 0 0 0 0 100 0 0

0 100 0 0 0 100 100 0 0 0 0 0 0 0 0 0 100 0

100 0 0 100 0 0 100 100 0 0 0 0 0 100 100 0 0 100

7

BACTNONE100100 80100 67100 90100 78 93 75100100 0 0 BACTSL 0 BACTABUN 05 POLLUT 05 CLEANIN 100 SAND 0 SILT 0 CLAY STSAND 95 0 STSILT 74 STCM 42 STCDLE STCMCDLE 26 42 STFDPE STCMDELE 26 11 STSILT STCMI 0 STFDPESI 05 STCDLESI 16 0 MEANDNT MEANDSL 26 MEAND 74 89 REGULNT REGULSL 05 05 REGUL PSIRR 95 PS30 0 PS45 0 05 PS75 PS90 0 05 PROCON SHADOW 79 05 SURURB SURFOR 47 SURWOB 47 SURFIE 05 42 SURGRA ISOL 0

0 0 20 20 80 0 0 40 0 40 80 0 40 0 60 0 20 40 40 0 60 60 0 40 60 0 0 40 0 40 60 20 40 0 0 60 0

0 0 0 0 10 0 07 25 0 0 0 100 20 0 33 0 0 0 0 0 22 0 0 0 20 19 75 13 24 31 22 07 0 0 0 100 40 14 50 19 62 44 11 07 0 0 0 100 100 100 100 100 86 88 89 100 100 100 100 0 0 0 0 0 14 13 11 0 0 0 0 100 20 0 08 0 05 0 07 0 0 0 0 0 100 95 50 81 38 25 67 20 25 0 100 0 0 0 0 0 10 0 0 0 0 0 0 0 80 43 17 31 05 06 0 07 0 0 0 0 100 62 17 06 19 0 0 0 0 0 0 100 0 10 08 19 05 0 07 0 0 0 0 0 20 29 0 07 0 0 0 06 0 0 0 0 0 10 0 0 0 0 0 0 0 0 0 0 60 52 75 69 76 88 100 87 50 33 100 100 0 05 17 0 0 0 0 0 0 0 0 0 0 0 0 06 05 0 0 0 0 0 0 0 20 19 25 38 24 0 0 25 67 0 0 0 20 14 75 44 67 100 100 60 100 100 100 100 20 38 17 31 29 0 33 0 0 0 0 0 60 48 08 25 05 0 0 07 0 0 0 0 80 71 17 31 24 0 0 67 07 0 100 0 0 05 08 19 0 13 0 0 0 0 0 0 20 24 75 50 76 100 33 80 100 100 0 100 80 81 25 50 19 0 13 0 0 0 0 0 0 0 0 0 10 0 67 0 0 0 100 0 20 05 25 13 29 50 0 47 100 67 0 0 0 14 42 31 38 50 33 27 0 33 0 0 0 0 08 06 05 0 0 13 0 0 0 100 20 10 67 44 48 19 11 13 0 0 0 0 100 76 67 81 43 13 89 27 0 33 64 100 0 14 17 19 0 06 11 20 0 0 0 0 100 52 17 38 29 0 67 0 0 0 18 0 0 38 42 31 38 13 78 20 0 0 64 0 0 0 17 06 24 06 22 40 25 0 0 0 20 43 58 56 57 100 56 67 75 100 55 100 0 0 0 0 0 0 0 67 0 0 100 0

97

®

• 13

a BACTABUN b SMELL

STSISA

SISAND

ISOL SHAPIRR

SURFIE

PS3

° SUMMER

9. SLTURB TEMPOR MEANDNT STSILT SURWOB RECULNT

1 4a S T S A N D SHADOWSAND c l , „ c r , o bUKhUK

5CREYBLAC , B A C T N O N E CREEN p S«0 ,-> SURCRA *BACTSLjU b^-TURBID a • .^---F>S45 o fi CLAY R E C U L S L . S T C M S I SURURB WATLF »ö - - S E E 7_ POLLUT CLEANIN R E C U L STCMDELE STFDPESI B ^ T V N , ' » P R O C O N PS75 P R O C O N K b b S T C D L E MEANDSL »12 11» ' SHAPLSRE c „ n p [ STFDP|3 4 » - - > STCDLESI ' * C 0 L 0 R L E S MEANDJ T C M A u f u M N . 2 s p R | N C

SHAPLSIR

K"

_ ^ - - < ; STCMCDLE 1»

YELLOW YELLOW

""PSIRR CLEAR

"TT



X-v

s - i \

3 C

< > / 1

/

l/ah-l/

~i

f«// ^1/

£.

1/1 l\>

-^K /j=r

» /"N /~C kz->

^

^ u . \ >u-»< ^tns_

*

)

00

^

•

^

Figure 6.3 Streams; ordination (DCCA) diagram for axes 1 (horizontally) and 3 (vertically). Thecontour line describes the total variation of site scores. Forfurther explanation see Sections 3.4.2 and6.2.3, andFigure 3.9. Abbreviations are explained in Table 10.5. 100

gravelandsandarepronounced. Thesitesarealso shaded. Nitrate concentration is high, andammoniumandsulphateconcentrationsare low (Table6.1). At theothersideofthefigure (sitegroup 8) the waters are deep, wideandregulated. Thiscorrespondswithasilty bottom,adensevegetation,anaverageconductivity,andhigh calcium andbicarbonate concentrations. Sitegroups 2,3and4resemble group 1. Inthese sitegroups,however,thefallismuch lower, there is more silt, and most chemicalvariables indicatemore anthropogenic influence. Sites ingroup 2 are slightly wider and much deeper comparedwith those ingroup 1. Physically sitegroup 2tends towards groups6, 7and8. Sitegroups3and4may dry up in summer, and ammonium concentrations arehigher insitegroup 3. Sites ingroup3 are all situated in woods and have more submerged vegetation (especially mosses)thangroup4. Physically sitegroup 3issimilar togroup 1. Sites ingroup4representwiderstreams than those in group 3. Sitegroups6and7arelikegroup 8. Inthisgrouporder averageconcentrations ofammonium, nitrate and phosphate decrease whereas calcium and bicarbonateconcentrations increaseasdoesthe totalvegetationcover. Thedimensionsofsites ingroups 6and7are smaller than those ingroup 8. Inthegrouporder 8,7and6,sites becomemore irregularly shaped,aremoreoftenshaded, and the fall andcurrentvelocity increase,butarestilllowcompared tothefirst fivegroups. Thesecond importantgradientrunsfromsitegroup 9 (above the centre of Figure 6.2) tositegroup13 (atthetop). Allthesesite groupsrepresent temporarywaters. Site group 13 represents pools that remain after thenatural streamhasdriedup. Eventhesepools maydisappear. Thepoolsdatarevealedhighpotassium andlow oxygen andcalciumconcentrations. Allwere sampled insummer. Sitegroup9 iscomparable togroup 13although thesiteshavehigherammonium and phosphate concentrations and less vegetation. These organically enrichedpoolsandsomeenriched, temporary, regulated streams are shaded. The organic enrichment isillustrated inFigure 6.3 where sitegroup 9issituatedclose tositegroup 5. Sitesof groups 10, 11 and 12 aretemporary,regulated streams. Sitegroup 12issmall and thewater isslightlyacidwithlow concentrations of chloride, sodium, bicarbonate and phosphate. Although theneutralsitesin group 11arealsopoor in nutrients, their dimensions are larger. Sitegroup 10resembles group 11but ismoreorganically enriched. Thethirdgradient is illustrated in Figure 6.3. The third ordination axis shows that site group 5 consists oforganically polluted siteswhicharesituated in natural and regulated streams (Table 6.1). Average ammonium, phosphate, chloride and sodium concentrations arehigh. Thestreambottomconsists ofa thick silt layer. The water smells and itisgrey/blackandpolluted. Site group 14isnotshowninFigures 6.2 and6.3,beingoneaberrant site which isheavilypolluted. Thedatasetisdescribedby three importantgradients. Thefirst gradient is related tothephysicalandhydraulic conditions ofthe streams. Sitegroupsarenotarranged inastraight line but on an arc alongthisgradient. This isduetothemore temporary character ofsitegroups 3and4,whichbothalsofitintothe second gradient related to drought, andthepollutionstatusofsitegroups6and7 whichcorresponds tothethirdgradientrelated toorganicpollution.

101

Forabiologicalverificationoftheabove gradients, the typifying taxa (resulting fromNODES)ofeachsitegroupwillbediscussedwith respect topublisheddata. Sitegroup1isinhabitedbyanumberofrheophilic speciesliving on or in stony substratum (e.g. Amphinemurastandfussi/sulcicollis, Hynes 1977), leafpacks (e.g. Brilliamodesta,Cranstonetal. 1983; Chaetopteryx villosa, Tolkamp 1980; Pedicia sp., Pomeisl 1953), detritus (e.g. Polypedilum laetum,Tolkamp1980)orsand (Sericostoma personatum, Higler 1975). Part ofthetypical taxaimmigratefrom springareasandarecoldstenothermic (e.g. Sperchon spp., Viets 1974; Dixa maculata,lilies 1952;Parametriocnemusstylatus,Lehmann 1971). Mostother taxaarecommonand/orabundant inunpolluted small streams. Sitegroup 8isinhabitedbyalargenumberoftaxacommon in slowly flowing or stagnant waters, with a richvegetation (e.g. Anisus vortex,Macan1977;Mystacides longicornis, Hickin 1967; Ophidonais serpentina, Verdonschot 1984) and well oxygenated (e.g. Limnesia undulata,Kreuzer 1940;Anabolianervosa,Lepneva 1971). There are several interesting taxa which are often found inslowly flowing waters and/or the littoral zone of lakes (wave action) (e.g. Mideopsis orbiculare, Berg1948;Glossiphoniaheteroclita,Elliott & Mann1979;Triaenodesbicolor,Lepneva 1971). Most other taxa are common but especially their population numbers indicate a well developed,eutrophic,unpollutedenvironment (e.g. Plea minutissima, Nieser 1982; Piona alpicola/coccinea, Viets1936;Corixapunctata, Macan1976;Cyrnus flavidus,Edington&Hildrew 1981). Sitegroup 2 has much in common with group 1, including some corresponding typifying taxa (Brilliamodesta,Chaetopteryxvillosa, GammaruspulexandPolypedilum laetum). Baetis vernus, Hygrobatus nigromaculatus, Lebertia inaequalis/insignis and Paratrichocladius rufiventrisarefoundinlarge, slowly flowing streams (see Macan 1970, Smissaert 1959 and Hirvenoja 1973,respectively). Onlyone taxon (Beraeodesminuta)istypicalofsmallstreams (Lepneva 1971). Some other taxaindicatemoresaprobic conditions (e.g. Limnodrilus hoffmeisteri,Kennedy 1965). All thetypifying taxaofsitegroup 3 mostly inhabit semi-aquatic environments (e.g. Ephydridae,Scatophagidae,Merrit&Cummins1984; Eiseniella tetraedra,Brinkhurst&Jamieson 1971). Sitegroup4 isalsocharacterizedby taxa which are resistant to desiccation (e.g. Nais elinguis, Rhantus sp., Klausnitzer1984; Chaetocladiuspigeragg.,Cranstonetal. 1983). Several taxa are clearly bound towater flow (Diplocladiuscultriger,Cranstonetal. 1983; Rhyacodrilus coccineus,Chekanovskaya 1962;Micropsectra fusca, Klink 1982). Sitegroup 5is typified by oligochaetes and chironomids. Their representatives are allresistant toorganicpollutionand/orbecome abundant inorganically enriched sediments. Examples are Chironomus sp., Tubifex tubifex (Hynes 1960), Limnodrilus spp. (Chekanovskaya

102

1962), Cricotopus gr. sylvestris (Saether 1979) and Prodiamesa olivacea. Under circumstances oforganicpollutionthelastoneis especially commoninflowingwaters (Tolkamp 1980). Thetaxatypicalofsitegroup 6allhave a wide distribution with respect to water flow. They inhabitmostlylenticaswellaslotie waters (e.g. Mideopsis crassipes, Berg 1948; Paracladopelma camptolabis agg., Lehmann1971;Cryptochironomus sp.,Pinder&Reiss 1983). Sometaxacanbefound in organically enriched conditions, e.g. Limnodrilus claparedianus (Pfannkuche 1977), Paratendipesgr. albimanus andProcladius sp. (Fittkau&Roback 1983). Manymore typifying taxacanbe foundinsitegroup 7. Site group 8 is evenricher. Sitegroup7isinhabitedbymoreorlessubiquitous taxa (e.g. Cloeondipterum, Macan 1979; Potamothrix hammoniensis, Brinkhurst 1964; Bathyomphaluscontortus,Macan 1977), someofwhich arebound to a dense vegetation (Physa fontinalis, Macan 1977; Graptodytespictus,Freudeetal. 1971), otherstoasiltysubstratum (e.g. Sphaerium sp.,Ellis 1978;Sialislutaria,Elliott 1977;Caenis horaria, Macan 1970), to water flow (e.g. Hygrobates longipalpus, Viets 1936)ortothe littoral zone (e.g. Limnephilus decipiens, Lepneva 1971). Glossiphoniacomplanataoccurs inalltypesofwater, including fastflowing streams,but ismostabundant inwaterswith a largepopulationofsnails (Elliott&Mann 1979). Sitegroup 13,anextreme inFigure6.2,istypifiedby a number of common coleopterans (Agabus spp., Helophorus spp., Hydrobius fuscipes). ColdstenothermictaxaareHydroporus discretus (occurring inspringsandsmallstreams),HydroporusmemnoniusandAgabus didymus (Freude etal. 1971). Sitegroup 9has fewtypifying taxa. Psectrotanypus varius favours the sediments of small, nutrient-rich, standingorslowly flowing waters,andalso those streamswhich dry up in summer (Fittkau & Roback 1983,Schleuter 1986). Limnophorasp. occursatthebanksof streams (Johannsen 1969). Sitegroup 12isinhabitedby severaltaxa which prefer swamps and temporary pools (Limnephilus centralisandL. auricula,Hiley1976; Paralimnophyeshydrophilus andLimnophyes sp.,Cranstonetal. 1983; Hydroporus planusandH. nigrita,Freudeetal. 1971). Lumbriculus variegatus (Verdonschot1987)andParalimnophyeshydrophilus are more abundant inslightlyacidwaters. Sitegroup11hasseveral taxaincommonwithgroups 10and12. Most taxa in this group prefer small,runningorstandingwaters (e.g. Ochtebius minutus, Freude et al. 1971; Haliplus lineatocollis, Cooling 1981) some mayoftenbe intermittent (Hydroporusdiscretus, Freudeetal. 1971;Limnephilus affinis, Lepneva 1971). Some are characteristic ofsmallacidpools (e.g. Hydroporuserythrocephalus, Freudeetal. 1971;Lumbriculusvariegatus, Macropelopia spp. and Callicorixa praeusta, Bernhardt 1985) or vegetated pools (e.g. Planorbarius planorbis, Macan 1977; Polycelis tenuis, Reynoldson 1978).

