Alowcostthreedimensionalsamplecollectionarraytoeva

ecologicalengineering33(2008)83–90availableat∗,F.J.AdamsenUSDA-ARS,USAridLandAgriculturalResearchCenter,21881NCardonLn.,Maricopa,AZ85238,UnitedStatesarticleinfoArticlehistory:Received2November2007Receivedinrevisedform7February2008Accepted17February2008Keywords:WetlandSamplingMonitoringabstractArtificiallyconstructedwetlandsaregainingacceptanceasalow-costtreatmentalterna-tivetoremoveanumberofundesirableconstituentsfromwater.Wetlandscanbeusedtophysicallyremovecompoundssuchassuspendedsolidsthroughsedimentation.Dissolvednutrients,biochemicaloxygendemand,heavymetals,andpotentiallyharmfulanthro-pogeniccompoundscanallberemovedinconstructedwetlandsthroughgeochemicalandbiologicalprocesses.Samplecollectiontoevaluatetheeffectivenessoftreatmentandtomonitorthestatusofwetlandsisusuallyonlyconductedattheinletandoutletofthewetlandduetocostconstraints.Tobetterunderstandtheinternalhydrologyandbiogeo-chemicalprocessesoperatingwithinthewetlandmoreintensivesamplingisneededthatdoesnotinterferewiththehydraulicsofthesystem.Anewrelativelylow-costsamplecollec-tiondesignhasbeendevelopedusingmostlyoff-the-shelfpartsthatallowsforpermanent,internal,three-dimensionalsamplecollectioninwetlands.Thedesignhasbeenusedtoconstructapermanentthree-dimensionalarrayof60-samplelocationsthatcanbesam-pledsimultaneouslythroughouta1.2haconstructedwetlandforlessthanUS$5000.Thesamplingarraywasusedinatracerstudyandshowedspatialandtemporaldifferencesintracerconcentrationwithinthewetland.Concentrationdifferenceswereseenandmea-suredinallthreedimensions.Thebasicfeaturesofthesystemaredescribedandanexamplehowtoconstructanarraythatcansuitanywetlanddesignisgiven.PublishedbyElsevierB.V.1.IntroductionWetlandsofferabiologicallydiverseecosystemcapableofremovingmanycontaminantsfoundinwater.Artificiallyconstructedwetlandsarealow-costtreatmentalternativeforremovalofanumberofundesirableconstituentsfoundinwastestreamsfromindustry,mining,agriculture,andurbanareas(Augustetal.,2002;MoustafaandHamrick,2000;Hambrightetal.,1998;GuardoandTomasello,1995).Wet-landscanbeusedformechanicalremovalofsuspendedsolidsthroughsedimentation(Schmidetal.,2004).Dis-∗Correspondingauthor.Tel.:[phone];fax:[phone].E-mailaddress:[e-mail](C.F.Williams).Abbreviations:3-D,threedimensional.solvednutrients,biochemicaloxygendemand,heavymetals,andpotentiallyharmfulanthropogeniccompoundscanberemovedinconstructedwetlandsthroughgeochemicalandbiologicalprocesses(Augustetal.,2002;Drizoetal.,2000;GuardoandTomasello,1995;Hambrightetal.,1998;MoustafaandHamrick,2000).Wetlandscanalsobeusedasacost-effectivefinalpolishingstepformunicipalsewageeffluenttoinsuredischargedwatermeetsstringentwildlifeprotectionstandards.AccordingtoBuchbergerandShaw(1995),thetwomostimportantprocessesindeterminingtheeffectivenessofcon-0925-8574/$–seefrontmatter.PublishedbyElsevierB.V.doi:10.1016/j.ecoleng.2008.02.00684ecologicalengineering33(2008)83–90structedwetlandsforwatertreatmentarebiologicalandhydrologicaldynamics.Biologicaldynamicsaredefinedasthebiogeochemicalprocessesthatremovecontaminantsfromwaterbeingtreatedinawetland.Theseprocessesareinturnlinkedtemporallyandspatiallythroughoutthewetlandviathehydrologyofthesystem.Thehydrologyofthewetlanddeterminesthedistributionandretentiontimeofnutrientsonwhichthebiotadepend,whilenutrientconcentrationsareinfluencedbythemetabolicprocessesofthebiota.Inaddition,thepresenceofmacrophytescanbecriticalindeterminingtheflowdynamicsofthewetlandduetosmall-scaleinter-ruptionstoflowpathscreatedbythingssuchasplantstems,whichresultsinanincreaseinsurfaceroughness(Schmidetal.,2004).Inthepast,constructedwetlandsweredesignedasaseriesofplugflowreactorsconnectingcontinuouslystirredreactors.Thevegetatedareaswereseenasplugflowreactorswhereflowislaminarandwellmixedintheplaneperpendiculartotheflow(Keefeetal.,2004).Openwaterareaswereviewedascontinuouslystirredreactorswherethewateriswellmixedinallthreedimensions.Usingthisconceptualframeworktheaspectratio(thelengthofthewetlandparalleltoflowdividedbythewidthperpendiculartoflow)becomescritical.Itwasgenerallythoughtthatthehighertheaspectratiothemoreideallythewetlandbehaved,withtheflowinthelongitudinaldirectionbeinglaminarandthevolumeofwaterperpendic-ulartotheflowbeingwellmixed.Morerecentlyithasbeenfoundthattheassumptionsofwellmixedplugflowarefaulty.Kadlec(2000)reportedthat,inanumberofwetlandtracerstudies,flowthroughconstructedwetlandswasnon-idealandshort-circuitedflowpathswerepresent.Short-circuitingwasevenfoundinlab-scalewetlandsassmallas1mlong×0.5mwide×0.5mdeep(Drizoetal.,2000).Non-idealflowthroughconstructedwetlandsismostoftenquantifiedusingtracerexperiments.Keefeetal.(2004)usedaconservativetracer,bromide,andareactivetracer,rhodamineWT,tocharacterizethreedifferentconstructedwetlandswithvariousseepagelosses.Theyfoundthatthevolumeofthewetlandthroughwhichthetracerflowedrangedfrom41to84%ofthetotalwetlandvolu