103

Sitegroup 10isalso inhabitedby taxatypical of temporary waters (e.g. Aplexa hypnorum, den Hartog & de Wolf 1962; Anisus leucostoma/spirorbis, Garms 1961; Pilaria sp., Brindle 1967; Trissocladius sp., Cranstonetal. 1983;Hydryphantesdispar/ruber, Wigginsetal. 1980). Some taxa inhabit detritus-rich or muddy sediments (e.g. Zavrelimyiasp.,Fittkau&Roback1983;Dinalineata, Elliott& Mann 1979). Stylodrilus heringianus is indicative of unproductive habitats (Brinkhurst&Jamiesson 1971); Dugesia lugubris ofproductive ones (Reynoldson 1978). Finally,sitegroup 14consists ofonesitethatcontains only three taxa. Mostsignificant isEristalis sp. whichcansurviveanaerobic conditionsbymeansofair-breathing.

6.3 Discussion

of data

A typological description isbasedonarecurringcombinationof taxa under comparable environmental circumstances. Major environmental variables are the results of climatological, geological and topographical processess. The presented sitegroupsareclearlya resultofsome of these major factors. The major environmental variables within the streams studied are dimensions,durationof droughtand 'stream-character'. Withintheoccurringcombinations of these factors theloadoforganicmaterialornutrientsdiversifies. The threemajorvariablesmentioned,however,donot occur in every independent combination. Regulation not only affects the 'stream-character'butalso implies increaseofsize,andanincreased drainage capacity can resultinintermittant streamflow. So,when size increases, 'stream-character' will decrease and discharge patterns can become even more irregular. Theallochthonous energy sourceofthenaturalstream (Cummins 1973)becomes an autochthonous one, especiallywhenthewoodedbanksarecleared. Butthere ismore thanalocalinteraction. Asastreamfollows itscourse it becomes larger and regulation of anupperreachwillaffectthemiddleand lowerreaches. Theenergy input (nutrientspiralling;Wallace et al. 1977) is changed andtheflowpatternisdisrupted. Taxadepositing theireggs intheupperreachmaydisappear despite the possibility that a suitable habitat forthelarvaeexists inthemiddlereach. Regulationaswellasdrought greatly diminish the natural stream community and often result inamoreorlessubiquitous fauna,also present instagnantwaters andwhichiscapableof resisting periods ofdrought. Doesthisgeneralpatternfitthe site groups described? Site group 1 has an optimal 'stream-character'; all sitesarefedby helocrene springsthroughout theyear. These small-sizedstreams are inhabitedbyarheophilicandaspring-inhabitingfauna. Thefaunais comparable to the community of helocrene springs described by Verdonschot & Schot (1987). Asdimensions increase thespringfauna gradually disappears. A fewrepresentatives remainin site group 2 wherenotonlywidthbutespecially depth ismuchlarger. Foralongtime, streams have been influenced by man mainly through agricultural activity in thecatchmentarea. Agricultural

104

activity is especially developed in the naturally level areas. Hill-ridges are less appropriate except for themore leveltops. Sites insitegroup1areallsituatedon the steep slopes of the ridges (fall is high) and therefore their natural characteris preserved. Because springs inthis province arise from a perched groundwater table (Verdonschot & Schot 1987),discharge isfairly constant throughout theyear. Butasthese springstreams flow down hill, the slope decreases and man's influence stronglyincreases. Thisprocess isalreadypresent insitegroup2 (e.g. regulation is traceable). Since theearlier investigationswerecarriedoutmostof thecharacteristic inhabitantsoftheupperand middle reaches have disappeared (Mol1986b). Agricultural activity anddrinkingwaterextractiondraintheland and lower the groundwater table. Streamswhicharenotfedbyan apparentgroundwater aquifercanbecome intermittent. Thesesitesare found insitegroup 3. Truerheophilic taxaareabsent. These truly rainwater-fedstreamsshowavery fluctuating discharge pattern in which even larger, downstream parts may dry up (sitegroup4 ) . Sometimes small isolatedpoolsremaininthese intermittent streams in summer. These isolatedpoolsareinhabitedbyapooraquatic faunaof mainlycoleopterans,notrestricted toa running water environment. Sitegroups13and9representthesummeraspectsofsitegroups3and 4,respectively. However,thedifferencebetween9and 13isalsodue to organic load. Themacrofauna inhighlyorganically-enrichedpools in9looks likethatinsomesmall,intermittent,organically-enriched regulatedstreams. Further downstream, as dimensions increase, there are no undisturbed catchmentareasorstreams intheprovince ofOverijssel. Those larger streams with a high 'stream-character' are always influenced by nutrient/organic inputand/orareregulated in (parts of)theupperreaches. Sitegroup6consistsofsitesthathave kept the physical part oftheirnatural 'stream-character'togetherwith sites that are regulated but which have retained their current velocity. Most sites are shadedcontrary tothesitesofgroup7. Site group 7 represents regulated, slowly flowing streams with vegetation. The differencebetweensitegroups6and7isalsoseen intaxoncomposition. Sitegroup 7 is inhabited by a number of vegetation-related taxa, while group 6represents acombinationof common,sub-rheophilictaxaand indicatorsof saprobity due to the presenceofawoodedbank. Sites ingroup 8arewideandcontaina zone with well aerated open water which, together with a large littoralzoneandalow organic enrichment,explains thehightaxondiversity. Organic enrichmentcausesadecrease indiversity andachange in taxon composition. Asorganic enrichment increases,itincreasingly dominatesalltheother,eventhemajor,environmental variables and thus sites from sitegroups4,6and7begintolookmore likesite group5. Only saprobic taxaaretobefound. Finally,sitegroup 14representsa state of almost completely 'dead'water (Verdonschot 1983). Regulated, temporary small streams fed by rainwater, become slightly more acid intheabsenceofotheranthropogenic disturbance (sitegroup 12). Sites insitegroups 10and 11areneutral and the sites in group 10areevenslightlyorganically enriched. Thetaxa

105

collectedaretypical of temporary waters but their distribution pattern sometimes looks more stochastically determinedwithinthe extremeconditions ofdroughtandacidity. Therelations foundbetweenthesite groups are a complex web resulting from interactionsbetweensomemajorenvironmentalvariables andfitthe general patterns indicated earlier in this section. Within thiswebcenotypes,asdefined inchapter2,canbedescribed. Sincecenotypes shouldbedescribed independentofseason, some site groups were combined in onecenotype. Further,somecenotypesare related,oftenreflecting anincrease inamountof organic material. Thefollowingschemeofcenotypes isproposed (Figure6.4):

environmentalfactor: ••dimension (regulation) *•organic enrichment ••drought • acidity — ^ — summer aspect {_) sitegroup i i related cenotypes

Figure 6.4 Streams; cenotypological relations. The arrows the direction of increase of the respective variables.

indicate

-permanent springstreams (sitegroup1 ) , -permanent,rainwater-fedupperreachesofnatural streams (site group2 ) , -temporary,smallupperreachesofnaturalstreams (site group 3 itssummeraspect,sitegroup13), -temporary,upperreachesofnaturalstreams (sitegroup4andits summeraspect,sitegroup9 ) , -temporary,smallupperreachesofregulated streams (site group 12), -temporary,upperreachesofregulated streams (sitegroup 10 and 11), -permanent,middlereachesofsemi-natural streams (sitegroup6 ) , -permanent,middlereachesofregulated streams (sitegroup7 ) ,

106

-permanent,lowerreachesofregulated streams (sitegroup8 ) , -organically-enrichedstreams (sitegroup5), -heavilyorganically-enrichedstreams (sitegroup14). Themiddlereaches (sitegroup6)togetherwith the described upper reaches of naturalstreams (sitegroups1,2,3,13,4and9)canbe generally indicatedasnaturalstreams. Themiddle (sitegroup7)and lower (site group 8) reachestogetherwiththeupperreaches (site groups 12,10and11)ofregulatedstreamscanbegenerally indicated as regulated streams. When upper and middle reaches become organically enriched, they move into both related cenotypes of organically-enriched streams (site group 5orevenintositegroup 14). Allstreamsstudied aremore (e.g. sitegroup5) or less (e.g. sitegroup 1)influencedbyhumanactivities. Regulation,especially, hascausedadramatic change in community composition and from a social-political point of view isalmost irreversible. Inplanning theaimsandmethods ofmanagement ofnaturaland regulated streams, oneshouldkeep inmind thenaturaloriginofregulated streams. Even inregulated streams it is possible to improve 'stream-character' without losing their agricultural function. Itisunrealistic to strive foranoriginalpristine stateatallsites for both natural and regulated streamsbutthewebofcenotypes (Figure 6.4) indicates directions inwhichthestreamecosystemcanbemademorenatural. A simplificationofthepreviously givenscheme isillustrated inFigure 6.5,where transverse andlongitudinalprofilesofanaturalanda

Figure 6.5 Scheme of the transverse and longitudinal profiles natural and a regulated stream with the cenotypes indicated.

107

of a

regulated streamaregiven on which the cenotypes are projected. Planningandmanagementmaybedirected towards improvement especially because onlyabout2%ofthetotallengthofstreams in the province of Overijssel has a more or less natural 'stream-character'and corresponding community. These2%areonlypreservedbecauseoftheir geographical positions on the steepest parts of theslopes. In general,anattemptatimprovement of 'stream-character' should be directed towards the physical and hydraulic conditions. This direction isindicated inFigure6.5 byarrows. Asalreadymentioned,thepresentedcenotypesareprovisional; in Chapter 10theywillbeelaboratedwithalltheotherdatacollected. Itisdifficult tocomparetheresultsof our study with other classifications of small streamsbecause ofthedifferentapproach. Yet, themajorenvironmental factors responsible for the cenotypes have already been mentioned individually as master factors,e.g. durationofdrought (Pennak1971,Williams&Hynes 1977),water source (lilies 1955), dimensions (Strahler1957)andregulation (Gardeniers & Tolkamp 1985).

108

7. MACROFAÜNAL COMMUNITY TYPES INDITCHES

7.1

Introduction

Despite the 350,000kmofditchesto be found in The Netherlands, knowledge is limited about thedistributionofmacrofauna inthese small, shallow, linear waters which are made and intensively influenced byman (Higler1976a;Beltman1976;Beltman1984;Scheffer et al. 1984; Verdonschot 1987). There exists no integrated, systematic approach forstudying theconsiderablevariabilitybetween andwithinditches (Higler,Torenbeek&Verdonschot 1986). Ditchesarethe most abundant physico-geographical water type present in the province. Some general physico-geomorphological characteristics oftheditchesstudiedare: -aregular,linearshape, -regulatedordug, -astreamdirection innoneortwoways (velocity0-0.05 m/s), -awidthup to10manddepthupto1m. Bottom composition is regarded as one of the most important environmental variables affecting the macrofauna composition in ditches. The126samplingsiteswereapriorispreadoverthe bottom types; sand, clay, ombrotrophic and minerotrophic (fenland)peat. Withineachoccurringbottom type,thesampling sitesweresituated in ditches of different water quality, profile and other important variables.

7.2

Results

7.2.1 Data collection Macrofaunaandenvironmental data were collected from 126 sites, sampled in1981and1983. Mostsiteswerevisited once,only5sites werevisited twice,andsampling dates were spread over the four seasons. Forty sites were sampled inspring,59insummer,22in autumnand5inwinter. A totalof71 environmental variables were used inthisanalysis. 7.2.2 Preprocessingofthedata Thisstudyrevealed 666taxawhich is a high number compared with other studies inditches (Garms1961,Beltman1984,Caspers&Heekman 1981, 1982). Aftercareful individualweighing 606taxawere finally included in the analysis. Further informationonthepreprocessing procedures isgiveninSection3.4.1. 7.2.3 Multivariate analysis The interpretation oftheresultsofclustering (programFLEXCLUS)and ordination (program CANOCO)ofthedataledtothedescriptionof11

109

sitegroups. Numbers of sites, arithmetical means and standard deviations of thequantitativeenvironmentalvariables,andrelative frequency ofthenominalvariablespersitegrouparegiven in Table 7.1. The resulting listofalltaxawiththeir typifyingweightper sitegroup (programNODES)isavailable fromthe author on request. The important typifyingtaxaarediscussedbelow. ThePCA-diagram (Figure7.1) isan illustration of the correlation matrix and shows themostobviousrelationsbetweenvariables,e.g. thedevelopmentof different vegetation layers in summer (higher temperatures), the relation between conductivity, calcium and chloride,therelationbetweenwidth and depth, and so on. Site groupsarenotprojectedbecausealmostallofthemarescatteredover thewholediagramandarethereforenotrelated to certain complexes of environmental variables, exceptforsitegroups1and2b. Those groupshavehighvalues ofthevariables in the lower left corner, suchasreedland, fenlandand irregularprofile. Theordinationresults (DCCA)aregiveninFigures7.2 and 7.3. The low eigenvalues (0.21, 0.14, 0.12 and0.09,respectively,forthe firstfouraxes)meanthattheextractedenvironmental gradients are short. Thelengthsofgradientoftheaxes lie intherangeof2-3 SD suggestinga75-80% turnoveroftaxaalong theenvironmental gradients represented by theaxes. Thescores (optima)ofmost taxatherefore lieoutside theregionofthe site scores and the probability of occurrence of such taxa increasesordecreasesuniformly alongthe sampled gradients instead of being unimodal (ter Braak 1986). ComparisonofDCAandDCCAshowsonlyaslightdecrease ineigenvalues whichmeansthatthe environmental variables describe the species variability well. The species-environmentcorrelations oftheaxes areallhigh (about0.95). Themeasured environmental variables are sufficient to explainthemajorvariationswithinthedataset. The first4axestogetheraccount for approximately 25% of the known variation inspeciesdistributions alongthemeasuredvariables. The firstaxisaccounts forapproximately 9%which indicates thatthereis no importantprimaryenvironmentalgradient. Although sitescanbeclassified intogroupsduetodifferences in frequency andabundance ofthetaxapresent,thedistributionpattern ofthese taxaisnotdiscontinuous. This isshowninFigures 7.2 and 7.3 wherethedottedlines indicate thetotalvariationwithinasite group. Theellipses represent the95%-confidenceregionsofthemeans ofsite scorespergroup. Therelationspresentedby thesitegroupswillbedescribed with respect to theimportantenvironmentalvariables (Table7.1) andthe ordinationresults (Figures 7.2 and 7.3). Avalidation of the site groups will be based on knowledge present of theautecologyof typifyingtaxa. Sitegroup1 FromFigures 7.2 and7.3 andTable 7.1 itcanbeconcluded that site group 1 is related toditcheswithalargecross-section (widthand depth), oftensituatedonpeat (65%)andwithin reedland. The bank profilewasoftenvertical (65%). Samplesweremostly takeninsummer (88%)whichexplains thehighaverage temperature. Conductivitywas 110

Table 7.1 Ditches; number of sites,means, and standard deviations of the quantitative environmental variables and percentages of the nominal environmental variables per site group.Abbreviations are explained in Table 10.5.

Quantitative environmental variables Site group Number of sites

PH sd NH4 sd NO3

sd 0-P sd T-P sd 02 sd 02% sd EC sd CL sd CA sd FE sd T sd W sd D sd S sd S-T sd %-A sd %-S sd %-F sd %-E sd %-B

1

2a

2b

3

4

5

6

7

8

9

10

11

17

14

16

41

3

4

4

6

8

2

2

1

7.7 0.4 0.4 0.1 0.2 0.3 .03 .02 .23 .16 8.6 3.7 94 45 556 147 73 43 77 33 .30 .36 20 3 6.0 2.9 .59 .24 .5 1.4 .11 .25 6 15 22 22 20 17 5 7 2

7.6 0.7 0.7 1.0 0.1 0.9 .12 .04 .35 .17 8.7 5.6 93 47 369 132 61 32 60 20 .34 .33 18 4 5.2 3.1 .50 .20 .0 .0 .10 .26 5 14 45 30 68 31 20 23 3

7.1 0.9 1.0 0.7 0.7 0.1 .03 .12 .18 .18 12. 4.8 94 59 344 155 51 44 50 21 .36 .54 8 7 3.6 1.2 .42 .17 .0 .7 .18 .30 2 3 23 33 6 9 3 5 2

7.5 6.9 0.7 0.4 1.5 0.7 2.5 0.2 0.6 6.2 1.3 4.3 18 .08. .42.03 .37.33 .47.15 7.2 9.5 3.6 0.9 68 74 35 7 460348 164150 75 31 50 19 57 62 18 39 .30.17 .38.21 14 5 8 0 2.7 3.8 1.3 5.3 .35.37 .22.06 2.9 .0 16. .0 .12.35 .22.22 10 0 25 0 32 0 39 0 12 0 22 0 13 63 22 25 3 5

5.9 1.8 1.4 0.9 0.3 0.2 13 .08 .16 .11 4.8 2.9 57 37 176 119 29 22 15 12 .38 .33 24 4 3.7 1.3 .28 .17 .0 .0 .25 .50 5 10 20 40 25 50 26 34 3

5.8 1.2 5.4 2.7 3.0 2.0 27 .47 .52 .57 11. 1.7 87 17 199 38 53 46 64 78 1.5 1.1 7 2 3.6 3.5 .36 .43 .0 .0 .03 .05 30 41 0 0 13 25 7 15 3

7.7 0.9 1.3 0.9 2.0 1.8 07 .08 .32 .32 11. 5.5 94 41 469 156 133 186 63 20 .62 .84 9 6 2.7 1.3 .33 .17 .5 .8 .05 .09 16 24 11 12 26 38 9 16 1

7.0 1.1 1.8 1.9 2.2 3.3 02. .01 .08 .05 7.1 2.4 80 35 548 223 44 11 58 30 .14 .29 20 6 1.1 0.5 .13 .11 .5 .8 .08 .17 1 4 7 14 0 1 14 14 1

5.9 5.6 0.3 0.1 2.0 2.9 0.1 0.4 9.8 5.7 0.3 6.1 041.42 .03 1.9 .18 1.5 .11 1.7 12. 8.0 0.1 0.8 92 65 1 11 240 283 28 152 43 47 1 33 30 30 4 13 .10 .90 .00 .57 5 6 0 2 1.8 2.4 0.2 0.2 .45 .48 .21 .11 2.5 1.5 .7 2.1 .02 .01 .02 .00 11 31 13 42 8 1 11 0 1 3 4 1 0 0 0 0 1 1

7.6 28. 0.0 .5 11. 4.0 32 650 51 75 .60 6 1.9 .35 .0 1.0 0 0 0 0 0

111

sd %-T sd %MM-B sd %MM-E sd %MM-F sd %MM-S sd %MMSILT sd %MMSAND sd %MMPEAT sd

1 49 25 19 13 12 16 27 19 20 17 14 12 2 7 6 9

3 94 11 13 13 16 16 24 11 25 11 10 10 0 0 10 10

3 32 38 33 20 16 22 5 10 20 22 18 14 1 5 4 11

4 54 41 18 18 18 20 11 13 29 22 19 15 2 6 1 4

5 68 27 13 11 67 12 0 0 0 0 20 0 0 0 15 19

5 55 42 15 30 25 30 10 20 25 19 10 12 0 0 5 10

5 40 43 30 35 18 21 15 30 23 29 5 10 5 10 0 0

2 51 29 13 16 30 28 10 17 17 20 25 12 5 10 0 0

2 22 27 21 17 25 26 8 10 15 18 21 14 9 15 0 0

0 18 18 20 0 0 0 55 7 0 0 15 21 10 14 0 0

4

5

6

7

8

9

0 31 0 41 40 0 0 20 0 0 0 0 0 10 0 14 20 100 28 10 0 14 0 0 0 -

Nominalenvi ronmental variables Sitegroup SPRING SUMMER AUTUMN WINTER SHAPREG SHAPIRR TEMPOR WATLF SAND CLAY PEAT FENLAND SEE COLORLES YELLOW BROWN GREEN SMELL CLEAR SLTURB TURBID POLLUT CLEANIN STNONE STDEPE STSILT STDEPES PSIRR PS45 PS75 PS90 SHADOW

1

2a

2b

3

0 88 12 0 94 6 0 0 24 18 65 0 0 18 77 0 6 0 71 18 12 18 47 18 18 29 35 6 0 35 65 24

7 71 21 0 64 6 7 0 7 21 71 0 7 7 86 7 0 0 64 36 0 0 7 7 64 7 14 50 14 7 29 21

44 6 44 6 94 6 0 19 44 13 38 6 19 44 56 0 0 0 75 19 6 6 31 19 13 44 25 25 6 38 38 19

37 42 20 2 98 2 0 12 42 34 24 0 17 27 63 7 2 2 73 17 10 5 37 17 2 68 12 12 0 59 29 15

112

10

11

0 0 100 67 13 100 100 100 0 0 100 0 17 88 0 0 0 0 0 0 0 0 0 0 100 0 17 0 0 0 0 0 100 100 100 100 100 100 100 100 0 0 0 0 0 0 0 0 0 50 25 33 63 0 100 0 33 25 0 17 25 100 50 0 0 25 50 83 100 100 50 100 0 0 26 17 0 0 0 0 100 0 0 0 0 0 0 0 0 75 25 0 0 50 0 0 0 0 0 25 17 63 0 0 0 0 33 38 0 0 0 0 100 50 75 33 63 100 0 0 0 50 25 17 0 100 0 0 0 0 17 0 0 0 0 0 0 0 25 0 0 0 0 100 67 50 50 83 88 100 50 0 0 50 33 50 50 0 13 0 0 17 0 0 0 0 0 100 0 0 0 25 17 0 0 100 0 0 75 50 50 100 100 0 0 25 25 17 25 0 0 0 0 0 25 0 13 50 0 0 67 25 50 83 63 50 50 100 0 33 50 0 0 0 0 0 33 25 25 0 0 0 0 0 0 0 0 0 0 0 0 0 67 25 50 83 88 100 100 100 0 50 25 0 13 0 0 0 0 38 50 67 25 25 0 100

SURURB SURFOR SURWOB SURFIE SURGRA SURREE PROCON

0 6 6 0 71 35 0

0 0 0 0 79 21 0

0 6 6 0 94 13 13

5 0 5 0 0 0 7 0 93 100 2 0 2 0

113

0 0 25 0 50 0 0

0 0 0 0 25 0 50 0 50 100 0 0 0 0

0 0 0 38 75 0 0

0 0 0 50 50 0 0

0 100 0 0 0 0 50 0 0 50 0 0 50 0

PH

HH A

EC CA h r^

CL A

r-

STSILT A ÎMMSILT CLAY A IR

/^

•1 SHAPREC A

SURCRA

TURBID 1 SURURB p S 7 5 COLORLES A POLLUT 1 1 02 G R E E N A T -P S AUTUMN . W / n SEE 1 SAND CLEANIN -1 ^ %-E CLEAR S-T;%-F %-T< 1 ÎMM-B r r A WINTER -1 . -'WATLF p T< S90< >FE -1 > 0 P %-s< + > SURWOB >PROCON IMM-S^ D^ > N03 SHADOW'' A „.,.. c ^ S T D E P E S ^ LL- V ^ ÏMMSAND a_F^ «MM-F YELLOW ÏMM-E PS15 SMELL TEMPOR, SURREE U STNONE O SURFOR , BROWN S U R F I E SUMMER^U SLTURB U PSIRR /_ FENLAND SHAPIRR V PEAT 02?, h

r-

7SPRING

>NH4

V STDEPE V ÏMMPEAT

Figure 7.1 Ditches; ordination (PCA) diagram for axes 1 (horizontally) and 2 (vertically) with only environmental variables indicated. For further explanation see Sections 3.4.2 and 7.2.3, and Figure 3.10. Abbreviations are explained in Table 10.5. 114

SUMMER

/

/ ' - " ' •'i

" "^S\

~~'-~~^'' 1\

- ' ' "-

1,\'" 1

'/*f™BäsSte CO c

c

c o

c o

•w

4-*

ro

ro >

TS

tu TJ

TJ u.

•a i-

ro TS

T3

C

ra

+-< 0n

159

> ro

•o

ro > (U

of

CU

T3

ra

Figure9.5 Scheine of the averages and the standard deviations transverse profiles per cenotype.

JC

TS

the

ro c

ro

+-» in

isopods. The theabsenceofvegetation. Ingeneral,the biological differences between the last four groups are small andtherearemany transitionsbetweenthegroups. The sitegroupsrepresentacomplexwhichcanbe separated into three groups of related cenotypes (Figure 9.4)namely;mesotrophic swamps (group 5), acidmoorlandpools (groups1, 2 and 3 ) , and stagnant, neutralpondsandlakes (groups4,6,7,8and 9). Thefirstcenotype consistsofonlythree sitesand isaberrant inthis data set. The second group ofcenotypescanbedividedaccording topermanencyand eutrophication (Figure9.4). Thethirdgroupofcenotypes represents awebofcontinuadominatedbydimensions (Figure9.5),eutrophication andbottomcomposition (thelatter is specifically related to the complex offactorscharacteristic forthefenland region).

160

10. AREGIONALTYPOLOGYOFSURFACE WATERS IN THE PROVINCE OF OVERIJSSEL

10.1

Introduction

InChapter 1itwasstatedthat the five main physico-geographical water types,asdescribed inChapters 5-9,donotrepresentdifferent ecological types. But only through an elaboration of all the collected data this statement and the intended typology canbe justified. Therefore,alldatawereanalysed together by the same multivariate analysis techniques asusedbefore (seeChapter 3). A strategybasedontheremovalofsetsof sites (site groups) along identified gradients, and a subsequentreordination (Peet1980)is used. Simple,unidimensional environmental gradients can be easily recovered by using nonlinear ordination techniques. But asthe complexity ofadataset increases, the efficiency of information recovery decreases because thespecies-environment relationsdonot correspond tooneconsistent,geometricmodel (incasutheunimodalor Gaussian response model). Thismeansthatordinationresultsbeyond twoorthreedimensionsaredifficulttointerpret in terms of the underlying environmentalvariables. AsPeet (1980)stated 'ordination canworkwell for the identification of the most important and conspicuous environmental trends in a data set with a simple underlying environmental structurebut these would also have been deduced by a competent field ecologist'. Ordination worksless effectively atresolvingenvironmentalrelations incomplexdatasets. Peet (1980) referred to principalcomponentanalysis (PCA). These problemswithPCA are partly solved within detrended (canonical) correspondence analysis (D(C)CA)(Hill&Gauch1980,terBraak 1986) because itsresponsemodel is not linear but unimodal. However, because the technique of detrending used isnotfullyeffective, higheraxesremainpartiallydependent. Thebasicmethodology chosenwastheprogressive removalofgroups of sites. After initialordination intwodimensionsusingDCCA,the resultingdiagram isexamined forrelationsbetween site groups and environmental variables. Distinctive sitegroupsareremoved (ora continuum ispartitioned)andthe remaining sites are reordinated. Thereby, theimpactoftheoriginally observedvariable(s)isgreatly reduced.

10.2 Multivariate

analysis

Theresultsofclustering andordinationofsites (programs FLEXCLUS and CANOCO), arrangementoftaxa (programNODES),andinformationon theautecology oftaxaledto42sitegroups (seebelow). A list of alltaxawith theirtypifyingweightper sitegroup (programNODES)is given (Appendix 1). Numbers of sites, averages and standard deviations of the quantitative environmental variables, and the relative frequencyofthenominalvariablespersitegroup are given

161

in Appendix 2. The final site groups were compared with the descriptions oftheformer sitegroupsdescribed for the five major water types, namely helocrene springs (Chapter 5), streams (Chapter 6), ditches (Chapter 7), rivers,canalsandlarge lakes (Chapter 8 ) , and ponds andsmalllakes (Chapter 9). Acodewasgiventoeachnew sitegroupwhichcorresponded tothemostsimilarsitegroup formerly presented for one ofthefivemainwatertypes. Hereby,theformer site group number is combined with the first letter of the correspondingmainwater typeasfollows: helocrene springs =H streams =S ditches =D rivers,canalsandlargelakes=R pondsand smalllakes =P Acomparisonofsitegroupswithintheformerandthenew arrangement is given in Table 10.1. Statusand thenumberofsiteswhichwere replacedbetweentheformerand thenewsitegroupsare given (Table 10.1).

Table 10.1Thechanges insitecompositionbetween theformersite groupsdescribed forthemainphysico-geomorphologicalwater typesand thenewsitegroupsdistinguishedwithinthetotaldataset.

cluster number

old number sites

HI H2 H3 H4 H5 H6 H7 H8 SI S2 S3 S4 S5 S6 S7 S8 S9 S10 Sil S12 S13 S14 Dl D2A

13 6 14 3 4 3 1 1 19 5 5 21 12 16 21 16 9 15 4 3 11 1 17 14

numberof removed sites 1 0 0 1 0 0 0 1 2 0 0 0 1 4 7 16 1 0 4 0 0 0 17 5

to

numberof received sites

Sl(l) 2 0 0 H6(l) 0 0 2 0 H6(l) 0 Hl(2) 1 0 0 0 R9(l) 0 S7(3),R9(1) 0 5 R9(5),D3(2) D3(6),R4(5),R9(5) 0 D8(l) 0 10 S10(4) 0 4 0 0 R2(14),R4(3) 0 D3(2),R2(1) P5(2) 0

162

from

Sl(2)

H8(1),H4(1)

Hl(l)

S6(3),R3(2)

D3(1),D7(5),S11(4) D10(2),D9(2)

D2B D3

15 42

15 2

P8(8),R4(6) S10(1),P5(1)

D4 D5 D6 D7 D8 D9 D10 Dil D12 D13 D14 D15 D16 D17 D18 D19 D20 RI

3 4 4 5 8 2 2 1 1 1 1 1 1 1 1 1 1 26

3 4 0 5 0 2 2 0 1 0 1 0 0 0 0 0 0 11

P5(2),D3(1) P3(2),P5(2)

R2

S12(2) S12(2)

P5(l)

R9(l)

R3(1),R4(2), R9(3),R12(5)

3 20

R5 R6 R7 R8 R9

6 1 4 9 28

1 1 0 0 2

RIO Ril R12 PI P2 P3 P4 P5

1 10 0 0 9 13 14 21 3

0 0 0 0 0 0 5 1 0

P6 P7

26 21

0 17

P8 P9 PIO Pli

24 26 0 0

4 3 0 0

664

1 17

"

12 46

Total

0 0 0 0 2 0 0 0 0 0 0 0 0 0 0 0 0

S10(5)

35

R3 R4

RI3

0 13

S7(2),R4(1) 1 D8(1),R2(1),R9(1) ) R12(13),P11(4) 23 R13(l) 0 0 R9(l)

-

0 0 19

R4(1),P7(1)

-

0 0 22 1 0 2 P2(2),P4(1),P7(2) 2 P10(l) 1 8 R4(4),R12(3),P6(2), P9(1),P11(7) D3(2),R2(1),R4(1) R1(1),R4(1),R12(1)

-

163

163

163

2 3 8 1 1 12

-

0

S8(6),S7(2),D2A(2) P8(2),D4(1) 53

-

0 0 4 0 10 0 0 1 0 1 0 1 1 1 1 1 1

S9(1),R4(1)

P9(l) D1(14),D2A(1), R4(1),P8(1)

RKD

16 52 10

S8(5),D2B(6),D1(3),R1(2), R3(1),R9(1),P7(4),P8(1) 49

-

5 0 4 9

S5(l) ,S6(1),S7(5),S8(5), D2B(1 D14(1),R1(3),R4(1),R6(1) 45

-

1 10 Rl(5) ,R4(13),P7(3),P9(1) 22 R5(l) 1 9 P3(2) 15 D5(2) 11 P3(l) 21 D2A(2),D3(1),D4(2), D5(2),D12(1) P7(2) R9(1),P3(2)

11 28 7 28 24 1 12

D2B(8) P7(l) P4(l) R4(4),P7(8)

664

Thenewsitegroupswillbedescribedwithrespectto: -theordinationresults (Figure10.1 -10.5), -thechanges insitecomposition (Table 10.1), -thehomogeneity ofeachsitegroup (Table10.2)which isdefined as the average resemblance of themembersofthissitegroup toits centroid. Ahomogeneity ofonemeansthatthegroup is completely homogeneous (thesiteshaveanidenticaltaxoncomposition)andwhen itiszero,thegroup iscompletelyheterogeneous (thesiteshaveno taxaincommon). -thetaxonomicalandenvironmentaldescriptiongivenforthe former, mostcomparable,sitegroup (Chapters5 - 9 ) , -the importantenvironmentalvariables within the total data set (Appendix2andTable 10.3),and -thechanges inthelistoftypifying taxa (Appendix1). Intotalfiveordination (DCCA)runswereexecuted toanalyse themost important relationspresentwithinthetotaldataset. Fromeachrun onlytheaxes 1and2were takenintoaccount. Sitegroupswith less thanfourmembersarenotrepresented inthefigures. Forthe interpretation of the ordination results the overall ordinationcharacteristics aregiveninTable10.4. Theeigenvalue isanumberbetweenzeroandone; the higher the value, the more important the ordination axis. In DCCA, the eigenvalue isameasure ofseparationof the distributions of taxa along the ordinationaxis (between-sitevariability) (Jongmanetal. 1987). Inotherwords,itindicates the importance of each axis. Eigenvalues of about 0.3 andhigherarequitecommoninecological applications (terBraak 1987). TheeigenvalueofDCCA-run1is large forthefirstaxisandmuchlowerforthesecondaxis. This indicates

Table 10.2Theaverageresemblance (homogeneity)ofthesites inthe sitegroups.

AR

AR Hl H2 H3 H4 H5 H6

0.6719 0.5234 0.5874 0.4892 0.5694 0.3912

Sl S2 S3 S4 S5 S6 S7 S9 S10 S12 S13

AR

0.6055 0.5132 0.4520 0.5460 0.5393 0.5495 0.4471 0.4458 0.4735 0.3853 0.4687

D2A D3 D6 D8

Sitegroup,AR-Average resemblance

164

0 0 0 0

5351 5230 4750 4971

AR Rl R2 R3 R4 R5 R7 R8 R9 RH R12

0.5578 0.5510 0.5260 0.5654 0.4392 0.5058 0.6447 0.5003 0.5829 0.5813

AR PI P2 P3 P4 P5 P6 P8 P9 PH

0.585 0.645 0.496 0.481 0.344 0.550 0.513 0.601 0.511

Table 10.3 Forward selection executed at each run. (a) Per variable included is given its contribution (e) to the sum of all eigenvalues of the CCA of environmental variables previously selected, (b) The results of the significance test of avariable at its point of inclusion.

(a) RUN 1 variable FALL NO3

T W WINTER MEAND STCDLE MEANDNT %MMGRAVE

D SHAPLSIR %MMDETR

S PH EC TEMPOR %MMSAND

%-F %MM-F STSAND SUMMER SAND SURWOO REGULNT

%-T SURHEA REGUL ISOL SHAPLSRE

CL STCM FENLAND

SEE SHAPREG PS90

CA STCMDLPE

NH4 %-S 02% %-E

RUN 2

e .33 .21 .18 .18 .17 .16 .15 .15 .15 .14 .14 .14 .13 .13 .13 .12 .11 .11 .10 .10 .10 .10 .09 .09 .09 .09 .09 .09 .08 .08 .08 .08 .07 .07 .07 .07 .06 .06 .06 .06 .06

variable

W

RUN 3

e

.13 .11 D .11 NO3 .11 TEMPOR .10 T .10 SAND .08 S .08 %-T .08 STSAND .07 SUMMER .07 %-F .07 %MM-F .07 FENLAND .07 NH4 .06 %MMSAND .07 REGULNT .06 SPRING .06 02 .06 REGUL .06 02% .06 SHAPLSRE .06 AUTUMN .05 MEANDNT .06 SHAPREG .05 ISOL .05 PS90 .05 SURREE .05 PH .05 0-P .05 %-S .05 PR0C0N .05 EC .05 T-P .05 %MM-S .04 %MMSILT .04 %-B .04 %MMPEAT .04 %-E .04 STCDLESI .04 SEE .03 FALL

variable

D W

RUN 4

e

.12 .11 REGULNT .08 %-T .07 REGUL .07 02 .07 NO3 .07 SHAPLSRE .07 02% .06 AUTUMN .06 STSAND .06 SPRING .06 EC .06 %-S .06 %MMSAND .06 SAND .05 ISOL .05 SHAPREG .05 FENLAND .05 PH .05 PR0C0N .05 %-F .05 %MM-S .05 SURREE .04 %-B .04 SURGRA .04 MEANDNT .04 %MMSTONE .04 CL .04 PS30 .04 %MM-F .04 NH4 .03 %MMPEAT .03 PS75 .03 FALL .03 %-E .03 SHAPIRR .03 %MMSILT .03 S .03 CA .03 PSIRR .03

165

variable

D N03 W %-T

RUN 5

e

.09 .08 .07 .07 STSAND .06 .06 SAND T .06 %-B .06 %-F .06 SPRING .06 PS90 .06 REGULNT .06 %MMSAND .05 FENLAND .05 SUMMER .05 PS30 .05 REGUL .05 MEANDNT .05 SURREE .05 SHAPLSRE .05 %MMSTONE .04 .04 PROCON S-T .04 .04 %MM-F ISOL .04 .04 %-S PSIRR .04 .04 S .04 SHAPREG .04 FALL SHAPIRR .03 CA .03 %MM-S .03 EC .03 CL .03 SURGRA .03 02 .03 %MMPEAT .03 CLEAR .03 PS75 .03 PH .03

variable

e

.07 .07 .06 .06 W .06 D .06 SUMMER .06 T .06 PSIRR .05 ISOL .05 NO3 .05 SHAPREG .05 SURREE .05 %-S .05 SHAPIRR .04 %-T .04 STSAND .04 EC .04 02 .04 S-T .04 SAND .04 S .04 %MMSAND .04 PS45 .04 %-B .04 NH4 .04 %MM-S .04 COLORLES .04 FENLAND .04 SEE .03 GREEN .03 PS75 .03 PS90 .03 %MMST0NE .03 %MM-F .03 02% .03 CLEAR .03 SURGRA .03 CLAY .03 PROCON .03 %-A .03 SHAPLSRE SPRING REGULNT REGUL

SPRING PEAT AUTUMN SURREE 02 SURGRA %MM-S %MMPEAT %-B SHAPIRR SHADOW 0-P COLORLES PROCON %MM-E STCDLESI T-P CLEAR %MMSTONE MEANDNT STSILT STFDEPE CLEANIN CLAY %MM-B STFDPESI BLACK

.06 WATLF .06 PS75 .06 PS30 .06 %MMSTONE .06 MEANDSL .05 MEAND .05 CA .05 CLEAR .05 PS45 .05 STSILT .05 %MM-E .05 BLACK .05 STCM .04 SURWOB .04 CLEANIN .04 SHAPLSIR .04 SURFIE .04 STFDPESI .04 STFDPE .04 %MMGRAVE .04 CLAY .04 SURWOO .03 SHADOW .03 STCDLE .03 SURURB .02 .02

.03 S-T .03 CLEAR .03 CLEANIN .03 COLORLES .03 GREEN .03 %-A .03 %MM-B .03 SEE .02 STSILT .02 STFDPE .02 %MMDETR .02 T-P .02 %MM-E .02 SLTURB .02 YELLOW .02 STCM .02 POLLUT .02 SURWOO .02 TEMPOR .02 SASILT .02 BLACK .02 .02 .01 .01

.03 WATLF .03 CLEANIN .03 GREEN .02 PS45 .02 SHADOW .02 SURWOB .02 STFDPESI .02 %MM-E .02 SHAPVIR .02 COLORLES .02 AUTUMN .02 02% .02 CLAY .02 %MMDETR .02 SLTURB .02 SURWOO .01 MEANDSL .02 %-E .01 STCM .01 TURBID .01 %-A T-P POLLUT BLACK

.03 .02 .02 .02 .02 .02 .02 .02 .02 .02 .02 .02 .02 .02 .02 .02 .02 .02 .02 .02 .02 .01 .01 .01

YELLOW FALL PH %-F CA %MMDETR SLTURB SHADOW %MMSILT %MM-B STCDLESI %MMPEAT BROWN SURWOO WINTER T-P %MM-E WATLF STFDPESI STCM MEANDNT TURBID POLLUT SHAPLSIR %MMGRAVE

.03 .03 .03 .03 .03 .03 .03 .03 .02 .02 .02 .02 .02 .02 .02 .02 .02 .02 .02 .02 .02 .02 .02 .02 .02

(b) significancetest

RUN1 variable FALL W EC N03

P= .01 .01 .01 .01

RUN3 variable

P=

W REGULNT FENLAND AUTUMN SPRING 02 EC D N03

.01 .01 .01 .01 .01 .01 .01 .01 .01

cumulationof variance explained .33 .45 .56 .64 cumulationof variance explained .12 .20 .25 .29 .33 .36 .40 .43 .46

166

RUN2 variable W NO3

cumulationof P— variance explained .01 .01

RUN4 variable

P=

D SPRING T N03 SHAPLSRE SURREE CL

.01 .01 .01 .01 .01 .05 .01

.13 .24

cumulationof variance explained .09 .15 .21 .25 .29 .32 .35

02% PSIRR %-S %-F REGUL RUN5 variable SHAPLSRE SPRING D T 02 N03

.01 .01 .04 .01 .01

.48 .50 .53 .55 .56

cumulationof P= variance explained .05 .05 .05 .10 .10 .05

.07 .13 .18 .22 .26 .29

Table10.4Theoverallordinationcharacteristics forallfive DCCA -runs.

DCCA run eigenvalue axis1 axis2

1

2

3

4

5

.41 .18

.20 .17

.18 .10

.15 .08

.11 .08

percentageofvariance ofspecies abundance dataexplainedbyspecies-sampleplot axis1 6.9 4.8 5.4 5.7 4.7 axis2 10. 8.7 8.4 8.5 8.2 percentage ofvarianceofspecies-environment tableexplainedbyspecies-environmentbiplot axis1 24 16 16 13 7.8 axis 2 35 30 25 20 13 sumofall eigenvalues

5.8

4.2

3.4

2.7

2.3

sumofallcanonical eigenvalues 1.7 P= .01

1.2 .01

1.2 .01

1.2 .01

1.4 .09

167

thatthere isastrongenvironmental gradientalongthefirstaxisand a much weaker one alongthesecondaxis. Thesamefeature,though weaker,ispresent inruns threeandfour. In the other runs the dominance of both environmental gradients is more equal. The eigenvaluesdecrease fromrun1torun5,indicating a weakening of thesuccessive environmental gradients investigated. Thustheroleof amaster factororsetofmasterfactors istakenoverby severalless dominantfactors. Thebiplot leadstoapproximatevaluesofthe species abundances alone and of the weighted averagesofspecieswithrespect tothe environmental variables. The percentage of variance of species-abundance data accounted for by thespecies-samplebiplot indicates thegoodnessoffit (thedegree of approximation) of the diagram with respecttothedistributionofspeciesabundances. The percentage of variance of species-environment data in the species-environment biplot indicates the goodness of fit ofthe diagramwithrespect totherelationbetween species abundances and environmental variables. Bothpercentages areverylowforallaxes inallruns. The total isnevera100%because ofnoise in the data (see Chapter 1) and because it depends on the largenumberof variableselaborated (terBraak 1986). Areference figure for these percentages cannotbegivenyet. Thesumofalleigenvaluesofcorrespondence analysis (CA) is a measure ofallbiologicalvariationpresentonallaxes. Forexample, inrun1thissumis5.8 ofwhich0.41 isdecribedby thefirst axis. Thesumofallcanonicaleigenvalues (eigenvaluesofCCA)isthatpart of the sum of all CA-eigenvalues which is decribed by the environmentalvariables investigated. The significance ofthesumofallcanonicaleigenvalues istested (Monte Carlo permutation test). Inthefirstfourrunsthis testis significantatthe1% significance level. Thus the environmental variables are related to the distribution ofspeciesabundances, despite therelatively lowpercentages ofvariance explained by all environmental variables. Thefifthrunisonlysignificantatthe9% significance levelindicatingonlyavaguerelationbetween variables andspecies. Intheoption 'forwardselection'of CANOCO (version 3.0), the program indicateshowwelleachindividualenvironmentalvariable can explainthespecies data. The measure given is the (canonical) eigenvalue thatwouldbeobtained iftheenvironmentalvariablewould be thesingle predictor variable. The program selects the best variable and continues to producealistofhowmucheachvariable would contribute ifthatvariablewouldbeaddedtotheonepreviously selected. The measure listed is the increment inthesumofall (canonical)eigenvalues. Theprogramagainselects thebestvariable, and so on. Ateachstepthesignificanceofthecontributionofthe variable tobe selected is tested by a Monte Carlo test. The selection process is stoppedwhenthevariable tobe selected isno longer significant (P< 0.10). The increase in the sum of the eigenvalues and the results ofthetestsaregivenforeachrunin Table10.3. 10.2.1DCCA -RUN1

168

Theresultsofordinationofsites and environmental variables are shown in Figure 10.1. The firstandmost important environmental gradient inFigure 10.1runsalongthehorizontal axisand is due to several sites situated in helocrene springareasandsmallstreams (sitegroupsHI, H2, H3, H5, H6, SI, S2, S3 and S12). The most important variables that significantly describe theordinationare fall,width,electricalconductivity,andnitrateconcentration (Table 10.3). In general,theenvironmentofthese sitegroups,asappears alsofromthearrows indicated inFigure 10.1,canbecharacterizedby the following variables. Thecurrentvelocity ishighatmostsites (10-33cm/s)andmore than50%ofallsitesareshaded,arepart of a meandering water course, and have an irregular profile anda substrate ofcoarsedetritus,leaves,sandandsometimes gravel. The temperature isrelatively low (3-11°C) (autumnandspring samples). Thechemicalvariables indicateahighconcentrationofnitrate (6.314.3 mg/1) and a relatively lowelectricalconductivity (175 -302 fiS). The sitesaremostlyoligosaprobicwhilethe trophic condition fluctuates considerably. With respect tosomephysicalvariables, there isadifferencebetween sitegroups S2, S3,S4 and S12 versus all the others. Withinthelastgroup,thesitesarenarrow (0.20.8 m ) , shallow (6-14cm),andthefall ishigh (16 -75m/km)while in the firstfoursitegroups,sitesarewider (1.0 -2.2 m ) ,deeper (19 -44cm),andthefall isrelatively low ( 1 - 4 m/km). Within thefollowingdescriptionof the new site groups, this general characterization should be takenintoaccount. One should further remember thatremarksontheformer sitegroupsweremadewith respect tothewater typeunderconsideration. Forexample,aremark oneutrophication foraformer streamsitegroup (Chapter 6)was made onlywithintherangeofthephosphate concentrationsmeasured forthe stream sites,notforallsites. Itshouldnowbe considered within the general environmental characterization givenforthisDCCArun. Thesenotesapply toalltheDCCArunsmade. SitegroupHI Theenvironmentaldescriptiongivenforspringsitegroup 1 (Chapter 5)alsoapplies tothisgroupdespite thesmallchanges incomposition ofsites (Table 10.1). Theaverageresemblance ofsites in group HI is high, compared to all other groups. Group HI has themost homogeneous compositionofsites (Table 10.2). Thetaxaindicated as indifferent in spring sitegroup 1,allbecometypifyingwithinthe totaldataset (Appendix 1). Furthermore, Plectrocnemia conspersa, Beraeamaurus,Brilliamodesta,andThaumasoptera sp. (all inhabiting springsand/orhygropetric zones;lilies 1952, Hickin 1967, Lehmann 1971), among others, were added tothelistoftypifying taxafor group HI. Site group HI represents oligo- to y8-mesosaprobic helocrene springs. SitegroupH3 Thecompositionofsites inthisgroupdidnot change compared with spring sitegroup3. Theaverageresemblanceofsites ingroupH3is high. The intermediatepositionwithrespecttotheenvironmental

169

Figure 10.1 Run 1; ordination (DCCA) diagram for axes 1 (horizontally) and 2 (vertically). The selected site groups of the total data set are shown. The contour line describes the total variation of site scores. Within this contour line the site groups not indicated by a centroid and a confidence ellipse are indicated by their main physico-geomorphological water type; D = ditches, S = streams, R = rivers and canals, P= ponds and lakes. Only environmental variables with an interset correlation greater than 0.4 are shown (arrows). For further explanation see Sections 3.4.2 and 10.2.1, and Figure 3.9. Abbreviations are explained in Table 10.5.

170

variables remainedandeventhepotassium andmagnesium concentrations do not differ. ThepHindicates slightlyacidwater,andphosphate andammoniumconcentrations are slightly lower than in group HI. About onethirdofallthesitesdryup temporarily. The indifferent taxaofhelocrene spring sitegroup 3become typifying for group H3 within the totaldataset. Nineothertaxaareaddedtothelistof typifying taxaofthisgroup,amongwhichDicranota bimaculata, Dixa maculata, and Gammarus pulex also typify groupHI, andNemurella pictetigroupH5. Thisunderlines theintermediatepositionof group H3betweengroupHIandH5. SitegroupH3represents neutral to slightly acid, oligo- to /3-mesosaprobichelocrenesprings. SitegroupH5 SitegroupH5didnotchange insitecomposition compared to spring site group 5 and the environmental description ofthelatteris applicable. Theaverageresemblanceofsites in group H5 is high. All sites areshadedandsituated inwoodswithasandybottom. The currentvelocity isrelativelyhigh. The indifferent taxa of spring site group 5 togetherwithnineother taxaareadded tothelistof typifying taxaofgroupH5,amongwhichKrenopelopiasp.,Macropelopia sp., Limnophyessp.,andPediciarivosaalsotypifygroupH3. Agabus guttatus,another typifying taxon,isoftencollected fromspringsand streams (Klausnitzer 1984). Thetypifying taxonSialis fuliginosais limited initsdistributiontofaststreamsandupperreaches (Elliott 1977). SitegroupH5represents slightlyacid,oligo- to /3-mesosaprobic, oligo-ionichelocrenesprings. SitegroupSI Theexchange ofsomesitesbetweengroupsSIandHIindicatesamutual resemblance (Table 10.1). The description ofstream sitegroup1 (Chapter 6) also fits group SI although in the latter the orthophosphate concentration is relatively high and thesulphate concentration isrelatively low but higher than in the helocrene springs. Ofallthestreamsitegroupsgroup SIhasaslightlyhigher fallbut itisnotcomparable tothatofthespringsites. The site composition of group SI ishomogeneous (Table 10.2). Fifteentaxa typifybothgroupsHI and SI, again indicating their similarity. Another fifteen taxa solely typifygroup SI. Mostofthetypifying taxahavealreadybeendescribed forstreamsitegroup 1 except, for example, Glyphotaelius pellucides which inhabits leaf packetsin woodland pools and lowland streams (Hiley 1976), Rhyacodrilus coccineus which occurs in silty and silty sand substrata (Chekanovskaya1962), Eukiefferiella sp. which primarily inhabits flowing waters (Cranstonetal. 1983), andConchapelopia sp. which isapoly-oxybiontic,moreorless cold stenothermic inhabitant of flowing watersand lakes (Fittkau&Roback 1983). Two taxahave lost their typifying character (Odagmia ornata and Halesus dlgitatta/radiatus). SitegroupSIrepresentsoligo- to/3-mesosaprobic springstreams.

171

SitegroupS2 SitegroupS2isidentical,insitecomposition and description, to stream site group 2. Thefall ismuchlowercompared tothegroups previouslydescribed. Theelectricalconductivity and the sulphate concentration arerelativelyhigh. All typifying taxaofstreamsite group 2havekept theirtypifyingweight,exceptfor Brillia modesta which is also abundant in springs. Some newtypifyingtaxaare Potamophylaxsp.,agenus inhabiting clean, well aerated, flowing waters (Hiley1976)andPolypedilumbreviantennatum,aninhabitantof flowingwaters (Lindegaard-Petersen 1972). SitegroupS2representspermanent,rainwater-fed, /3-mesosaprobic upperreachesofnaturalstreams. SitegroupS4 Stream sitegroup4didnotchangeinsitecompositionanddescription within the total data set (Table 10.1). WithinthisDCCA-runthe concentrations ofbicarbonate andchlorideofgroup S4arehigh. The current velocity andelectricalconductivity arealsohigh. However, theammoniumconcentration islowcompared toother streamgroups but higher thaninthehelocrene springs. Someofthesitesare temporary (57%). Theindifferent taxaofstreamsitegroup4alsoprove to be typifying for group S4 within the wholedataset,exceptforthe commonTubificidae. Thenewtypifying taxaareStenophylax sp. (an inhabitant of fast flowing streamswhichcandiminish ordryupin summer,Hiley 1976)andStictochironomus sp. (aninhabitant of soft or sandy sediments in streamandlakes,Pinder&Reiss 1983). The morecommontaxa,Rhantus sp. andMicropsectra sp.,areremoved from thelistoftypifyingtaxa. SitegroupS4represents temporary,/3-mesosaprobic upper reaches ofnaturalstreams. SitegroupH2 This group H2 completely corresponds with the composition and description of spring site group 2. Despite itslargeconfidence ellipse (Figure 10.1), theaverageresemblance isintermediate (Table 10.2). Several newtypifying taxaofgroupH2alsotypifygroup H3, suchasPtychoptera sp. andKrenopelopiasp. Several common taxa, like Tubifex tubifex, Limnodrilus udekemianus, Polycelis sp.,and Chironomussp. provenottobe typifyingofgroupH2within thewhole dataset. SitegroupH2represents temporary or desiccating, neutral to slightly acid,/3-mesosaprobic seepagemarshes. SitegroupS3 SitegroupS3isidentical,insitecomposition and description, to stream site group 3. Formerly indifferent taxalikeNemouracinerea andDiplocladius cultriger (bothbound to water flow, Hynes 1977, Cranston et al. 1983)togetherwithNatarsia sp. (aninhabitantof streams,springsandthelittoralzone of lakes, Fittkau & Roback 1983)areaddedtothelistoftypifying taxaofgroupS3.

172

Sitegroup S3represents temporary, a-mesosaprobic, small upper reachesofnaturalstreams. SitegroupH6 SitegroupH6 isacombinationofsites (springsitegroups4, 6 and 8), allsituatedonthewesternhillridge (Figure1.6). Theaverage resemblanceofsitesingroupH6is low, indicating that it is a heterogeneous group. Theaveragetemperature islowbecause thesites ofgroupH6arefedbyasmallgroundwater aquiferwith a small and seasonal (only in winter) discharge. Theionicconcentrations are low, exceptforsulphate. This indicatesa low retention time for rainwater intheaquifer. The isolated sitesaresurroundedbywoods (60%)orheath (40%), andaresituatedonasandybottom. Noneofthe sites discharges in a truestream. Somecommoncoleopterans,like Anacaenaglobulus andHelophorusaquaticus/grandis,are removed from the list of typifying taxa andtheformerly indifferentlarvaeof Hydroporinae have become typifying group H6. In general, taxa composition ispoor. SitegroupH6represents temporary,acid,oligo-ionic, oligo- to /3-mesosaprobicseepagemarshes. SitegroupS12 Sitegroup S12 isacombinationofstreamsitegroup 12andbothditch site groups 9and10. Mostsitesofgroup S12maybecome desiccated (71%), areslightlyacid,andoccur inregulated,andrelatively small streams. The environmentaldescription,givenforstream sitegroup 12, stillholds. Thecalcium,chloride,andsodiumconcentrations are low but as for all other major ions, concentrations arehigher comparedwithhelocrene springs. Thesulphateconcentration ishigher compared with all other stream site groups. The phosphateand nitrogenconcentrations are high, indicating eutrophication. The sites of group S12occur inveryslowly flowing streamscoveredwith filamentousalgae (22%)and situated in an agricultural landscape (grassland 71%, woods 29%). Theyweresampled inspring. Theditch sitegroupswereaddedduetotheircurrent flow and slightly acid character. They wereresponsible forthelowaverage resemblanceof sites ingroup S12. Several taxatypifying streamsitegroup 12prove not to be typifying of group S12withinthewholedataset(e.g. Limnophyes sp.,Limnephilusbipunctatus, Hydroporus nigrita, Sigara striata, andlarvaeofHydroporinae). Ontheotherhand,Polypedilum uncinatum (aninhabitantofoligotrophicwater,Moller Pillot 1984), andLimnephilusvittatus (aninhabitantofsandyorsiltypools,Hiley 1976)areaddedtothelistoftypifying taxaalongwithseveral taxa formerly typifyingofditchsitegroups9and10,suchasStylodrilus heringianus,Hydroporuspubescens,andLimnephiluscentralis. Sitegroup S12represents temporary,slightlyacid, a-mesosaprobic upperreachesofregulated streamsorditches.

Thesecondenvironmental gradient in Figure 10.1 runs along the vertical axis and isduetoanumberofsiteswhichareoligo-ionic, oligotrophic,andacid. Themost importantenvironmental characterof

173

the three distinguished site groups arethelowconcentrationsof major ions,alowelectrical conductivity (72-122/uS),alowpH (45), and lowconcentrations ofphosphate (t-P: 0.06 -0.13 mg/1)and nitrate (0.1 -0.2 mg/1). Mostsitesarewellvegetated (65 - 97%), especially in the submerged part (20 - 40%). The water level fluctuates (33 -67%),thesitesare mostly isolated (82 - 100%), often shaded (33 -55%),andsituatedonanombrotrophicpeat (3353%)orasand (56-73%)bottom. Thesitesaresurrounded by heath (27-67%)orwoods (55 -89%). Theywereoftensampled inspring (36 - 89%). SitegroupPI SitegroupPIrepresents theformerpond site group 1 and did not change insitecompositionanddescription (Table 10.1). Theaverage resemblance ofsitesingroupPiis high, indicating a homogeneous site group composition (Table 10.2). The high average ammonium concentration is the highest in the total data set. All taxa typifying pond site group 1 remain typifying of group PI. Psectrocladius sp.,aneurytopic genus (Cranston et al. 1983) is added. Site group PI represents temporary, acidified, oligo-ionic, a-meso- topolysaprobic,mesotrophicmoorlandpools. SitegroupP2 Thesmallchanges incompositionofsitegroupP2compared with pond site group2didnotalteritsenvironmentalcharacter. Theexchange oftwositesofpondsitegroup 3indicates the resemblance to site group P3. Group P2hasahomogeneous sitecomposition. Limnophyes sp. provednot tobe typifyingofthisgroupwithin the whole data set. Sometaxawereaddedtothelistoftypifying taxaofgroup P2, like Telmatopelopia nemorum (an inhabitant of temporary, acid, oligotrophicwater;Fittkau 1962),Holocentropus dubius (aninhabitant ofstillwater,Edington&Hildrew 1981),Naiscommunis and Agrypnia varia (bothinhabitantsofshallow,stagnantwater;Timm 1970,Sedlak 1985), andHydroporuserythrocephalus (aninhabitantofacid moorland pools,Freudeetal. 1971). Site group P2 represents permanent, acid to acidified, oligo-ionic, a-mesosaprobic to polysaprobic, mesotrophic moorland pools. SitegroupP3 Thecompositionofsitegroup P3,comparedwith pond site group 3, shows onlyminorchanges. Itshouldbenoted thatthesiteshave,on average,athicklayeroforganicmaterial. The list of typifying taxaofgroupP3islongcomparedwithpondsitegroup 3 (Appendix1 ) . Several taxawhichalsotypifygroupPIand/or P2,wereadded to this list. Therearesome interestingnewtypifyingtaxaofgroup P3, e.g. Corixapunctata (a widely distributed corixid also occurring in dystrophicwater,Bernhardt 1985),Rhantus suturalis (aninhabitantof stagnant,vegetatedwaters;Freudeetal. 1971), Dixella aestivalis (an eurytopic species also recorded from acid,oligotrophic peat

174

pools;Disney 1975), and Notonecta viridis (collected in small, temporarymoorlandpools;Bernhardt 1985). Site group P3 represents permanent, slightly acid to acid, oligo-ionic,a-mesosaprobicpools. 10.2.2DCCA -RUN2 Theresultsofordinationofsitesandenvironmentalvariables of the second DCCA-run (in which,thesitegroupsdiscussed forrun1were skipped fromtheanalysis)are illustrated inFigure 10.2. Firstly, the aberrant site group D6, situated inthelowermiddleofthis figure,isdiscussed. SitegroupD6 SitegroupD6didnotchange insite composition compared to ditch site group 6. Despite the slightly acidconditionandthelower electrical conductivity,theaverageconcentrations ofmostmajorions and iron are highcomparedwithallothergroups. Withinthisrun, thephosphate concentration isnothighand thenitrate concentration is relatively low. Theaveragecoverpercentageoffilamentous algae ishigh (30%)comparedwithallothergroups. Several taxa, formerly typifying ditch site group 6, areremoved (e.g. Ptychopterasp., Chaetocladius sp., Culiseta sp., and Chaoborus crystallinus) but widespread taxa (e.g. Agabus sturmi, Haliplus heydeni, and Limnephilus marmoratus) become typifying of group D6, including Hydroporus angustatus (an acidophilic inhabitantofshadedwoodland pools;Freudeetal. 1971), Agrypnia pagetana (an inhabitant of slowly flowing waters and lake shores,Tobias&Tobias 1981),and Hebruspusillus (often collected from Sphagnum vegetation, Nieser 1982). SitegroupD6 represents acid, oligo-ionic, a-mesosaprobic to polysaprobic smallditches.

Thedifferencesbetweenthefirstandsecondenvironmental gradients, asjudgedby theeigenvalues (Table 10.4),aresmallanddivideFigure 10.2 intofourmainparts. The lower right corner of the figure consists of small,shallow,temporarywaters (groups D8, S9,S10and S13). Theupperrightcornerconsistsofstreamswithanincrease in current velocity and fall, goingfromgroupR9, S7, S6toS5. The uppermiddle sectionofthe figure shows a group of large, very slightly flowing tostagnantwaters. Finally,theremainingsitesare allsituated inthemiddleandleftofthefigure. Widthand nitrate concentration appear to be significantly describing theordination (Table 10.3). Ingeneral,theelectricalconductivity ishighforall sites (400-533 ßS) which isalsothecase fortheconcentrations of ammonium (1.6-2.5mg/1), totalphosphate (0.45-2.53 mg/1) and major ions. The nitrate concentrations are lower than those in the helocrene springsandtheupperreachesofstreamsbuthigher than in most other groups. Thedimensions arelarger thaninthehelocrene springsand theupperreaches of streams. Several groups have a medium currentvelocityduetoaslight fall. Thebottomconsistsof sand (80%)and ispartially coveredwithsilt (>50%). Oftensitesare

175

Figure 10.2 Run 2; ordination (DCCA) diagram for axes 1 (horizontally) and 2 (vertically). The selected site groups of the reduced data set are shown. The contour line describes the total variation of site scores. Within this contour line the site groups not indicated by a centroid and a confidence ellipse are indicated by their main physico-geomorphological water type; D = ditches, R = rivers and canals, P = ponds and lakes. Only environmental variables with an interset correlation greater than 0.4 are shown (arrows). For further explanation see Sections 3.4.2 and 10.2.2, and Figure 3.9. Abbreviations are explained in Table 10.5. 176

surroundedbygrasslands (>40%)andhave a steep profile (25-64%). The arrows pointing to the upperleftandlowerrightcornersare associatedwithvariationspresentwithintheremaininggroups. SitegroupD8 SitegroupD8 iscomparable toditchsitegroup8, although most of the percentages givenearlierdifferbyabout10%. Thepresenthigh totalphosphate concentration (Appendix2)isduetoonlyoneaberrant site (thereisahighstandard deviation). Theaverage temperatureis high (indicatingsummer samples). Several taxa, like Micropsectra sp., Hydroporus striola, Colymbetes fuscus, andChaetocladiussp. provenottobe typifyingofgroupD8withinthetotaldataset. The new typifying taxa of groupD8areHydroporusumbrosusandRhantus sp., both also typifying of group PI. Haliplus lineatocollis, although widely distributed (Freudeetal. 1971)alsotypifiesthis group. Site group D8 represents temporary, very slightly flowing, a-meso-ionic,a-mesosaprobicsmallditches. SitegroupS13 Sitegroup S13represents streamsitegroup13withinthe total data set. Theenvironmentaldescription isstillappropriate. Besidesthe alreadymentioned typifying taxa of stream site group 13, Aplexa hypnorum (an inhabitant of temporary waters,denHartog&deWolf 1962)andLaccobiusbiguttatusareaddedtothelistoftypifying taxa of group S13. Only the genus Culex isremoved fromthelistof typifying taxacompared tostream sitegroup13. Sitegroup S13represents thesummer aspect with a-mesosaprobic conditions oftemporary,smallupperreachesofnaturalstreams. SitegroupS10 Sitegroup S10isacombinationofstreamsitegroups 10and 11, and ditch site group 7. All three formerdescriptions resembleeach other. The sitesoccur inregulatedwaters (88%),aretemporary (76%) and have a sandy bottom (96%) withasiltysubstrate (80%). The phosphate andnitrogenconcentrations are average within the whole data set. Most sites are situated ingrassland (72%). Thetotal vegetationcover isrelativelyhigh (41%),partlyduetothe presence of filamentous algae (19%). Thesamplesweretakenmainly inspring (88%). Severaltypifying taxaofstreamsitegroup 10provenottobe typifying of group S10withinthewholedataset (e.g. Stylodrilus heringianus,Hydrobius fuscipes,andLaccobiusminutus). Almost none of the taxatypifying streamsitegroup11orditchsitegroup 7are typifyingofgroup S10. Thenewtypifying taxaare Macropelopia sp. (an inhabitant of fine sediments of coolwaterbodies,Fittkau& Roback 1983),Hydryphantus sp. (an inhabitant of small, still or slowly flowing waters; Davids 1979), and Tubifex tubifex (an inhabitant of extreme, often organically enriched environments; Verdonschot 1987). Sitegroup S10represents temporary,a-mesosaprobic, flowingupper reachesofregulated streamsorditches.

177

SitegroupS9 Theenvironmentaldescriptionof group S9 corresponds to that of stream site group 9. The site composition ofgroup S9isquite heterogeneous (Table 10.2). Thenewtypifying taxaof group S9 are Agabus didymus (a commonly distributed taxon whichpreferssmall ditches,Freudeetal. 1971),Veliacaprai (a common inhabitant of shaded upper andmiddlereachesofstreams,Bernhardt 1985), Smittia sp. (amostly terrestrial livinggenus,Cranstonetal. 1983), and thesemi-aquatic genusTipula (Theowald 1967). Sitegroup S9 represents the summer aspect with a-meso- to polysaprobic conditionsoftemporaryupperreachesofnatural streams ortemporary,a-meso- topolysaprobic regulatedstreams. SitegroupS5 Theenvironmental descriptionof group S5 corresponds to that of stream sitegroup 5. Notethehighaveragecurrentvelocity (26cm/s) andtheprofile consolidation (63%)which isoftenfound. To thelist of typifying taxa of stream site group 5,thefollowing taxaare added;Dicrotendipes gr. notatus (a common inhabitant of stagnant waters,Fittkau&Reiss 1978), Conchapelopia sp. (amoreorlesscold stenothermic inhabitantofflowingwaters andlakes,Fittkau& Roback 1983), and Tubificidae. Limnodrilus profundicolaprovednottobe typifyingofgroupS5. Sitegroup S5representspolysaprobicupperandmiddlereaches of naturalandregulatedstreams. SitegroupS6 Thedescriptionofstreamsitegroup6corresponds to the group S6. Group S6 has a lower averagephosphateandnitrogen concentration comparedwithgroup S5andthesiltlayer ismuch thinner. Additional typifying taxa of group S6areMicropsectrasp. (aninhabitantof muddydesposits instreamsandsmallrivers, Pinder & Reiss 1983), Dicranota bimaculata (an inhabitant ofstreams,Tolkamp 1980),and Hygrobatus longipalpus (aninhabitantofflowingwaters,Viets 1936), Limnodrilus hoffmeisteri and Tubifex tubifex (bothassociatedwith organic enrichment,Kennedy 1965), andAnabolianervosa (aninhabitant of slowly flowing,clearwaterorsunlitstreams;Lepneva1971)among others. Site group S6 represents a-mesosaprobic middle reaches of semi-naturalstreams. SitegroupS7 Duetothechanges insitecompositionofstreamsitegroup 7compared to group S7, the averages of some environmental variableshave changed. ThisgroupS7hasa heterogeneous site composition. The ammonium, nitrate, calciumandbicarbonate concentrations arehigher andthephosphateconcentration lower,comparedwithgroup S6. Width and depth areslightly larger thaningroup S6. Mostsitesarepart ofregulated streams (84%)and thecurrentvelocity is moderate (22 cm/s). The new typifying taxa of group S7areParatendipesgr.

178

albimanus,Limnephiluslunatus,andAnabolianervosa (all three taxa also typifyinggroup S6),Hydropsycheangustipennls (ataxonwhichis quitecommoninstreamsandtolerateshigh temperatures, low oxygen concentrations andlowwatervelocities;Edington&Hildrew 1981),and Sigaradistincta (which is common in vegetated waters; Bernhardt 1985). Thewidely distributed taxa,formerly typifying streamsite group 7, Bathyomphalus contortus, Planorbarius corneus, Physa fontinalis,Erpobdellaoctoculata,Sigarastriata,andCloeondipterum provenottobe typifyinggroupS7. Site group S7 represents a-mesosaprobic middle reaches of regulatedstreams. SitegroupR5 Insitecompositionanddescription,sitegroupR5 is comparable to river site group 5 (Chapter 7)although itscompositioninsitesis fairlyheterogeneous (Table 10.2). Thelargedimensionsof sites in this group contrast with those in the other groupswithinthis DCCA-run. Notethatthere isnorelationtoa bare sandy substrate (0%; Table 10.2)althoughsuggestedby thearrow inFigure10.2. The ammoniumandphosphate concentrations are low within this run and there is no current. Both taxaindifferent inriver sitegroup5 become typifyingofgroupR5withinthewholedataset. SitegroupR5representsa-mesosaprobic, fairly large regulated riversorstagnantcanals. SitegroupR9 Therearemanymore sites ingroupR9 than in river site group 9. Most ofthesitesaddedformerlybelonged tostreamsitegroups7and 8. Although riversitegroup9issmallcompared tootherriver site groups,groupR9,withcomparabledimensionsasriver sitegroup 9,is muchlarger thangroup S7. Thesampleswere takeninspring (44%)and autumn (53%). The nutrient concentrations aremostlyhighandthe currentvelocity ismoderatedespite therelativelyminor fall. The formerly typifying butwidelydistributed taxaofriversitegroup9 (e.g. Cloeondipterum,Procladius sp.,and Neumania deltoides) and taxa more abundant in upper and middle reachesofstreams (e.g. Micropsectra sp.,Baetis sp., Centroptilum luteolum, Conchapelopia sp., andAulodriluspluriseta)areremoved fromthelistoftypifying taxaofgroupR9. Formerly indifferenttaxabecometypifyingofgroup R9 withinthewholedataset (e.g. Cryptochironomussp.,Limnodrilus claparedianus,Hygrobates longipalpis (allthree taxa also typifying group S6), Polypedilum gr. bicrenatum (an inhabitant oflakes, Brundin 1949), andValvatapiscinalis (acommoninhabitant of slowly runningwatersofallkinds,Macan 1977). Site group R9 represents ar-meso-ionic, a-mesosaprobic lower reachesofregulated streamsorslightlyflowingvery smallrivers. 10.2.3DCCA -RUN3 Theresultsofordinationofsitesandenvironmentalvariables of the third DCCA-run (forwhichthesitegroupsdiscussed forruns 1and2

179

wereremoved from theanalysis)areillustrated in Figure 10.3. In general, the sitegroupsdistinguishedwithinthisDCCA-runhavelow averagenitrate concentrations (0.1-2.0 mg/1) and high bicarbonate concentrations (80-197 mg/1) duetoarichvegetation (>50%)atthe momentofsampling. Thesampleswere takenmainly in summer, which also explains the higher temperatures. One site group (R7)is aberrant. Ahighnumber of variables describes significantly the ordination (Table 10.3). Mostarrows inthediagramareexplained in thedescriptions ofthedifferent sitegroups,suchasthe increase in dimensionsanddecrease invegetationcovergoingfromgroup P5toD2A toD3orthelowelectricalconductivity ofgroupP4orthehighpHof group R7. SitegroupR7 Site group R7 completely replaces river site group 7. The environmental characterizationofthelatteralsoapplies togroup R7. Limnophyes sp. andStictotarsus duodecimpustulatus provednot to be typifying of group R7withinthewholedataset. Thenew typifying taxa of group R7 are Potamothrix moldaviensis (see group R5), Arrenurus sp., Eylais sp., Limnesia sp.,Dreissenapolymorpha (an inhabitantoflarge slowly flowingorstagnant waters, Ellis 1978), Pionaalpicola/coccinea,andOulimniussp. SitegroupR7 represents oligo- to y8-mesosaprobic, medium to fairly large stagnantcanals. SitegroupP5 SitegroupP5isacompilationofpondsitegroup 5andseveral sites from ditch sitegroups4and5. Group P5has themostheterogeneous compositionofsites of all the groups (Table 10.2). Both the environmental characterization andlistoftypifying taxaofpondsite group 5markedly changed forgroupP5. The sites of group P5 are small and veryshallowwithadensevegetation (57%),especially the submergedpart (46%). Thebicarbonate concentration ishigh and the oxygen, potassium and phosphate concentrations areslightly lower compared toothergroupswithinthisrun. Mostsites have a linear shape (73%)andathicksiltlayerorareswampy. Thetypifying taxa ofgroupP5areDixellaamphibia (an inhabitant of sedge and reed swamps, hydroseres and emergent beds ofvegetation;Disney 1975), Agabus sturmii,Helophorus minutus, Haliplus flavicollis, Enochrus testaceus, Helophorus sp.,andHydroporuspalustris (allsixtaxaare widelydistributed coleopterans,Freude et al. 1971), Xenopelopia nigricans (an inhabitantofsmallwaterbodies andthelittoralzone oflakes,Fittkau&Roback 1983), Arrenurus fimbriatus, Segmentina nitida (aninhabitantofdrainageditches inmarshes andoccasionally inponds,Macan 1977), Planorbarius corneus, and the taxa already discussed: Haliplus ruficollis, Odontomyia sp., Enochrus melanocephalus, Enochrus coarctus, Bathyomphalus contortus, Cyphonidae,andPlanorbisplanorbis. SitegroupP5 represents permanent, a-mesosaprobic, eutrophic, very shallow (swampy),smallditches.

180

Figure 10.3 Run 3; ordination (DCCA) diagram for axes 1 (horizontally) and 2 (vertically). The selected site groups of the reduced data set are shown. Within this contour line the site groups not indicated by a centroid and a confidence ellipse are indicated by their main physico-geomorphological water type; R= rivers and canals, P = ponds and lakes. Only environmental variables with an interset correlation greater than 0.35 are shown (arrows). For further explanation see Sections 3.U.2 and 10.2.3, and Figure 3.9. Abbreviations are explained in Table 10.5. 181

SitegroupD2A Despite the loss of five sites (Table 10.1), the environmental description of ditch site group2Aapplies togroupD2A. Compared with group P5, sites are wider and deeper with much floating vegetation (77%). Theaverageconcentrations ofmajor ions,nitrate andammoniumarerather low. Thesubstrate often consists of fine detritus and minerotrophic peat (89%). Thenewtypifying taxaof groupD2A are, among others, Argyroneta aquatica and Hydroporus erythrocephalus (both taxa alsotypifygroupPIand P3),Hydroporus angustatus (anacidophilic taxon inhabiting Sphagnum and woodland pools, Freude etal. 1971), Agabusundulatus,Xenopelopianigricans (seegroup D6),Dicrotendipes gr. lobiger, and several widespread molluscs and coleopterans. Odontomyia sp. andColymbetessp. are removed from thelistoftypifying taxacompared toditch site group 2A. SitegroupD2Arepresentspermanent, /3-meso- to a-mesosaprobic, small,shallowditches. SitegroupD3 Thechanges insitecompositionofditch site group 3 compared to group D3 didnotaffecttheenvironmentalcharacterization. Several sitesofstreamsitegroups7andespecially 8 are included. Most sites havearather steepprofile (53%)andaresituated ingrassland (94%). Thedimensions arecomparable tothoseofsites in group D2A but thestandarddeviations aremuchlarger indicatingawiderrange. Thenitrate concentration islow,thesitesarelessvegetated andthe oxygen content islowercompared togroupD2A,but theconcentrations ofnutrientsandelectricalconductivityarehigher. Thisindicatesa greatereutrophicationcompared togroupD2A. Somenewtypifyingtaxa of group D3 (e.g. Xenopelopia nigricans, Haliplus ruficollis, Arrenurus globator, Bathyomphalus contortus, Planorbariuscorneus, Noteruscrassicornis,andStagnicolapalustris)also typifygroupD2A. Other new typifying taxa are, e.g. Polycelis sp., Hydroporus palustris,Lymnaeastagnalis,Bithynia leachi, Planorbis planorbis, Sphaeriumsp.,Erpobdellaoctoculata,andGlossiphoniacomplanata(all commoninhabitants ofvegetated,stagnantwaters;Freudeetal. 1971, Macan1977,Reynoldson1978,Dresscher&Higler 1982). SitegroupD3representspermanent,a-mesosaprobic, shallow,small ditches orstagnantregulatedstreams. SitegroupP4 SitegroupP4iscomparable,insitegroupcomposition,to pond site group 4. Thepresentedenvironmental description strongly dependson thecomparisonwithpondsitegroups PI, P2and P3. In general, the electrical conductivity andtheconcentrations ofnitrate,oxygenand sulphate arelowandphosphateconcentrations arehigh inthesesmall, isolated pools. The new typifying taxaofgroupP4areChaoborus obscurus,Chaoborus flavicans, Notonecta obliqua (all three also typify group P3),Oligotrichastriata (aninhabitantofstagnantand slowlyrunningwaters,usually swampy,withbrownpeaty water, often in peat deposits and draining ditches; Lepneva 1970), Acroloxus

182

lacustris (acommonly found taxon attached to vegetation in hard water, Macan 1977), and juvenile Coenagrionidae andtheirparasite Arrenuruscuspidator (Stechmann 1978). TheCyphonidae andTubificidae were removed from thelistoftypifyingtaxaofgroup P4withinthe totaldataset. SitegroupP4represents slightlyacidtoneutral,a-mesosaprobic, vegetation-rich,small,shallowpools. SitegroupP6 The sitesofgroupP6havearelatively low phosphate and nitrogen concentration, and a high oxygen concentration,although theyare stilleutrophic. Theenvironmentaldescriptionofpond site group 6 is applicable to group P6. ThecompositionofsitesofgroupP6is reasonablyhomogeneous. Thesites have a thick organic or silty substrate and a regular shape. Thenewtypifying taxaare,among others,Holocentropus dubius (atrichopteran restricted to stagnant waters, Edington & Hildrew 1981),Gyrinusmarinus (acommontaxon, Freudeatel. 1971),Agrypniapagetana (a characteristic inhabitant of lakes, Lepneva 1970), Parachironomus gr. arcuatus (aninhabitant ofallkindsofwater, locally abundant in small peaty lakes on Stratiotes; Higler 1977), Ecnomus tenellus (aninhabitantofstagnant and slowly flowing waters, in plant thickets; Lepneva 1970), Ablabesmyia longistyla (an inhabitant ofditches,streamsandpeat pits;MollerPillot 1984), Endochironomus tendens, Polypedilum gr. sordens, and Chaoborus flavicans (allthreetaxaalso typifygroup P3). Neumanialimosa isremoved fromthelistof typifying taxa of pondsitegroup6compared togroup P6. SitegroupP6represents clear, well oxygenated, /3-mesosapobic, meso- to eutrophic waters (peat pits)witharichvegetationona minerotrophicpeatbottom. 10.2.4DCCA -RUN4 Theresultsofordinationof sites and environmental variables of DCCA-run four (inwhichthesitegroupsdiscussed forruns1,2and3 wereremoved from theanalysis)are illustrated inFigure 10.4. The important environmental variables which describe the ordination significantly aregiveninTable 10.3. Thefirstand most important environmental gradient runs alongthehorizontal axisand isdueto twosetsofsitegroups. Onthe right side of the figure, three groupsoccur (R3,R8andRll). Mostsites inthesegroupshaveahigh electrical conductivity,andhighconcentrations of total phosphate, nitrate and major ions. Theammoniumconcentrations arelow. The sitesarewide,deepandalmostfreeofvegetation. Theyhaveasandy bottom with a silty substrate. The sitesweremainlysampledin spring,thewater isslightlyturbid to turbid, and there is some current. Ontheleftsideofthefigure,thegroupsP8andR2occur. Incomparisonwiththegroupsalreadydiscussedabove,the electrical conductivity and the concentration ofphosphate arelower,andthe bicarbonate concentration andthecoverpercentage of vegetation is higher (about 50%), especially thefloatingpart (27%). Group P7, above themiddle in Figure 10.4, differs in biotic and abiotic characteristics frombothabove-mentioned setofgroups.

183

Figure 10.4 Run 4; ordination (DCCA) diagram for axes 1 (horizontally) and 2 (vertically). The selected site groups of the reduced data set are shown. Within this contour line the site groups not indicated by a centroid and a confidence ellipse are indicated by their main physico-geomorphological water type; R= rivers and canals, P = ponds and lakes. Only environmental variables with an interset correlation greater than 0.35 are shown (arrows). For further explanation see Sections 3.4.2 and 10.2.4, and Figure 3.9. Abbreviations are explained in Table 10.5. 184

SitegroupRil SitegroupRllcompletelyreplaces the river site group 11. The composition of sites of groupRllisslightlyheterogeneous (Table 10.2). Theenvironmentaldescriptionofriver sitegroup 11 is also valid for group Rll. The new typifying taxa are Potamothrix moldaviensis (seegroupR5)andPiscicolageometra (an inhabitant of clean, fairly large tolargewaterswithasparsevegetationdueto itsrelation tofish;Garms 1961). SitegroupRllrepresentsß-meso- toa-mesosaprobic,a-meso-ionic, mesotrophic, large, linear, slightly flowing rivers or stagnant waters. SitegroupR8 SitegroupR8completelyreplacesriversitegroup8withinthe whole data setandtheenvironmentalcharacterizationofriver sitegroup8 isstill applicable. The composition of sites of group R8 is homogeneous. The taxa Nais elinguisandLimnophyessp.,typifying riversitegroup8,areremovedwhilesomewidespread taxainstagnant waters, like Hygrobates nigromaculatus, Midea orbiculata, Sigara falleni,Psammoryctidesbarbatus (seegroup Rll), Oecetis lacustris (an inhabitant ofsolidbottoms insmall,eutrophic lakes,mouthsof slowrivers,andponds;Lepneva 1970), andPolypedilumgr. bicrenatum (seegroupR9)areaddedtothelistoftypifying taxaofgroup R8. SitegroupR8represents/3-mesosaprobic,a-meso-ionic,verylarge, roundto irregularly shapedlakes. SitegroupR3 SitegroupR3resemblesriversitegroup 3 in site composition and description. Theenvironmentaldescriptionofthelatteralsoapplies togroupR3. Allmajor ionconcentrations are relatively low. The samples were mostly takeninthevegetationofthelittoralzoneor from thebaresandybottom. Severaltaxaareremoved fromthelistof typifying taxa of group R3 compared to river sitegroup3(e.g. Zavrelimyia sp.,Nemouracinerea,Nais elinguis, Conchapelopia sp., Uncinaisuncinata,andRheotanytarsus sp.). Thenewtypifying taxaof group R3 are Cladotanytarsus sp., Cryptochironomus sp., Oecetis lacustris, andPotamothrixmoldaviensis (fordetailsofallfourtaxa seegroupR8),Anodontaanatina (aninhabitantofflowingwatersanda sandy bottom, Ellis 1978),Naisbarbata (aninhabitantofmiddleand lowerreachesofstreams,Learneretal. 1978), Unio pictorum (an inhabitant ofslowrivers,canals,lakesandlargeponds;Ellis 1978) andThienemaniella sp. (aninhabitantoflotiehabitats,Cranston et al. 1983). SitegroupR3represents aamesosaprobic, medium-sized, slightly meandering,slowly flowingsmallrivers. SitegroupP7 SitegroupP7isalmostcompletely different insite composition and description compared with pond site group 7 (Table 10.1). The important environmental variables which describe the ordination

185

significantly are given in Table10.3. Mostsiteswere sampled in summer (71%). Theelectricalconductivity andthenutrientand major ion concentrations are relatively low. Thesitesaredeepandthe surface area is fairly large. There is a moderate vegetation development,thewater isclear (86%),shaded (86%),andsurroundedby woodsorwoodedbanks (86%). Theprofile shape is often irregular (86%), the sitesareisolated (86%)andthebottomconsists ofsand. Thewaterlevelcanfluctuate (57%). Thetypifyingtaxaof group P7 are Psectrocladius sp. (aneurytopic taxon,Cranstonetal. 1983), Ablabesmyiamonilis/phatta (aninhabitantofstagnantwaters, Fittkau 1962), Cladotanytarsus sp. (common, Pinder &Reiss 1983), Cyrnus flavidus (aninhabitantofstillwaters, Edington & Hildrew 1981), Erythromma najas (aninhabitantofvegetation-rich,stagnantwaters; Gardner 1954),Mideopsis orbicularis (common), Cryptochironomus sp. (common,MollerPillot 1984), Caenis luctuosa (aninhabitantoflarger streams,rivers,andmeso- to slightly eutrophic, stagnant waters; Malzacher 1986), Centroptilumluteolum (ataxonfoundonstonyshores oflakesand in slowly flowing sections of streams and rivers, especially amongstvegetationandonsandybottoms;Elliott&Humpesh 1983), Cloeonsimile (aninhabitant of slowly flowing sections of streamsandrivers,smallpondsandamongstvegetation indeeperparts ofpondsandlakes (Elliott & Humpesh 1983), Demicryptochironomus vulneratus (found in lakes and rivers, Lehmann 1971), Gerris argentatus (common,especially invegetation;Nieser 1982), Hydrodroma despiciens (aninhabitantoflakesonasandysubstrate,Davids 1981), Pseudochironomus sp.,Stictochironomussp.,andTribelos intextus (all three chironomids preferring sandy bottoms in larger, oligo- to mesotrophicwaterbodiesorthelittoralzonewithwaveaction;Pinder & Reiss 1983, Buskens 1987),Molannaangustata,Mesoveliafurcata, Mystacides sp.,Ranatra linearis (commoninstagnant, well vegetated waters; Bernhardt 1985), andAblabesmyia longistyla (aninhabitantof welloxygenated,vegetatedwaters;Fittkau 1962). SitegroupP7representsß-mesosaprobic,clear, well oxygenated, meso- to eutrophic, medium-sized, deep stagnant waters rich in vegetation. SitegroupR2 SitegroupR2isacombinationofriversitegroup 2 and ditch site group 1. TheaverageresemblanceofsitesingroupR2isreasonably high. The sitesaremedium-sized,havealinearshape (90%) with a steep profile (83%) andaresituatedonaminerotrophic peatbottom (60%). Thesitesaresurroundedbyreed- (54%) or grassland (52%). They weresampled inspring (67%)orsummer (33%). Theenvironmental descriptiongivenforriversitegroup 2also applies to group R2. Thenew typifying taxaofgroupR2are,amongothers,Limnesiaconnata (aninhabitantofstagnantwatersbetweenvegetation,Besseling 1964), Tiphys ornatus (a species occurring in spring,Davids 1979), and Oecetis furva (aninhabitantofvegetatedponds,Tobias&Tobias 1981) compared to river sitegroup2. Onlyafewtypifyingtaxaofditch sitegroup 1arealso typifying of group R2 (e.g. Endochironomus albipennis,Pionaalpicola/coccinea). Argyronetaaquatica,Segmentina nitida,Bathyomphalus contortus,andHippeutiscomplanataare removed fromthelistoftypifying taxaofriver sitegroup 2withinthewhole

186

dataset. SitegroupR2represents/3-meso-toa-mesosaprobic, large ditches andsmallcanalsonaminerotrophicpeatbottom. SitegroupP8 Group P8ismainlycomposedofsitesformerlybelonging to pond site group 8 and ditch sitegroup2B. Theenvironmental descriptionof boththelattersitegroupsdoesnotfitwiththatofgroup P8. The sites of group P8 were mainly sampled insummer (75%). Theyare sometimes shaded (46%)andsituated ingrassland (64%). Thesitesare smaller and shallower than the sitesofgroup P7. Theprofileis irregular (54%)andthesubstrateconsistsofsilt (61%)deposited on all kinds of bottom. Themajor ionandnutrientconcentrationsare comparable togroupR2. Almost 50%ofthis group is hydrologically isolated compared to6% ingroupR2. Thenewtypifying taxaofgroup P8are, among others, Holocentropus dubius, Cymatia coleoptrata, Gerris odontogaster (for details of allthree taxaseegroupP2), Hippeutiscomplanata (common, particularly in closed ponds; Macan 1977), Notonecta glauca (common, Freude etal. 1971), Erythromma najas (seegroup P7),Anatopyniaplumipes (an inhabitant of muddy, vegetated littoralzoneoflakes;Fittkau 1962),Agrypniapagetana (an inhabitant of lakes, Lepneva 1971), Theromyzon tessulatum (an inhabitantofshallow,stagnantandslowly flowingwaters;Dresscher & Higler 1982), Sialislutaria (an inhabitant of sluggish parts of streams and rivers, Elliott 1977), Paramerinacingulata (foundin lakesandslowly flowingrivers,Fittkau 1962), Tanypus villipennis (aninhabitantofstagnantwaters,MollerPillot 1984), andTriaenodes bicolor (aninhabitantofvegetation-richponds,Tobias&Tobias 1981) compared to pond site group 8. However,Bathyomphaluscontortus, Planorbarius corneus,Bithynialeachi, Physa fontinalis, Stagnicola palustris,Anisusvortex,andPionaalpicola/cocclneaprovednottobe typifyingofgroupP8withinthewholedataset. SitegroupP8represents/8-meso-toa-mesosaprobic, medium-sized, stagnantshallowwaters. 10.2.5DCCA -RUN5 Theresults ofordinationofsitesandenvironmentalvariablesof the fifth DCCA-run (in whichthesitegroupsdiscussedforruns1,2,3 and4wereremoved fromtheanalysis)areillustrated inFigure 10.5. There is noclearenvironmental gradientpresent (seesection10.2). Ingeneral,sites of all groups in run five have high average temperatures,afairlyhighelectricalconductivity (390-501 nS), high concentrations ofsodiumand chloride, are well oxygenated (10-13 mg/1), oligo-y8-mesosaprobic,andeutrophic tohypertrophic. Allsites arewideanddeep (176-358cm)orhaveafairly large surface area. The water ismostlyyellowandslightlyturbid. Thebottomconsists mainlyofsand (42-78%)orclay (10-38%)withasilt layer (42-75%). Thesitesareoftensurroundedbygrassland (50-96%). SitegroupP9

187

Figure 10.5 Run 5; ordination (DCCA) diagram for axes 1 (horizontally) and 2 (vertically). The selected site groups of the reduced data set are shown. Only environmental variables with an interset correlation greater than 0.35 are shown (arrows). For further explanation see Sections 3.4.2 and 10.2.5, and Figure 3.9. Abbreviations are explained in Table 10.5. 188

Sitegroup P9 isstrongly similar topondgroup 9insite composition and description. The environmental characterizationofthelatter fitsgroupP9reasonablywell. Thecompositionofsitesof group P9 ishomogeneous (Table 10.2). Thesurfaceareaofmostsites islarge. Someofthesitesareisolated (46%), somehavean irregular profile (67%) and the water is sometimes green (42%). Thebicarbonate concentration ishighandthenitrateconcentration low. The sites are sometimes situated inreedland (42%). Compared topondgroup9 thenewtypifying taxaofgroupP9are,amongothers,Tanypuskraatzi, Theromyzon tessulatum (both taxa see group P8), Dero digitata, Polypedilumgr. sordens,Endochironomus albipennis (both the latter chironomids inhabit tubes invegetation inpondsandlakes,Shilova 1968), Parachironomus gr. arcuatus (an inhabitant of lakes and rivers, Moller Pillot 1984), andPhryganeabipunctata (aninhabitant ofvegetation-rich,stagnantwaters;Tobias&Tobias 1981). Allthree species of thegenusLimnodrilusandValvatapiscinaliswere removed fromthelistoftypifying taxaofgroup P9. SitegroupP9representsa-mesosaprobic, fairly large ponds or smalllakes. SitegroupR4 SitegroupR4iscompiled fromsomeofthesitesofriver sitegroup4 and a number of other sites from differentgroups. Despitethe changedsitegroupcompositioncomparedtoriversitegroup 4, group R4 is a reasonablyhomogeneous group. Thesiteshavealowcurrent velocity,arerelatively shallow (withinthisDCCA-run),havealinear shape (88%), and aresurroundedbygrassland (96%). Thesiteswere mainly sampled inspring (47%) and summer (41%), and the bottom consistsofsandwithasiltlayer (67%). Theelectrical conductivity ishigh. Compared toriver sitegroup4the new typifying taxa of group R4 are, amongothers,Cryptochironomus sp. (aninhabitantof lakes,streamsandr.ivers;Pinder&Reiss 1983), Laccophilus hyalinus (common, Freudeetal. 1971), Mystacides sp.,Eylaisextendens (both inhabitants of vegetation rich, stagnant waters; Hickin 1967, Besseling1964), Oulimniustuberculatus (foundinunpolluted, stagnant waterswithwaveactionorstreams;Cuppen 1984), Parachironomus gr. vittiosus (an inhabitant oflakesandrivers,Reiss 1968), Bithynia leachi (common, Macan 1977) and several common mites. Haliplus ruficollis proved not to be typifyingofgroupR4within thewhole dataset. SitegroupR4representsa-meso-ionic,jS-meso-to a-mesosaprobic, linearshapedsmalltomedium-sizedwaters. SitegroupPll Sitegroup Pllisacombinationofsitesformerly belonging to pond site group 7 and river site group 4. Besides the general environmental descriptionapplicablewithinthisDCCA-run, group Pll itself has onlyafewenvironmentalcharacteristics. Mostsitesare isolated (67%)andshaded (75%). Thesamples were mainly taken in summer (92%). The average nitrate concentration is low. The typifying taxaofgroupPllare, among others, Gerris sp., Gerris odontogaster (an inhabitant oflakes,swamps,andtemporarywaters;

189

Bernhardt 1985),Arrenurusglobator,Cyrnusflavidus (aninhabitantof stagnant or slowly flowingwaters,Lepneva 1971), Piscicolageometra (aninhabitantofclean, larger, less weedy waters; Garms 1961), Unionicolaaculeata,Dicrotendipesgr. nervosus (collectedfromlakes andrivers,Reiss 1968), Cryptochironomus sp. (aninhabitantoflakes and rivers, Pinder &Reiss 1983),Nanocladius sp. (commoninlakes andrivers,Cranstonetal. 1983),Oecetislacustris (an inhabitant of ponds and small eutrophic lakes,Lepneva 1970), Polypedilumgr. sordensandEndochironomus albipennis (bothchironomidslivingon the vegetation instagnant,eutrophicwaters;MollerPillot 1984), several widespread taxa (e.g. Pionaconglobata,Hydrachna globosa, Limnesia sp., Laccophilus sp., Enochrus sp.,andHaliplus immaculatus),and several taxaalsotypifyinggroup P7 (e.g. Hydrodroma despiciens, Mesovelia furcata, Mystacides sp., Ranatra linearis, Ilyocoris cimicoides,andAblabesmyia longistyla). Site group Pll represents y8-mesosaprobic, medium-sized, deep stagnantwaters. SitegroupR12 SitegroupR12isacombinationofsitesformerlybelonging to river sitegroups1,4and 12,andpondsitegroups 7and9. GroupR12 isa fairlyhomogeneous group. Again, beside the general environmental characterization of allsiteswithinthisDCCA-run,groupR12 itself hasfewcharacteristics. Thephosphate concentrations are relatively low andthebottom issilty. Thesamplesweremainly takeninspring (77%). Onlyathirdofthesitesisisolated fromotherwaterbodies. Thesitesareslightlydeeperandlargerthanthoseofgroup Pll. The typifyingtaxaofgroupR12are Cladotanytarsus sp. (an eurytopic taxon, Pinder & Reiss 1983), Cryptocladopelma gr. laccophila (an inhabitantofwelloxygenated lakes,Lenz 1962), Oecetis furva (an inhabitant of vegetated ponds, Tobias & Tobias 1981), Ophidonais serpentina (aninhabitantofstagnant,vegetated waters; Verdonschot 1984), andseveral taxaalsotypifyinggroupR3 (e.g. Pionapusilla, Corynoneurasp., Caenls horaria, Ecnomus tenellus, Nais barbata, Psammoryctides barbatus,andParachironomus gr. arcuatis),andgroup Pll (e.g. Piscicola geometra,Dicrotendipes gr. nervosa, Mystacides sp., Polypedilum gr. sordens,Unionicolacrassipes,Endochironomus albipennis,Glyptotendipes sp.,andAblabesmyia longistyla). SitegroupR12represents ß-meso- to a-mesosaprobic, meso- to eutrophic,large,lessdeepstagnantwaters. SitegroupRl SitegroupRl iscomparable,insitecompositionand description, to river site group 1. The site group composition is reasonably homogeneous. Theenvironmentaldescriptionofriver sitegroup 1fits group Rl. In addition, thecalciumandnitrateconcentrationsare high,thesitesarelinear (94%),theprofile is often consolidated (56%) and there is somecurrent. Thesiteswere sampled insummer (63%). Compared toriversitegroup1thenewtypifyingtaxaofgroup Rl are Gerris sp.,Potamothrixmoldaviensis (aninhabitantoflower reaches of rivers, Timm 1970), Glyptotendipes gr. caulicola, Branchiura sowerbyi (an inhabitant of silt in stagnant waters,

190

Chekanovskaya 1962), Limnodrilusclaparedlanus (asaprobic inhabitant of lakes and rivers in mud andmuddy sand,Dzwillo 1966), Cyrnus trimaculatus (an inhabitant of lower reaches of large rivers, Eddington & Hildrew 1981), taxa also typifying group Rll(e.g. Dreissenapolymorpha,Dicrotendipes gr. nervosa, and Psammoryctides barbatus), and group Pll (e.g. Endochironomus albipennis and Glyptotendipessp.). SitegroupRlrepresents/3-meso-to a-mesosaprobic, medium-sized tolargeveryslowly flowinglowercoursesofstreamsandrivers. 10.2.6Theenvironmental characterizationofthesitegroups Intheenvironmentalcharacterizationofthesitegroups,thephysical and chemical variables were weighed against each other but an evaluationisonlypossiblebycomparingthemtootherdata. Thenutrientcontent is mostly expressed as the load and/or concentration of phosphate and nitrogen. Several trophic classificationsbasedonnutrient concentrations were made but it shouldbeborne inmind thattherelationbetweennutrientbalanceand productivity isnotastraightforward one (Elster 1962). For the present study only concentrationswereavailable. Leentvaar (1979) introduced a trophic classification based on orthophosphate and nitrate concentrations. Within thatclassification, thegroups PI, P2,P3andPllareclassifiedasmesotrophic,groupP6,P7andR12are meso- to eutrophic,andgroupP5iseutrophic. Severalothergroups alsohaveeitheraloworthophosphate oralow nitrate concentration indicating meso- or eutrophy,buttheothervariable thenindicates hypertrophy. This trophic indication isgiveninthecharacterization of the site groups. One should consider that the nutrient concentrations arestronglyrelated to the season, especially the orthophosphate concentration. Vollenweider (1968)publishedatrophic classification based on total phosphate concentration. In his classification most of the presented site groups prove to be polytrophic,exceptforgroupsH4, H5,H6, PI,P6 and P7 which are eutrophic to polytrophic. Infact,theinorganicnitrogencompound shouldalsobeconsidered (Vollenweider 1968). Looking atthenitrate concentration incombinationwithtotalphosphateconcentration,only groupsPIandP7provenottobepolytrophic. It can be concluded that therearealmostnotrulyoligotrophicwaters intheprovinceof Overijsselnowadays. Formost of the sites investigated, neither phosphatenornitrateprobably limitsprimaryproduction. Thenitrate concentrationswereevenexcessivelyhigh in smaller waters in the Pleistocenepartoftheprovince (springs,streamsandditches). The indicationofsaprobity,usedforthecharacterizationof the site group, isbasedontheammoniumconcentration (Wegl1983). The highsaprobity ofacidwatershasalreadybeen discussed in Section 9.3. The totalionconcentration isclassifiedbyOlsen (1950). Most of our groups are ,8-meso-ionic, exceptforthose indicated inthe characterizationofthesitegroups. According toOlsen (1950), all the groupshavesofttomedium-hardwater,exceptforgroupR12which hashardwater. Thealkalinity isslightlybelow or above average, exceptforgroupsS12,PI, P2, P3,Rl,andR3 inwhich itislow. Thephysicalvariables (such as width, depth and shape) have

191

alreadybeendiscussed inthesitegroupdescriptions. The indication ofthephysico-geographicalwater type (Table1.2) is related to a combination ofimportantphysicalvariablesandisused inthatsense forthecharacterizationofthesitegroups. Notethatnotallwaters which correspond to aphysico-geographicalwater typebelong tothe same site group. On the contrary, the indication of a physico-geographical type isonlyoneofthecomponents contributing tothecharacterizationofasitegroup. The typifying taxaofeachsitegroupwithinthe whole data set change when compared withtheanalyseswithinoneofthemainwater types. Thefollowingchangesoccurred: -Formertypifying taxabecame indifferentortheir typifying weight decreased. Thisconcernstaxawhichweretypifyingacertaingroup withinamainwater typebutforwhich theenvironmentofthisgroup is less than optimal. Their optimal environment ispresentin another (main)water type. For example, taxa common in streams (e.g. Micropsectra sp.,Conchapelopia sp.,andBaetissp.)typified riversitegroup9withinthemainwatertype "rivers and canals" but theiroccurrence inthesesites islimitedandsuboptimal. Due totheirwidedistributionrangewithinstreams they were removed from the list of typifying taxa forgroup R9. Anotherexample concerns taxawhichare widely distributed but locally abundant (e.g. Chironomussp. andTubificidae). Theirtypifyingcharacter inseveral groupswithinthemainwater types iscoincidental. -Former indifferent taxabecame typifyingortheir typifying weight increased. Thisconcerns taxawhichwerewidelydistributedwithin amainwater typeandwere lessabundantor lacking in all other main water types. Forexample,severaltaxatypically inhabiting smallstreams (e.g. Dlcranotabimaculata,Prodiamesaolivacea, and Brilliamodesta)arecommonwithinthestream sitegroupsandtypify themwithinthewholedataset.

10.3 The biological

similarity

between the site

groups

Ahierarchical dendrogram ofsitegroupsbasedonthetotal data set reflects the biological similaritybetweenthegroups (Figure 10.6). Thisdendrogram isbasedonagglomerative clustering of site groups. Each site group is compared to allothersanditsresemblance (a measure ofdistancebetween thecentroidsoftwogroups)to the most similar site group iscalculated (centroidclustering;vanTongeren 1986). The twomostsimilargroups (i.e. those with the highest resemblance) are fusedandanewcentroid iscalculated for thisnew group. Again,anewfusionisexecuted andthisprocess is repeated until all the groups are fused. InFigure 10.6theresemblances between (combinations)ofgroupswereplottedon the vertical axis. The site groups areplottedonthehorizontalaxis. ThegroupsRl, R4, Pll,R12,R2andR8arequitesimilar to each other while they differ markedly from groups H5 and H6,whicharealso completely different fromeachother (Figure 10.6). Inthisway the similarity between all the sitegroupscanbeseen. Inthedendrogram,atthe differentdivisions,themost important variables or complexes of variablesrelated tothatdivisionare indicated. The twomostaberrant sitegroupsareS14andDil. Theyareequal tostream sitegroup 14andditchsitegroup 11respectively. They

192

a - •— ri^ij-r--'—o o i r » ( N ( v n o m p o r > J , ~ » — = » " ' " ' v ) ( j i ( N t f ) f f i ' - W r - i n ^ m N m i - r - M r - I D

in

i

,

2£sS

J*Is?J E

( 0 ( 0 >AE (A

ÏSîJJ

L. a n j u v> a)

•O U

Oc Si« "O

u » j D— y *•= ™ O r u » [

-S.5-

•S3"»«»! £ O , L. , £ u)

Xl

auaig>:v

.y?

•ils

.era 0.-~i

»saa,£ r-g-30101 _ o a_ «~ 3 Ea ? = am O f ID ^ *

' t IT! « - l - 1

tftQ.

fee UI « vraj

5 u

t

T

>H

:s =

Figure10.6 Hierarchical potentially responsible

a: