SubmittedtoAlianzaCeiboByJenaWebb,PhDDate:December19,2018

Technicalreport:ClearWaterDomesticrainwaterharvestingsystemsevaluation

“Youhaveto really takecareofwater,careforitmorethangold.”- CommunitymemberfromSanVictorian

AbouttheAuthor 5ParticipantsintheEvaluation 5SUMMARYOFEVALUATION 6SummaryofResults 6

INTRODUCTION 7QualityofRain-harvestedWater 9TheCeiboAllianceandClearWater’sWaterProgram 11

AIMSANDOBJECTIVESOFTHEEVALUATION 12Aim 12Objectives 12

METHODOLOGY 12ChoiceofIndicators 12SamplingandQuestionnaires 14PilotProject 14TotalandFecalColiformAnalysisWithaPortableLab 14Questionnaire 14In-DepthInterviews 15EthicsApproval 15UseofData 15

RESULTSANDDISCUSSION 15UseoftheSystems 16MaintenanceoftheSystems 17StateoftheSystems 17Health 18ProgramAcceptability 18Sampleanalysisofdomesticrainwaterharvestingsystemsandalternatesourcesofwater 19FecalandTotalColiform 19Hydrocarbons 21Metals 21Mercuryandlead 21

Zinc 22QualitativeAnalysis 22Health 22TheRainwaterHarvestingSystem 23SecurityandEaseoftheSystem 24OtherBenefitsoftheProgram 25

CONCLUSIONS 26APPENDIX1:RECOMMENDATIONS 27

SecondWaveofMaintenanceWorkshops 27MaintenanceGuide 27NeedsAnalysis,Sampling,InterviewsandTraininginWaoraniTerritory 27TrainingandProperInstallationofSystemsinTarabiaya 27WomenCommunityWaterTechnicianTraining 28AcquisitionofaPortableLaboratoryandRegularEvaluationofSystems 28InstallaFirst-FlushSystem 28SystemDatabase 29NewRoundofRainwaterHarvestingSystemsinCommunitiesAlreadyServed 29OverflowMechanisms 29StagnantWaterControl 30

APPENDIX2:WATERSAMPLING 31APPENDIX3:QUESTIONNAIRE 32APPENDIX4:INTERVIEWGUIDE 34APPENDIX5:SUMMARYOFRESULTSFROMTHEEXTERNALOBSERVERS(NSANDCMC) 36APPENDIX6:RECOMMENDATIONSOFCALEBMATOSCHÁVEZ,RAINFORESTFLOW 37

AbouttheAuthorDr.JenaWebbhasbeenworkingwithindigenouscommunitiesintheEcuadorianAmazonsince2001 to better understand and respond to contamination in water and one of the mostimportant food sources in the region - fish. She earned a doctorate fromMcGill Universityconducting a studyonmercury andhydrocarbonpollution in threewatershedsof theUpperAmazon. Dr. Webb is committed to conducting and sharing research in a way that is bothrigorousandbeneficialtoindigenouspeople.

ParticipantsintheEvaluationNicolasMainville,Monitoringteam,AmazonFrontlinesDarwinYiyoguaje,Sionateam,AlianzaCeiboAliciaSalazar,Sionateam,AlianzaCeiboFlorTangoy,Sionateam,AlianzaCeiboArmandoPiaguaje,Sionateam,AlianzaCeiboHurlemPayaguaje,Secoyateam,AlianzaCeiboDannesPayaguaje,Secoyateam,AlianzaCeiboGladysVargas,Kofanteam,AlianzaCeiboAlexLucitante,Kofanteam,AlianzaCeiboWilsonQuieta,Kofanteam,AlianzaCeiboEmergildoCriollo,Kofanteam,AlianzaCeibo

SUMMARYOFEVALUATIONThisreportpresentstheresultsofanevaluationofthejointCeiboAllianceandClearWater(aproject of the non-governmental organization, Amazon Frontlines) domestic rainwaterharvestingsystemprogramwhich,from2012to2018,provideddomesticrainwaterharvestingsystems to 1164 households in 78 communities of the Northeastern Ecuadorian Amazonbenefitingover6000people.Thestudyevaluatedthreegeneralaspectsoftheprogram1)thequalityofthewater,2)theacceptabilityoftheprojectand3)impactsofthewatersystemsonpeople’slivelihoodsandhealth.SummaryofResultsThe study evaluated water quality and user observations in a representative sample of thecommunitiesvisited.Sixty-twohouseholdheads(39menand23women)respondedtoashortquestionnaire on use, maintenance, health and satisfaction of the water and systems. Onehundred percent of the respondents were satisfied with the taste of the water and thefunctioning of their system. 86% of people reported observing improved digestive health(reduceddiarrhea,vomitingandstomachpain) since the installationof thesystems.Noneofthedrinkingwatersampleshaddetectablelevelsofpolycyclicaromatichydrocarbons(PAHs)orthe heavy metals, mercury and lead. Zinc levels in all samples were below a level whichoccasionally leads to unpleasant taste (4mg/l). 85% of systems had ideal water quality asindicatedbytheabsenceoffecalcoliforms,4%hadsatisfactorywaterqualityand11%showedsigns of fecal coliform levels higher than indicated by health advisories. Water from thealternativesources (rivers, streamsandsprings)had lowerwaterqualityas indicatedbyhighlevels of fecal coliform. People clean their systems on average 4 times per year. 70%of thewatersystemswereratedasinverygood(46%)orgood(24%)condition.The qualitative analysis showed a high level of satisfactionwith thewater and the systems.Comments pointed to improvements in health and workload, with the system reducing thenumber of hours and the effort required to collectwater. Small critiques anduser problem-solvingtechniqueswerecollectedandcanbeelaboratedoninsubsequentinformationsharingwithcommunities.Aseriesofrecommendationsarepresented.

INTRODUCTIONAlthough water abounds in the Ecuadorian Amazon, local communities face numerouschallenges to securing a safewater supply for their families. Over the past half-century, theEcuadorianAmazonhasbeencrisscrossedbymore than9500kmof roads–or1.5 timestheEarth’s radius– built to connect themore than3430 oilwells tomarket via pipeline. In theperiodfrom2005to2015alone,oilcompaniesreleasedover350,000barrelsofcrudeoilintothe rivers, streams and soils of the Ecuadorian Amazon, amounting to an average of 4000barrels of a toxic chemicalmix released every day into the environment1. Hydrocarbons areknowncarcinogens.Concurrently,Africanpalmplantationshave led to thedeforestationof anequivalentof 100000 soccer fields (according to GFW)2 and the use of a large quantity of pesticides andfertilizers, some of which end up in the local waterways. Some pesticides are consideredcarcinogenic.Accessroadsenticecoloniststosetupsmallfarms,resultinginthedeforestationof vast tractsof land.Amazonian soils contain largeamountsofnaturallyoccurringmercury,accumulated thereovercenturiesofnearbyvolcanicactivity3.Mercuryexists in thesoils inaharmless form, but once soils erode into waterways following deforestation, themercury isconverted into a toxic form, easily entering the aquatic food chain and humans via fishconsumption4. Illegal gold mining, booming in certain areas of the Amazon, can also be animportantsourceofmercuryintoaquaticecosystems.5Further, erosion provoked by deforestation and mining has led to changes in the waterchemistry of local streams and rivers6. Meanwhile, increasing population and inadequatesewage-water treatmenthas led toan increase inexposure to fecal coliforms.Fecal coliformingestionleadstostomachpain,diarrhea,and,inextremecases,death.

1MAEPRAS:SUBSISTEMADEINTELIGENCIADEESTADÍSTICASSOCIOAMBIENTALESDELASACTIVIDADESPRODUCTIVAS(SIESAP)http://pras.ambiente.gob.ec/web/sinari/siesap2http://www.ecuadorencifras.gob.ec//documentos/web-inec/Estadisticas_agropecuarias/espac/espac_2014-2015/2014/Presentacion%20de%20resultados%20ESPAC_2014.pdf3Mainville,N.,J.Webb,M.Lucotte,R.Davidson,O.Betancourt,E.CuevaandD.Mergler(2006)."DecreaseofsoilfertilityandreleaseofmercuryfollowingdeforestationintheAndeanAmazon,NapoRiverValley,Ecuador."ScienceoftheTotalEnvironment368:88-98.4Webb,J.,N.Mainville,D.Mergler,M.Lucotte,O.Betancourt,R.Davidson,E.CuevaandE.Quizhpe(2004)."Mercuryinfish-eatingcommunitiesoftheAndeanAmazon,NapoRiverValley,Ecuador."EcoHealth1(suppl.2):59-71.5Akagi,H.,O.Malm,Y.Kinjo,M.Harada,F.J.P.Branches,W.C.PfeifferandH.Kato(1995)."MethylmercurypollutionintheAmazon,Brazil."ScienceofTheTotalEnvironment175(2):85-95.6McClain,M.E.andR.J.Naiman(2008)."AndeanInfluencesontheBiogeochemistryandEcologyoftheAmazonRiver."BioScience58(4):325-338.

In2014,EcuadorwasamongthefivecountriesinLatinAmericawiththelowestdrinkingwatercoverage.7TheUnitedNationsnamedaccess tocleandrinkingwaterabasichumanright in2010. In that same year, 89million people across the globe dependedon rainwater as theirmainsupply.8Rainwater is largelyfreeof impurities9.Whenproperly installedandmanaged,Domestic RainwaterHarvesting (DRWH) Systems are considered inexpensive and low-tech;and vastly superior tomost alternatives, such as surfacewater,which contain high levels ofcontaminantsandfecalcoliforms.Ecuadorisoneofthefewcountriesinwhichbottledwaterisan importantsourceofdrinkingwater10.Buyingbottledwater,whether itbesingleuseor inlargerquantities,isnotanoptionformostofthecommunitiesintheAmazonbecauseitiscostprohibitiveandaccessis low.Bearinginmindthebenefitsofrainwaterharvesting,expandingcapacityindevelopingcountrieshasbeensetasanewWorldHealthOrganizationgoal(6.a).3AccordingtotheWorldHealthOrganization(WHO),thebiggestpreoccupationwithwaterforhuman use in rural communities is via contamination by coliforms from human or animalfeces11.Consumptionofthesebacteriaindrinkingwater,andespeciallythosefromthefamilyofEscherichiacoli,cancausediarrhea,bloodinthefeces,vomiting,kidneydamageandurinarytract infections. Fecal coliforms can enter rain-harvestedwater through the feces of animalsspendingtimeontherooforbirddroppingsontheroof.Ifthesystemisnotwellclosed,fecalcoliformscanalsooriginatefromanimalsthatobtainaccesstothesystem.

7WorldHealthOrganization(WHO)andtheUnitedNationsChildren’sFund(UNICEF)(2014)Progressondrinkingwaterandsanitation.2014UpdateWHO/UNICEFJointMonitoringProgrammeforWaterSupplyandSanitation.Geneva.8WorldHealthOrganization(WHO)andtheUnitedNationsChildren’sFund(UNICEF)(2012)Progressondrinkingwaterandsanitation.2012UpdateWHO/UNICEFJointMonitoringProgrammeforWaterSupplyandSanitation.Geneva.9WorldHealthOrganization.(2008)RainwaterHarvesting.Chapter6.11In:GuidelinesforDrinking-waterQuality,SecondAddendumtothe3rdEdition.Geneva.10WorldHealthOrganization(WHO)andtheUnitedNationsChildren’sFund(UNICEF)(2017)Progressondrinkingwaterandsanitation.2017UpdateWHO/UNICEFJointMonitoringProgrammeforWaterSupplyandSanitation.Geneva.11WHO.(1997)Guidelinesfordrinking-waterquality.SECONDEDITION.Vol3:Surveillanceandcontrolofcommunitysupplies.Geneva.http://www.who.int/water_sanitation_health/dwq/gdwqvol32ed.pdf

QualityofRain-harvestedWaterFour factors affect the quality of harvested rainwater: 1) the air; 2) the roof; 3) filtrationmethods;and4)thestoragetank.

Moststudieshavereportedlevelsbelowmaximumpermissibleconcentrationsfor metals, hydrocarbons and pesticides in rain-harvested water12. Surfacewater, and not rainwater, was found to be a source of mercury andhydrocarbonsamongstAmazonianwomen13,14.

Studies have shown that one of themost appropriate roofing materials forcapturing rainwater are zinc roofs, such as those commonly used in theIndigenousvillageswhereCeiboandAFwork,becausezinc,whilenotharmfultohumans,iseffectiveinreducingmicrobialactivity15.

Sand filters eliminate the vast majority of bacteria and particulate boundpollutants in rain-harvestedwater16.Microorganisms thatmayenter the firsttank from roofs and other sources are removed by as much as 81-100% byfilterssuchasthoseusedbytheCeibo/AFsystems17.

Ingeneral,storagepurifieswaterduetoasettlingoutofcontaminantsandthecreationofanutrientpoorenvironmentunfavorabletobacterialgrowth.Thiseffectisenhancedinlargetanks,suchastheonesinstalledbyCeiboandAF11.

VeryfewstudiesonthequalityofrainwaterhavebeenconductedintheAmazon.Somestudiesshowthatrainwatercontainsverylittlecontaminationatthemomentthatitprecipitatesoutof

12Abbasi,T.andS.A.Abbasi(2011)."Sourcesofpollutioninrooftoprainwaterharvestingsystemsandtheircontrol."CriticalReviewsinEnvironmentalScienceandTechnology41(23):2097-2167.13Webb,J.,O.T.Coomes,D.MerglerandN.Ross(2016)."MercuryConcentrationsinUrineofAmerindianPopulationsNearOilFieldsinthePeruvianandEcuadorianAmazon."EnvironmentalResearch151:344-350.14Webb,J.,O.T.Coomes,D.MerglerandN.A.Ross(2017)."Levelsof1-hydroxypyreneinurineofpeoplelivinginanoilproducingregionoftheAndeanAmazon(EcuadorandPeru)."InternationalArchivesofOccupationalandEnvironmentalHealth91(1):105-115.15Stewart,C.,N.D.Kim,D.M.JohnstonandM.Nayyerloo(2016)."HealthHazardsAssociatedwithConsumptionofRoof-CollectedRainwaterinUrbanAreasinEmergencySituations."InternationalJournalofEnvironmentalResearchandPublicHealth13(10):1012.16Helmreich,B.andH.Horn(2009)."Opportunitiesinrainwaterharvesting."Desalination248(1):118-124.17Peter-Varbanets,M.,C.Zurbrügg,C.SwartzandW.Pronk(2009)."Decentralizedsystemsforpotablewaterandthepotentialofmembranetechnology."WaterResearch43(2):245-265.

the atmosphere18, even in areas close to petroleum operations19. However, the practice offlaring off unwanted gases that escape during the extraction process could be a source ofhydrocarbonstorain-harvestedwater.Morerecentstudies,usingnewtechnology,showthatflares letoffmorepollution that calculatedusing traditional techniques20. Thisprocess couldalso be releasing heavymetals, such asmercury and lead, into the atmosphere,where theycouldadheretoraindroplets.Severalstudieshaveshownthatrainharvestedfromzincroofscanleadtohighlevelsofzincinwaterdestinedforhumanconsumption21.However,accordingto thesesamestudies, there isnonegativehealthconsequence fromconsumingzinc.Wherezinc can have an impact is in the taste of thewater and, therefore, the acceptability of thesystems.TheexpectedimpactoftheCeibo/AFprojectisanimprovedqualityoflifeandhealththrougha secure and accessible source of clean water. Improved health is expected through theeliminationorreduction indiseasesandillnessesassociatedwithsubstandarddrinkingwater,primarilydigestivetroubles,suchasdiarrhea.In2016,water,sanitationandhygiene(WSH)wasresponsible for 1.9%of the global burden of disease22. The latest data available for Ecuadorestimates the total WSH-related deaths at 88,200, or 3.9% of the deaths in 2004. Chronicdiarrhea, especially in children, can lead tomalnutrition. Treatments for stomach complaintscan sap scarceeconomic resources from rural familieswith limited revenue, forcing them towork longerhours,gowithoutotheressentialsorexploitnatural resources.Finally, theworkassociatedwith collectingwater from traditional sources is tiresome, lengthyand sometimesdangerous23, such as when heavy rains make climbing river banks slippery. This burden isdisproportionatelycarriedbywomen.Ithasalsobeenshownthatprogramssuchascommunityinstallationofrainwaterharvestingsystemsinpollutedareasaccompaniedwithinformationonhealthrisks,leadstogreaterawarenessaboutenvironmentalhealthingeneral,withassociatedimprovementsinhealthandwellbeing24.

18Bendixetal.(2004)GradientsoffogandraininatropicalmontanecloudforestofsouthernEcuadoranditschemicalcomposition.Proceedings3rdInt.Conf.onFog,FogCollectionandDew.19Lewisetal.(2012)Airborneconcentrationsofmetalsandtotaldustduringsolidcatalystloadingandunloadingoperationsatapetroleumrefinery.InternationalJournalofHygieneandEnvironmentalHealthVolume215,Issue5.Pages514-52120Conrad,B.M.andM.R.Johnson(2017)."FieldMeasurementsofBlackCarbonYieldsfromGasFlaring."EnvironmentalScienceandTechnology51(3):1893-1900.21Stewartetal.(2016)HealthHazardsAssociatedwithConsumptionofRoof-CollectedRainwaterinUrbanAreasinEmergencySituations.InternationalJournalofEnvironmentalResearchandPublicHealth.13,101222http://www.who.int/gho/phe/water_sanitation/burden_text/en/23Inter-agencyTaskForceonGenderandWater(GWTF)(2015).Gender,WaterandSanitation:APolicyBrief,UnitedNations.24Davis,L.F.,M.D.Ramirez-Andreotta,J.E.T.McLain,A.Kilungo,L.AbrellandS.Buxner(2018)."Increasingenvironmentalhealthliteracythroughcontextuallearningincommunitiesatrisk."InternationalJournalofEnvironmentalResearchandPublicHealth15(10).

TheCeiboAllianceandClearWater’sWaterProgramThemissionoftheCeiboAllianceandClearWater’sWaterProgramistoimprovethehealthstatusandthequalityoflifeofindigenouscommunitiesintheAmazonthroughareliableand safe water supply. Based on the suitability of rainwater harvesting in such a humidenvironment as the Amazon, the Ceibo Alliance and ClearWater (now a sub-project ofAmazon Frontlines) aimed to provide the indigenous people of this contaminated regionwith domestic rainwater harvesting systems. Between 2012 and 2018, more than 1164water systems were installed in 78 communities, providing more than 6000 people withcleanwatertodrinkandusedomestically.Inthesystems installedbyCeibo/AF,rainwater isharvestedfromrooftopgutter-spoutsonpeople’shomes,with the flowdiverted intoa first tankwherethewaterpasses throughaspecially-designedbiosandfilter.Thewaterpassesthroughfourlayersoffiltration:

1. abiologicallyactivesurfacelayer,2. afinesandlayer,3. alayerofcrushedquartz,andfinally,4. alayerofcoarsegravel.

The top hypogeal layer (called a “Schmutzdecke”), contains microorganisms that removebacteria,trapcontaminants,andbreakdownotherincomingorganicmaterial.Thenexttwolayersworktogethertocreateacomplexmazeofsandgrainsthatmicrobesgettrappedinanddie. Theyalso trap contaminants suchas toxicmetals andpetroleumpollution,whichstick to thesandas they flowby inaprocesscalledadsorption.Finally, the layerofgravelservesasasupporttothesandandquartz layerssonothingflushesoutofthetankasthecleanwaterflowsintoasecond,largeanti-bacterialstoragetank.

AIMSANDOBJECTIVESOFTHEEVALUATIONAimThe purpose of this evaluation is to provide a rigorous and systematic evaluation of thedomesticrainwaterharvestingsystemsinstalledbytheCeiboAllianceandClearWater.Theaimwastodesignanevaluationwhichwouldprovideinformationonthe:

- Qualityofthewater;- Efficiencyofthefilters;- Healthandqualityoflifebenefitsofthesystems;- Levelofusersatisfaction;- Useandmaintenanceofthesystems.

ObjectivesTheCeiboAllianceandClearWater’swaterprogramevaluationhasseveralgeneralobjectives:

1. Evaluate the water quality of rain-harvested water from systems installed by the Ceibo Alliance andClearWatersince2012;

2. Comparethequalityoffilteredwaterwiththeun-filteredwater(fromthefirsttank);3. Compare the quality of the water from the systems with the water from rivers, streams and springs

constitutingtheprevious/alternatesourceofwaterforthecommunities;4. Determineifuseofrainwaterhasledtohealthandqualityoflifeimprovements;5. Determinethelevelofusersatisfaction;and6. Evaluatethestateofthesystems.

METHODOLOGYThis mixed-method design combines indicators from quantitative data and insights fromqualitativedatainordertorespondtothegeneralobjectivesabove.Thespecificobjectivesoftheevaluationwerethefollowing:

1. Measurea)thelevelsoffecalcoliformsintheunfilteredandfilteredwaterofarepresentativesampleofsystemsandb)thelevelsofPAHs,mercury,leadandzincinasubsampleoffilteredandunfilteredwater;

2. Measurelevelsoffecalcoliforms,PAHs,mercury,leadandzincintheprevioussourcesofwater;3. Carry out a survey among a representative sample of users addressing quality of life, health and

maintenanceissues;4. Carry out in-depth interviewswith a subsample of users to delvemore deeply into issues surrounding

health,qualityoflifeandsustainability;and5. Inspectthestateofarepresentativesampleofsystems.

ChoiceofIndicatorsSincethewaterinthesystemsinstalledbytheCeiboAllianceandClearWatercomesfromruralareas,themainhealthconcernwithrain-harvestedwaterintheareaisthroughfecalcoliformsenteringthesystemsfromanimalfecesontheroofor inthepipingusedtocollectrain.Raw,un-treated water (from different sources) generally contains 100-100,000 E coli per liter.11

Water temperatures and nutrient conditions in the type of biofilm created by the watersystemsarenotconducivetotheEcolistrainofbacteriaand,therefore,thepresenceofE.coliisconsideredasevidenceofrecentfaecalcontamination.25This was, therefore, the primary focus of the present evaluation. Our first indicator for thismeasure was number of fecal coliform colonies in drinking water from the systems. Theobjective fordrinkingwater supplies is zero fecal coliform;however,anything less than10 isconsidered passable and the WHO recommends setting intermediate goals in underservedareas so as to not condemn water sources that are otherwise superior to alternatives. TheWorldHealthOrganizationmaintainsthat:

“Inmanydevelopingcountries,highqualitywatermeetingtheE.colicriterionisnotreadilyavailable,anduncriticalenforcementoftheguidelinemayleadtocondemnationofwatersourcesthatmaybemoreappropriateormore accessible thanother sources, andmayeven forcepeople toobtain their water from more polluted sources. Under conditions ofwidespread faecal contamination, national surveillance agencies arerecommendedtosetintermediategoalsthatwilleventuallyleadtotheprovision of high-quality water to all, but will not lead to impropercondemnationofrelativelyacceptablesupplies.”26(p.21)

Our second indicator for thiswas information from several survey questions: “Do you sufferfromdigestiveproblems?Stomachpain?Diarrhea?Vomiting?”and“Haveyounoticedachangeintheamountortypeofdigestiveproblemssincetheinstallationoftherainwaterharvestingsystem?”Samplestakenbeforeandafterfiltrationandfromprevioussourcesofdrinkingwaterallowedforcomparisons.Asub-sampleofwaterwasanalyzedforPAHs,mercury,andleadtoevaluatetheimpactsonairandrainwaterqualityoftheoilindustryactivitiesclosetoindigenouscommunities.Finally,thissamesubsamplewasanalyzedforzinctounderstandwhetherzinclevelsareaffectingthetasteof thewater.At levelsof4mg/l, zinc leavesanunpleasant,astringent taste to thewater.11Aquestiononthesurvey,likewise,assessedpeople’ssatisfactionwiththetasteofthewater.

25WHO.(2011)Guidelinesfordrinking-waterquality.FORTHEDITION.Geneva.26Fewtrell,Lorna&Bartram,Jamie.(2001).Waterquality:guidelines,standardsandhealth:assessmentofriskandriskmanagementforwater-relatedinfectiousdiseases/editedbyLornaFewtrellandJamieBartram.Geneva:WorldHealthOrganization.http://www.who.int/iris/handle/10665/42442

SamplingandQuestionnairesOver six months, teams of monitorsand one of two project leads (JenaWebb and/or Nicolas Mainville) wentto 17 communities to take samples ofwater (see table 1 for a summary ofsamples taken), conduct a shortquestionnaireandappraisalofsystemsand carry out in-depth interviews. Astandardized procedure for collectingsamples (seeAppendix2)andcarryingout the basic questionnaire andappraisal of the systems wasdeveloped.In-depthinterviewswerecarriedoutbythesameinvestigator(JW)onavarietyoftopics.Thequestionnaireandall interviewswereconductedinSpanishwithatranslatorfromtheCeiboAlliance team,onhand to translate intonative languageswhennecessary. Ineachcommunity,roughly15%ofthesystemsweresampled.PilotProjectA first iteration of the evaluation was carried out in November 2017. In this phase, severalmembers of the Ceibo Alliance carried out the sampling and the questionnaire in severalhouseholdsoftheirowncommunities.Atotalof24householdsweresampledandsurveyedinthisfirstphase.Thisphaseallowedustoevaluatethefeasibilityofthefullevaluation,toadjustthesamplingprocedureandvalidatethequestionnaire.TotalandFecalColiformAnalysisWithaPortableLabSevensystemswerealsoanalyzedusingaHach©MEL/MFTotalColiformportablelaboratory.Onesamplefromthefirsttankandtwosamplesfromthesecondtankateachsystemwerecollected insterile filtrationunits (Filter:GN-6Metricelwith0.45µmporesizeand47mmdiameter.Funnel:PolypropylenePetriDish:polystyrenelidandpolypropylenebaseNutrientPad:Cellulose.).ThesampleswerekeptinacooleroniceuntilreturningtoLagoAgrio,approximately2-3hoursaftercollection.OnceinLagoAgrio,watersampleswerefilteredthroughthepetri,thereagentwasaddedandpetrieswerestoredat37°Cfor24hours.After24hours, the number of bacterial colonies (red) and fecal colonies (blue) were counted andrecorded. A water engineer from Rainforest Flow, CalebMatos Chávez, was responsible fortakingthesamplesandprovideduswiththeresultsandrecommendations(seeappendix6).QuestionnaireA questionnaire was conducted with the owner of each system sampled and consisted ofquestions on the use of the systems (number of users, number of liters per day, etc.);Maintenanceof thesystems(howoften,etc.); theroof;previoussourcesofwatercollection;health(digestiveproblems,infections,etc.)andsatisfaction.(seeAppendix3)

In-DepthInterviewsIn-depth interviews were carried out byJenaWebbwithat leastonememberofeach community where sampling tookplace.Thetopicscoveredincluded:useofthe systems (ease, installation,maintenance, etc.); health (changes inhealth, types of health issues, etc.) andgeneralobservations.An interviewguidewascreated(seeAppendix4)andusedtoorient discussion but not adhered tostrictly.Conversationswererecordedandlistened to by the interviewer to gleanthemostsalienttopics.Themostrepresentativequoteswerethenselectedandarepresentedintheresultssection.Interviewswereconducteduntilsaturationofissueswasreached.EthicsApprovalThis evaluation represents a Quality Assurance, Testing and Performance Review of theCeibo/AF rundomestic rainwaterharvestingsystemandas suchdidnot requireethicsboardapproval(accordingtoCanadianPanelonResearchEthicsModule2TCPS2).UseofDataTheresultsfromthisevaluationwereintendedtobeusedinfourdistinctways.1)toinformtheusers of their water quality, 2) to inform users of commonmaintenance issues and how toprevent or resolve them, 3) to orient new iterations of the Ceibo/AF water program (seerecommendationsinAppendix1),and4)asacommunicationstoolinoutreach.

RESULTSANDDISCUSSIONTable 1 summarizes the samples taken and the interviews conducted in the context of thisevaluation. Sixty-two household heads (39 men and 23 women) from 17 Kofan, Siona andSeikopaicommunitiesrespondedtotheshortquestionnaireonuse,maintenance,healthandsatisfactionofthewaterandsystems.

Table1:SummaryofWaterSamplesandInterviewsConducted Kofán Siona Seikopai TotalWater samples for fecal coliform,pre-filter

18 20 14 52

Water samples for fecal coliform,post-filter

22 17 15 54

Water samples for fecal coliform,previousdrinkingwatersource

11 9 4 24

WatersamplesforPAHs,pre-filter 6 3 7 16WatersamplesforPAHs,post-filter 5 3 15 23Water samples for PAHs, previousdrinkingwatersource

4 9 3 16

Watersamplesformetals,pre-filter 6 3 6 15Watersamplesformetals,post-filter 5 3 14 22Water samples for metals, previousdrinkingwatersource

4 9 4 16

Shortinterviews 22 17 18 57In-depthinterviews 4 3 4 11UseoftheSystemsOn average there are 5 people per household and families use the system 4 times per day,consuming amean of 46 liters per day. 96% of families use thewater for drinking, 60% forwashingplates,13%forbathingand11%forwashingclothes(seeFigure1).Onlytwofamiliesdid not drink the water. One family in Puerto Bolivar had a municipal rainwater harvestingsysteminstalledwhichtheypreferredbecauseitwasclosertothekitchen.Thesecondfamily,fromTarabiaya,reportednotdrinkingthewaterandthatthewaterhadlittleredwormsinbothtanks.

MaintenanceoftheSystemsThreequantitativequestionsaddressedusermaintenanceoftheirsystem:“Whenwasthelasttimeyoucleaned the tank (days)?” “Whenwas the last timeyoucleaned thedrainagepipes(days)?”and“Howmanytimesperyeardoyouconductmaintenance/cleaningofthesystem?”Thecleaningof thetanksandthetubingwasquitevariableranging fromnevertoyesterday,and 0 to 12 times a year. 40% of the people reported never cleaning their tubing and 13%reportednevercleaningthetanks.Onaveragepeoplecleanedtheirsystems4times/year.StateoftheSystemsThepersonfillingoutthequestionnairereportedontwoquestionsevaluatingthestateofthesystem:oneaddressedtheroofofthehouse(installedbytheowner)andtheotherthewatersystem,bothonascaleof1to4,with4beingverygood,3good,2satisfactoryand1bad.46%of thesystemswere judgedtobeverygood,24%weredeemedgood,12%weresatisfactoryand18%wereinbadcondition(seeFigure2).Manyofthesystemshadfallenroofs.Thesystemroofsservetocoolthesystemsothatthewaterdoesn’tgethot.Further,bacteriareproducelessquickly incoolerwater.34%of thehouse roofswere judged tobeverygood,23%weredeemedgood,14%weresatisfactoryand29%wereinbadcondition.

HealthThis section presents quantitative data from the short questionnaire pertaining to healthquestions. Two sectionsof thequestionnairedealt specificallywithhealth:digestive troublesandinfections.86%ofpeoplereportedobserving improveddigestivehealth(reduceddiarrhea,vomitingandstomachpain)sincetheinstallationofthesystems,whiletherest,exceptforonewhoreportedworsenedhealth,reportednochange(seeFigure3).The results on infections (ear, urinary and skin) were more mixed: 44%, reported animprovement, while 56% had not observed a change in infections since the systems wereinstalled.Noonereportedworseninginfectionssinceinstallation(seeFigure4).

ProgramAcceptabilityTwoquestions,“Areyousatisfiedwiththetasteofthewater?”and“Areyousatisfiedwiththeoverall functioning of your system?” yielded quantitative data (yes/no) regarding theacceptability of the water systems. 100% of people responded yes to these quantitativequestions.Theaimoftheevaluation,namelytoreportbackonthequalityofthewaterandtoincorporate user feedback to improve the program, was explained to all respondents andpeoplegenerally, freelyprovidedsuggestions, small criticismsandbothpositiveandnegativefeedback,therefore,itdoesnotappearthattheresponsetothesequestionsweredesignedtoplease.Theresultsseemtorepresentagenuine,overallsatisfactionwithboththetasteandthegeneralfunctioningofthewatersystems.

Sample analysis of domestic rainwater harvesting systems and alternate sources ofwaterFecalandTotalColiformThe majority of drinking water samples collected for this evaluation contained no fecalcoliforms.83%ofpre-filtersamplesand85%ofpost-filtersampleshadnofecalcoliforms.Thefindingsweregroupedinathree-tierclassification27:zerofecalcoliform/100mlfor“Idealwaterquality”(classone),onetotenfecalcoliform/100mlfor“acceptablewaterquality”(classtwo)andmorethantenforfecalcoilform/100mlfor“lowwaterquality”(class3).Onatotalof54samples, forty-six of the post-filter samples (85%) were of ideal water quality, 2 wereacceptablewaterquality(4%),andsixofthepost-filtersampleswouldbeclassedaslowwaterquality(11%)(seeFigure5).

There is an average decrease of 68% fecal coliform between the pre-filter tank (mean fecalcount=27.5) and the post-filter tank (mean fecal count=8.7), although this difference is notsignificant (Wilcoxon test P=0.58). The high percent reduction probably does not yield asignificantresultbecauseofthelownumberofpost-filtersystems(n=8)inwhichfecalcoliformispresent;withahighersamplesizethismightbecomesignificant.

27Gould,J.&Nissen-Petersen,E.1999RainwaterCatchmentsSystemsforDomesticSupply:Design,ConstructionandImplementation.ImmediateTechnologyPublicationsLtd,London

Therewasnoassociationbetweenthelevelofwaterinthewatertanks(evaluatedasfull(2),partfull(1)andempty(0)bythepersontakingthesample)andtheamountoffecalcoliformobservedineitherthepre-orpost-filtersamples.Inotherwords,emptysystemsdidnothavehigherlevelsasmightbeexpectedsinceasthebiofilterdriesupitsabilitytofilteroutbacteriaisreduced.Neitherwasthereanassociationbetweenthe“stateofthesystem”andtheamountoffecalcoliform.Thiscouldbeeitherduetothelownumberofsystemswithpresenceoffecalcoliform,asabove,orbecausetheevaluationgiventothesystembydifferentmembersoftheteamvaried.Thestateoftheroof,asevaluatedbytheteammembertakingthewatersample,alsowasnotassociatedwithfecalcoliformlevels.Nosignificantrelationshipswerefoundbetweenthelevelsoffecalcoliformsinthepre-orpost-filter samplesandeitherof thecategories“last timethesystemwascleaned,”“last timethetubing was cleaned,” “last time the tank was emptied” or “number of times the system iscleanedperyear.”Thiscouldbedue,again,tothesmallnumberofsystemsthattestedpositiveforfecalcoliformsorrecallerroronthepartofinformants.The external observers alsomeasured fecal coliform levels in pre-filter and post-filter tanks.They found acceptable levels in the pre-filter tanks (average= 5.14 counts; where anythingbelow10ispassable)andacompleteabsenceoffecalcoliformsinthepost-filterreservoirs.Oursamplingstrategydidnotsystematicallyanalysesamplesfortotalcoliforms.Totalcoliformsarenotharmfultohealthbutaresometimesusedasanindicatorfortheoverallcleanlinessandintegrityofsystems.28Theexternalobservers(NSandCMS)wereabletotestfortotalcoliformsandallsampleshadarelativelyhightotalcoliformcountinboththepre-andpost-filtertanks,indicatingthatasubstantialamountoforganicmatterisenteringthesystems(seeAppendix5).Functioningmeshfiltersattheentrancetothesystemandafirst-flushapparatuswouldreducetheamountoforganicmatterenteringthesystem,improvingwaterqualityandincreasingthelife-spanofthesystems(seerecommendationsinAppendix1).Thealternativewatersourceswereclassifiedintosurface(riversandstreams)andsub-surface(springs) sources.On average the surface sources had 79 fecal coliform counts and the sub-surfacehad12.5,bothwouldfall intoclassthree(waterunacceptablefordrinkingpurposes).ComparisonofthemeansusingthenonparametricWilcoxontest(P<0.0001)showsthatthereis significantly less fecal coliforms inboth thepost-filter andpre-filter rainwater than in thealternatewatersources(seefigure6).28WHO.(2011)Guidelinesfordrinking-waterquality.FORTHEDITION.Geneva.

Figure 6: Box-plot of fecal coliforms in alternate surfacewater sources vs.rainwater.

Thisevaluationfoundthatthevastmajorityofrainwaterharvestingsystemshadlevelsoffecalcoliformsidealfordrinkingwaterandthatthesystemsoverallaresignificantlybettersourcesofdrinking water than the available alternatives, be it surface or sub-surface. Some simpleadjustmentsandmaintenancecould improvethesituationfor thosehouseholdswhere levelswerelessthanideal(11%ofsystems)(seeAppendix1:Recommendations).HydrocarbonsLevels of Polycyclic Aromatic Hydrocarbons (PAHs) weremeasured in a subset of the watersystems(n=39)andalternatewatersources (n=16).Noneof thesamples,whethertheywerefromthepre-filter,post-filter,oralternativewatersources,hadmeasurableamountsofPAHs(detection limit=0.00012). These results do not indicate that there is no petroleumcontaminationinthewater,butratherthatlevelsarebelowthedetectionlimitinwater.Othercompartments of the ecosystem where hydrocarbons accumulate, such as sediments andbottom feeders, arebetter indicatorsof petroleumpollution in theecosystem. These resultsonly indicate that even in the communities closest to oil operationswhere gases are openlyflared,thereisnohealthconcernfromPAHsinrainwater.MetalsMercuryandleadSimilarly,noneoftherainwaterharvestingsystemssamples(n=37)oralternativewatersourcesamples (n=16) had detectable amounts of mercury (detection limit=0.005 mg/l) or lead(detectionlimit=0.05mg/l).Theseresultsdonotindicatethatthereisnomercuryorleadinthewater. Levels of mercury in water are often very low; however, predatory fish are more

responsivetodifferencesinenvironmentalHglevels,concentratingupto106timesthelevelsfound in water29 These results indicate only that rain water represents no health hazard intermsofmercuryorleadconsumption.ZincMost samples (n=53), bothpre- (n=15) andpost-filter (n=22) and thealternatewater source(n=16),haddetectablelevelsofzinc.Allthesamples,however,hadlevelsbelowthatassociatedwithanunpleasanttaste(4mg/l).Pre-filterlevelswerehigher(mean=1.7mg/l)thanpostfilterlevels (mean=1.5mg/l),butthisdifferencewasnotsignificant (WilcoxonP=0.37).Thefiltersremovedonaverage12%ofthezincthatenteredthefirsttank.Thestateoftheroofwasnotassociatedwithzinclevelsineitherthepre-orpost-filtertank.Thematerialoftheroof(zincvs.duratecho)wasalsonotadeterminantofzinclevels,possiblybecauseonly3peoplereportedhaving only duratecho on their roof. There was a negative association, but not significant,betweenyearssincetheroofwasinstalledandtheamountofzincinthewater.QualitativeAnalysisThissectionpresentsobservationsmadebyintervieweeseitherduringtheshortquestionnaireorthein-depthinterviewsandisorganizedintothefollowingsections:Health,TheRainwaterHarvestingSystem,SecurityandEaseof theSystem,andOtherBenefitsof theProgram.Thecomments were overwhelmingly positive. Several recommendations were given and haveservedtoorienttherecommendationssectionbelow(seeAppendix1).“People are very satisfiedwith the rainwater harvesting program. Because before, local andnationalgovernmentsdidn’t focusonwatercollection.Andthanks to theCeiboAlliancetheyhavegainedthistremendousbenefitfortheirfamiliesandforalloftheSionaoftheCuyabenoregion.Peopleareverysatisfiedwiththesesystems.”(Vic-01-A,D)HealthAnoverwhelmingmajorityofpeoplecommentedon improvedhealthsincethe installationofthewatersystemsandexplicitlyassociatedthisimprovementwiththesystems.“Thesystemhaschangedeveryone’slife,becausenowwedon’tgetsicknesses:wedon’tthrowup,wedon’thavestomachachesanymore,soithaschanged,nowwelivebetter.Thekidslivewellandmywifetoo.”(SP-01-A,D)Agrandmother(SP-04-A,D)explainedthatwhenshewaslittle,theydidn’tsufferfromstomachpain, fever and diarrhea, but that her children did. She associated the change with thecontaminationoftherivers.Shethenexplainedthathergrandchildren,whodrinkwaterfromtherainwaterharvestingsystems,donotsufferfromtheseailments.

29BaudoR,GiesyJ,MuntauH(1990)Sediments:chemistryandtoxicityofin-placepollutants.CRCPress,Norvara

“Whatwouldwedoifweranoutofwater?We’dhavetosuffer,likewedidbefore…itwasverydifficult…”(SP-04-A,D)Peoplealsoappreciatedthetasteofthewater,animportantcriterionintheacceptabilityofaclean water program, which also has health promotion effects through the consumption ofadequatequantitiesofwater.“Itgivesagoodfeelingwhenyoudrinkit.Itdoesn’thavethatsmellthatithaswhenitcomesfromaspring,whenamuddyodorcomeswithit.It’snotlikethat.Ithasahealthytaste,sortofperfumed.”(UK-01-02)The only health issue that was raised and associated with the systemwas stomach troublewhenthesystemdriesup,andthisonlybytwopeoplefromthesamecommunity(BV).“Our health is better now, but it is possible that there aremore digestive troubles when itdoesn’train.”(BV-01-AD)TheRainwaterHarvestingSystemWhenaskedaboutthefunctioningofthesystemsthemajorityofpeopleweresatisfiedandhadonly minor complaints with certain aspects of the system, namely the roofs and the meshfilters.Atthesametime,peoplerecognizedthattheywereresponsiblefortheupkeepoftheirown systems. Several interviews addressed the information that people received during theinitialworkshops.Thosewhospokeof thisweregenerally satisfiedwith the information thattheyreceived.Severalpeoplementionedthatitwouldbeusefultoprovideinformationtonewarrivals in a second workshop; however, new arrivals explained that they did receiveinformationabout thesystemeither fromfamilymembersor, in thecaseofone intervieweewhowasnot living in thecommunitywhenthesystemswere installed (Vic-01-A,D), fromthepresidentofthecommunity.Thisindicatesthatinformationaboutthesystemscirculatesinthecommunities independent of interventions by Ceibo/AF. In in-depth interviews people gavetheirsystemsanaverageof20years.“The tank is functioningwell…this is so important for drinkingwater. That iswhy I’m clearlytellingyouthatitissohelpfulthatyoubuiltthesesystemslikethis.Becausetheygetreallyfull.It is so easy for me. I can’t say that they don’t help. Really, I say they are working well,perfectly.”(SP-04-A,D)“Mycriteriaisthatitdependsoneachofushowwemaintainthesystem,becauseifwedon’tkeep that up, that’s the point. It someone gives you something, if you maintain it, it lastslonger.Butifyoudon’ttakecareofit,itwillbreak,right.”(Bol-03-A,D)“Wehavetotakecareofourownsystems…this iswhatwasexplainedintheworkshop…takecareofit,cleanit,it’sfordrinking.”(Bol-03-A,D)

Manypeoplenotedthatthemostdifficultparttoupkeepwasthisplasticroofoverthesystem.Inalonginterview(SP-04-A,D)arespondentwithafallenroofexplainedthattheywerehardtomaintainbecause thestructureof the roof ismadeofwood,which rots,and that theplasticrips.Butshealsoconfirmedthatthetasteofthewaterdoesnotchangewhenitiswarm.Otherpeoplerecognizedthattheysimplyhadn’thadthetimetofixitafterithadgottenoutofplace.E.gSP-07-A,Dsaidthathehimselfwasfullyresponsibleandthattherewasreallynotatechnicalissuethere.Another issuewith thesystemsthatcameupwasthemesh filterat theentranceto the firsttank.Severalrespondentscommentedthattheyrippedorthattheycloggeduptooquickly.“Yes, therewas ameshwhen the systemwas installed, itwas kindof delicate,white, but itdidn’tlast,justtwomonthsandthenitwasallripped.Ichangedittoafinecloth,notsofineasclothing,butalmost.Itlastedthreemonths.Whenitisdelicate,ithastobechangedoften,sothatfrogsdon’tgetin.”(SP-01-A,D)TheonlynegativecommentsaboutthesystemcamefromseveralresidentsofPuertoBolivarand Tarabiayawhowere skeptical of the sand filters. Itwould be important to re-explain tothese communities the importance of the filters in the filtration process and their benignnature.SecurityandEaseoftheSystemAnotherpointthatwasrepeatedlymadeintheinterviewswasthecontributionofthesystemsto reducing the time required and the energy necessary for collectingwater. Irrespective ofwhere the previous water source was located, the systems being right by the doorstepconstitutedanimprovement.“Thetanksmakeiteasierbecausewehavewaterrighthere.Beforewehadthetanks,wehadtogodowntotheriver,lugupbuckets,andgoback,upanddown,often.Wehadmorework.Butnow,no,wehavethetankrightatthefootofthehouse.Iusedtogettired.”(Bol-03-A,D)Thesystemsalsocontributedtoasenseofwatersecuritywhichwascommentedonbyseveralrespondents. People were able to identify when they were at risk of running low on water(December, January and February, at the end of the dry season) and also water consumingbehaviors to reduce the inconvenience of lacking water. For example, Vic-01-A,D, from SanVictorianosaid,“Here inDecember, JanuaryandFebruary it’s themostcomplicated,becausetherivergetstotallydry,just50cm.Itdependsoneachfamily’s[use],tonotbewashingplates,onlyfordrinkingandcooking,thenwashtheplatesintheriver,aswedofortheclothes.Youhavetoreallytakecareofthewater,careforitmorethangold.”(Vic-01-A,D)“Having rainwaterharvesting tanks is reallypractical for the familybecause thenyou’re surethat you’re going to collectwaterwhen it rains.And ifwedidn’t have the systemyoudon’tknowwhattocollectthewater in,youcouldbeusingtanksthataren’tsuitableforcollectingrainwater.Weusedtocollectwaterinmetaltanksandyouknowmetalletsoffanacidorsome[chemical].Thissystemisbetterbecausethetanksaresuitablefordrinkingwater.”(Vic-01-A,D)

“Wefeelsecurewiththesystem.”(Bue-01-A)“Kidsjustcomeanddrinkwater.”(Bue-02-A,D)OtherBenefitsoftheProgramOurresultsdidnotallowustodetermineiftherainwaterharvestingprogramledtoincreasedawareness of environmental health, as has been shown in previous studies30, since no pre-training baseline information was collected. However, respondents appeared to have asophisticatedunderstandingof theconnectionsbetween theenvironment,waterandhealth,asthefollowingquotesdemonstrate:“Now there are so many people, so much contamination, open gas-flares right nearby,sometimes it rainsandtheopen-pitwastepoolsdrain thewastes, theblackpetroleum,rightintotheriver.”(SP-01-A,D)Aninterviewee(Vic-01-A,D)explainedclearlythatheknewthatownersofthesystemsshouldnotkeepsoaporpesticidesnearthesystem.Helaughedsheepishlyashemotionedtohisownsystemwhereapesticidetankwasrestingonthebaseofthewatersystem.Thisindicatesthatwhiletheinformationhasbeenassimilatedtheassociatedbehavioralchangeshavenotalwaysfollowed.InresponsetoaquestiononchangesinthelocalRiver,Waiya,sincetheinstallationoftwooilplatforms in the headwaters an informant said “Yes, it’s changed. Mostly in terms of skininfections.That’swhywedon’tbathetheremuch.Justalittle.”(WY-02A,D)“Beforewelivedbetter,becausetherewasnocontaminationbyTexaco.Wedidn’tsufferfromillnesses.Mygrandmotherlivedtobelike100.Butshewalkedwell,andsheworked.Andnowwiththiscontamination,itisworse…[theelders]aterightfromtheriver,drankrightfromtheriver,butitwasn’tcontaminated,therewasnopetroleuminit.Theydidn’tgetfevers,stomachaches,diarrhea.That’swhyshelivedtobe100,andwalkedwithoutacane.”(SP-04-A,D)In communitieswith roads nearby, such as BellaVista, residents spokeof concernover dustbillowingupfromthepassingtraffic.Oneresident(BV-04-A,D)explainedthough,thatthisdustis a problemwhen they collect rainwater in a tankwithout a filter, but that in the systemsinstalledbyCeibotheydonotnotethedustinthefilteredwater.30Davis,L.F.,M.D.Ramirez-Andreotta,J.E.T.McLain,A.Kilungo,L.AbrellandS.Buxner(2018)."Increasingenvironmentalhealthliteracythroughcontextuallearningincommunitiesatrisk."InternationalJournalofEnvironmentalResearchandPublicHealth15(10).

CONCLUSIONSTheresultsfromthequantitativeandqualitativedatacollectedinthisevaluationpointtohighwaterquality,improvedhealthsincetheinstallationofthesystems,easeofuseandresoundingusersatisfaction.85%ofsystemshadnofecalcoliformswhatsoeverandonlysixsystemshadunsatisfactorywaterquality.Amajorityofpeople(86%)pointedto improveddigestivehealthsince the installation of the systems. In-depth interviews, especially those carried out withwomen,indicatedthatthesystemshavereducedtheirworkload.Finally,100%ofrespondentsreported being satisfied with both the taste of the water and the system itself. Smalladjustmentswhichwould reduce both the amount of total and fecal coliforms in thewatercould be implemented to aim for 100% Class 1 water, ideal for drinking (see Appendix 1);however,evenas is, thedrinkingwater representsa substantial improvementoveralternatesourcesofwater,whichwerefoundtobe,onaverage,unacceptableforhumanconsumption.In closing, the ensemble of this evaluation reveals a program that is working to provideexcellentdrinkingwatertosomeoftheAmazon’smostremotecommunities.

APPENDIX1:RECOMMENDATIONSThisappendixpresentssomerecommendationsforthedirectionoftheCeiboAlliance/AFwaterprogram.Theyarecategorizedintohigh,mediumandlowpriority.SecondWaveofMaintenanceWorkshopsPriority:highDespite the fact that interviews with new-arrivals indicated that there has been sufficientuptake of the program for the auto-circulation of necessary information on the use andmaintenance of the systems, several people mentioned that it would still be good to haveanotherworkshopsothatpeoplewhomarriedintothecommunityoryouthwhorecentlycameofagecouldgettrainingintheuseandmaintenanceofthesystems.Onewomen(Bol-03-A,D)mentionedthatwhileshethinksthatsheunderstoodalltheinstructionsinSpanish,Spanishisnotherfirstlanguage.Therefore,theresultsofthisevaluationpointtotheneedforasecondsetofworkshopsontheuseandmaintenanceofthewatersystemsandthattheseworkshopsbefullytranslatedinthelocallanguage.Amaintenancevideocouldbemadetoshowtheexactstepsnecessaryforregularmaintenance.Thisworkshopcouldincludeinformationon:

● Thepurposeandnon-toxicnatureofthequartzandfiltercomponents● Thebiologicallayerandtheimportanceofkeepingitwet● Commontechnicaldifficulties(mesh,roof,dripping,sandinsecondtank)● Bestmaintenancepractices(howtoaccessandcleanthetanks,tubing,etc.)● Whattodoincaseofmajordamage

MaintenanceGuidePriority:highThestepsforregularmaintenanceandhowtoresolvecommonproblemsshouldbewrittenupintoashortandeasytoreadguidewithampleimagestoaccompanydescriptions.Muchofthesameinformationshared intheworkshopshouldbe included inthisguideand leftwitheachhousehold.NeedsAnalysis,Sampling,InterviewsandTraininginWaoraniTerritoryPriority:highSamplingandinterviewsneedtobeconductedinWaoraniterritory.Aprogram to return toWaorani communitiesandcarryoutanevaluation similar to theoneconductedhereaswellasfurthertraining(asabove)andaneedsanalysiswouldbeimportantasanextiterationoftheCeiboAlliance/AFwaterprogram.TrainingandProperInstallationofSystemsinTarabiayaPriority:high

Throughinterviewsandinspectionsofthesystemswelearnedthatmanyofthesystemswerenot properly installed in Tarabiaya. The Ceibo Alliance/AF should plan to return to thiscommunity and assist residents in the proper installation of their filters. According tointerviewees, the initial technician is no longer available tohelp residentsof this communitywiththeirsystemsandtherefore,anewtechnicianshouldbetrained.WomenCommunityWaterTechnicianTrainingPriority:medium-highManystudieshaveshownthat including theactiveparticipationofwomen inWASHprojectsgreatly increases their sustainability due to women’s central role in household watermanagementand their greater likelihood to remain in theirnative communities. “A studybytheInternationalWaterandSanitationCentre(IRC)ofcommunitywaterandsanitationprojectsin 88 communities in 15 countries found that projects designed and run with the fullparticipation of women are more sustainable and effective than those that do not. Thissupports an earlier World Bank study that found that women’s participation was stronglyassociatedwithwaterandsanitationprojecteffectiveness.”31Itwouldbeadvisable,inaseconditeration of the project, to actively recruit and trainwomen from the communities aswatertechnicians and have thesewomen participate actively in theworkshops as native languagefacilitators.Thiscouldbeaprogramwideactivityoritcouldbeginwiththecommunitieswhich,foronereasonoranother,nolongerhaveatechnician(e.g.Tarabiaya).AcquisitionofaPortableLaboratoryandRegularEvaluationofSystemsPriority:mediumA portable lab would facilitate a continuous evaluation of the systems, especially incommunitiesmorethan6-hoursfromCoca.Itwouldbeadvisabletosporadicallytestsystemsineachofthecommunitiesatleastonceayear.Realisticgoalsforprogressiveimprovementofwater quality indicators could be set and monitored through a yearly quality rating (for anexample seeWHO, 2011, p. 91).With a portable lab this could be done in conjunctionwithotheractivities.Theexternalobserverswhocametoseethewatersystems(NSandCMC)haveaHACHMELlabandrecommendedittous.TheHACHMELlabisvaluabletooltomonitorwaterqualityandtoeducatethepeopleaboutunseenfecalandtotalcoliformsinthewater.TheysharedthenameoftheHACHDonationCommitteeChairandsuggestedtogetintouchtoseeiftheycouldgiveCeibo Alliance/AF an In-Kind donation of a MEL Laboratory and supplies. TIM SCHMITT|Technical Support/Donations Committee Hach CommunityOutreachhachhelps@hach.comP970.669.3050x 6210 |F970.619.5058 Hach Company|www.hach.com|InstallaFirst-FlushSystemPriority:medium

31Inter-agencyTaskForceonGenderandWater(GWTF)(2015).Gender,WaterandSanitation:APolicyBrief.UN-WaterandtheInteragencyNetworkonWomenandGenderEquality(IANWGE),UnitedNations.

Allliteratureconsultedforthisreportindicatesthatfirst-flushsystemsimprovewaterquality.Afirst-flushsystem,whichwould ideallydivert the first5mmof rain,preventsorganicmaterialaccumulatedontheroofsincethelastrainfromenteringthefirsttankorcloggingupthemeshfilter at the entrance to the tank. Thehigh levels of total coliformsobserved in the samplescollectedwiththeexternalobservers(NSandCMC)couldbeavoidedbydivertingthefirstrainand the organic material that comes with it. The system maintenance is also reduced byavoidingtheentryofasubstantialpartoftheorganicmatterfoundontheroof32,33.First-flushsystems are not difficult to install. Abbasi34 suggests how to calculate and install first-flushsystems.Itwouldbeadvisabletoinstallafirst-flushsystemonallnewsystemsandconductanevaluationof thenewsystemsspecificallyaccessedtocollectuser feedbackonthefirst-flushsystem.Ifthereisuseracceptabilityofthefirst-flush,itwouldbeadvisabletoretro-fitafirst-flushonpreviouslyinstalledsystems.SystemDatabasePriority:mediumAdatabaseof all the systemswould facilitatemaintenance, futureevaluations andplanning.TheBluePlanetNetworkprovidesa freeplatform for cleanwaterorganizations to storeandeasily update water system information (http://blueplanetnetwork.org/programs/platform).Theplatformseemstobeavailableonly inEnglish,butsomethingsimilarwouldallowfortheinput informationonthesystemsfromthefieldviacellphones.Consultationofthedatabasepriortoscheduledvisitswouldalerttechnicianstoanyrecentneeds.NewRoundofRainwaterHarvestingSystemsinCommunitiesAlreadyServedPriority:mediumSinceinstallation,somecommunitieshavegrownandnewarrivalsdon’talwayshaveasystem.At some point it would be good to do a second round of installations in each of thecommunitiesbecausepeoplehavegrownupandmovedinsincethefirstinstallation.Itwouldbegood togiveample forewarning thatpeoplecanbuildpillars toput the systemuphigherbeforetheinstallationteamcomes.Innewsystems,itismoreefficienttoincreasethesizeofreservoirthanthefiltertankbecausethenpeopledon’teverhavetowaitforwater.Acommoncalculationisthatthereservoiris8timesthesurfaceareaoftheroof.OverflowMechanismsPriority:mediumWhile conducting interviews,we learned that therewere no overflowmechanisms (tubo dedesfogue), that control the level of water automatically, installed on the systems in San

32deKwaadsteniet,M.,P.H.Dobrowsky,A.vanDeventer,W.KhanandT.E.Cloete(2013)."DomesticRainwaterHarvesting:MicrobialandChemicalWaterQualityandPoint-of-UseTreatmentSystems."Water,Air,&SoilPollution224(7):1629.33Helmreich,B.andH.Horn(2009)."Opportunitiesinrainwaterharvesting."Desalination248(1):118-124.34Abbasi,T.andS.A.Abbasi(2011)."Sourcesofpollutioninrooftoprainwaterharvestingsystemsandtheircontrol."CriticalReviewsinEnvironmentalScienceandTechnology41(23):2097-2167.

Victoriano,sotheyhavetoturnitonandoffaccordingtotherain.(Vic-01-A,D)SeveralofthehousesinDaiparealsodidn’thavetheoverflowmechanism.Itwouldbegoodtoverifythisinallhouses inSanVictorianoandothercommunitiessusceptibletohavehadasimilar installationmethodandretro-fitthesystemswiththeoverflowmechanism.StagnantWaterControlPriority:lowStagnant water was observed near water systems in many households. Stagnant water is avehicle for water borne diseases and also creates breeding habitats for vectors, such asmosquitos,ofvectorbornediseases,suchasdengue. Itwouldbegoodtodiga1.5by2 foothole in frontofthefaucetandfill itwithriverrockstoavoidpoolingbelowthefaucet. If theCeiboAlliance/AFdoesnottakethistaskon,itcouldbesomethingthatisrecommendedthatusersdothemselvesattheworkshop.

APPENDIX2:WATERSAMPLINGAstandardizedsamplingprocedurewasestablishedandconsistedofthefollowing:

1. Putongloves2. Takeoutanew,cleanbottleandcap3. Openthefirsttankandfillthebottlebyinsertingthebottleunderwateruntilitiscompletelyfull4. Drythebottle5. Identifythebottlewiththesamplenumber,dateandtime6. Sealthebottlewithtape7. Putthesampleimmediatelyonice8. Takeoutasecondbottleandcap9. Turnonthetapofthesecondtankandletitrunfor5seconds10. Fillthebottleuntilitoverflows11. Repeatsteps4-712. Aftereachhouse,disposeofgloves13. OnceallsamplesaretakengoimmediatelytotheLABSUlaboratorytodeliversamplesforE.colianalysis

andwithinnomorethan6hoursofsampling.14. SamplesforPAHandmetalanalysisarekeptinarefrigeratoruntilbeingsenttoQuitobybusforanalysis

atLABANCY

To take samplesof the riversandstreamsa similarprocedure to sampling the first tankwasused.Threesamplesweretakenseveralmetersapart.GPSpointswererecorded.The fecal coliformanalysiswas carriedout the the laboratory LABSU,Coca, Ecuador and thePAHandmetalanalysiswereconductedatLABancy,Quito,Ecuador.

APPENDIX3:QUESTIONNAIRE#demuestra:__________________Fecha:_____________Horadetomademuestra:_______

Comunidad:__________________CoordenadasGPSdelsistema:________________________

Nombreyapellidodeldueñodelsistema:___________________________________________

¿Vivenpermanentementeenlacasa?SiNo:¿Conquéfrecuenciaestán?__________

Usodelsistema

¿Cuantosusuariosdelsistema?:____________Añodeinstalacióndelsistema:_____________

¿Cuántasvecesaldíaseusaelsistema?_____________________________________________

¿Paraqueseuseelsistema?______________________________________________________

¿Cuantasollas(ogalones)deaguaseusadiarioaproximadamente?______________________

¿Cuántosdíasdesdelaúltimalluvia?____Elsistemaestá:lleno___mediovacío___vacío___

Mantenimientodelsistema

¿Cuándofuelaúltimavezqueselimpióelsistema?___________________________________

¿Cuándofuelaúltimavezqueselimpiaronlostubosycanalespararecolectarelaguadel

techo?

_____________________________________________________________________________

¿Cuándofuelaúltimavezquesevacióelsegundotanque?_____________________________

¿Cuántasvecesalañoustedhaceelmantenimientodelsistema?________________________

Estadodelsistema:MuybuenoBuenoSatisfactorioMal

Techodelacasa

¿Dequéestáhechoeltecho?_____________________________________________________

¿Eltechofuepintado?SiNo¿Cuándo?_____________________________________

¿Cuándofueinstaladoeltecho?___________________________________________________

Estadodeltecho:MuybuenoBuenoSatisfactorioMal

Ubicación:sombrasol

Observacionesycomentarios______________________________________________________

______________________________________________________________________________

Otrasfuentesdeagua

¿Donderecolectabaelaguapotableantesdelainstalacióndesusistema?

______________________________________________________________________________

¿Ustedsigueusandootrasfuentesdeaguapotable?SíNoSilarespuestaessí:

¿Cualfuente?________________________________________________________

¿Conquéfrecuencia?¿Cualuso?

________________________________________________________

¿Hayunafuentedecontaminacióncercadelacasa?¿Dóndeestáelmecherolomáscercano?

______________________________________________________________________________

Salud

¿Ustedessufrandeproblemasdigestivas?:¿Dolordeestómago?Si.No

¿Diarrea?Si.No¿Vomitó?Si.No

¿Hannotadocambiosenlacantidadotipodeproblemasdigestivas(despuésdelainstalación

delsistema)?___________________________________________________________________

¿Ustedessufrandeinfecciones?:¿Deloído?Si.No

¿Infeccionesurinarias?Si.No.¿Infeccionesdepiel?Si.No

¿Hannotadocambiosenlacantidadotipodeinfecciones(despuésdelainstalacióndel

sistema)?______________________________________________________________________

¿Cómoeselsabordelagua?_______________________________________________________

¿Estánsatisfechosconelagua?Si.No

¿Estánsatisfechosconelsistemadeagua(lugar,manteamiento,etc.)?Si.No

¿Cómohacambiadosuvidadesdelainstalacióndelsistema?___________________________

______________________________________________________________________________

APPENDIX4:INTERVIEWGUIDEAcontinuación,encontraráunapropuestadeguíadeentrevista.Esteesunmarcodediscusióngeneral que es flexible y abierto. Hemos agrupado aquí las ideas de las preguntas. En estesentido,estaguíasepuedeadaptaromodificarsegúnseanecesario,segúnlasparticularidadesdecadacomunidadysegúnlasrespuestasdelosparticipantes.Existetambiénuncuestionarioconpreguntasmáscerradasquetodosdebenresponderyquedebeserhechoconcadafamiliaqueproporcionaunamuestradeagua(verdocumentoX).Etapa1:DescribabrevementeelproyectodeevaluaciónEtapa2:LaentrevistaNombre y apellido:______________________________________________________________Edad: ________________________. Comunidad:____________________________________Númerodelamuestra:________________________Sexo:FMNota:siunamuestranofuetomadaalacasa,llenarelcuestionariotambién.Dividimosnuestraentrevistaen3secciones:1.Lossistemasdeagualluvia2.Lasalud3.ConclusionesIndicaalinicioque“Sepuedehablarlibremente,nohayunarespuestacorrectaoincorrecta.Losdatosserántratadosconfidencialmenteparanoasociarpalabrasconunapersonayconfinesdemejorarelprograma.”Sección1–LossistemasdeaguaDesde 2011 la Alianza Ceibo y Amazon Frontlines han instalado XX sistemas de agua en 72comunidades.¿Cuálfueelproblemainicialolasolicitudinicialquehallevadoalainstalacióndesistemasdeagualluviaensucomunidad?¿Haevolucionadoestasolicitudconeltiempo?¿Dequémanera?¿Haparticipado/austedmismo/aenalgunadimensióndelproyecto?¿Dequéforma?¿Cómoestáfuncionandosusistema?¿Llenacadalluviaonollena?¿Cuántasvecessesecaporaño?¿Cómolohace?¿Recibióinformaciónsobreelmantenimientodelsistema?¿Cuándo?¿Sesientequetienetodalainformaciónquenecesitaparaelbuenmantenimientodelsistema?¿Cadacuántohacelalimpieza?¿Quiénhacelalimpieza?

¿Qué se puede hacer cuando hay problemas? ¿Cómo usted se resuelva los problemas oinconvenientesqueseencuentra?¿Cuáles fuentesdeaguaalternativasutilizausted?¿Encuálessituaciones? ¿Quétratamientohaceustedalaguaalternativa?¿Estásatisfecho/aconelagua?¿Conelsistema?¿Quéopinausteddelossistemasdeagua?Mirandohaciaatrás, ¿cuál fueel impactode la instalaciónde sistemasdeagua lluvia?¿Paraustedydentrodesupropiacomunidad?¿Havistounarecepciónpositivadelossistemasdeagualluvia?¿Quién,quéactoresestuvieronmásafavordelproyecto?¿Hacambiadoestasituaciónduranteelproyecto?Porelcontrario,¿presencióalgunadisensiónentrelaspartesinteresadasenlaconduccióndelproyecto?a.¿Quéactoresteníanreservasacercadelproyecto?b.¿Quéreservasseexpresaron?Sección2-SaludDesdequelascompañíaspetrolerashanllegado,ciertosmiembrosdelascomunidadessehanpreocupadaporlasaluddelagente.Lassiguientespreguntassonsobreestapreocupación.¿Estápreocupadaporsusalud, lasaluddesufamiliao lagentedeestacomunidad?¿Encuálsentido?¿Cómo va la salud de los adultos de la comunidad? ¿Ha notado algún cambio desde lainstalacióndelossistemasdeaguadelluvia?¿Cómovalasaluddelasniñasylosniñosdelacomunidad?¿Hanotadoalgúncambiodesdelainstalacióndelossistemasdeaguadelluvia?¿Quiénhace la recoleccióndeaguaen la familia? ¿Hubo cambiosenel tiemponecesarioenelesfuerzonecesariopararecolectaragua?Sección3-Conclusiones¿Hahabidounimpactodelossistemasdeaguaamáslargoplazo(meses,añosdespués)enlacomunidad?¿parausted?¿Cómo calificaría la satisfacción de su comunidad con el resultado del proyecto? ¿Y susatisfacciónpersonal?Explicar.¿Hahabidoalgúnimpactodelproyecto,acortoolargoplazo,conrespectoalasensibilizacióndelagentedecuestionesdelacontaminaciónambiental?Enretrospectiva,¿quépodríahabersemejoradoenlainstalaciónolaparticipacióndelossociosduranteelproyectodesistemasdeagua?¿Tienealgunasugerenciaparamejorarelprograma?Graciasporsucooperación.

APPENDIX5:SUMMARYOFRESULTSFROMTHEEXTERNALOBSERVERS(NSANDCMC) SanPablo Bavoroe Upirit

u Total

#samples

#samplestestpositive

Averagecount

#samples

#samples testpositive

Averagecount

#samples

#samples testpositive

Average

#samples

#samples testpositive

Averagecount

Fecalcoliform, pre-filter

3 2 11 2 1 1 2 1 0.5 7 4 5.14

Fecalcoliform,post-filter

3 0 0 4 0 0 4 0 0 11 0 0

Totalcoliform, pre-filter

3 3 666.66

2 2 1000 2 2 1000 7 7 857.14

Totalcoliform,post-filter

3 3 40.66 4 4 538.5 4 4 105.25

11 11 245.18

Fecalcoliform,municipalwater

1 1 6 0 na na 0 na na 1 1 6

APPENDIX6:RECOMMENDATIONSOFCALEBMATOSCHÁVEZ,RAINFORESTFLOWEl proyecto de agua de Amazon Frontlines es muy innovador y alternativo con una visiónfuturista, que podría ser replicado en las selvas de todo el mundo, creemos que lasrecomendacionesdadas,sonparamejorarestossistemasyquepuedanserreplicandosdeunamejormanera.INSTALACION CORRECTA

1.Elniveldeaguaentreelingreso(aguadelluvia)ylasalidadelbio-filtronodebesermenosde35cmnimayora40cm.2.Enesta foto (abajo)semuestraquehaymásde1metrodediferenciadealtura, loquecreaunafuertepresióndeaguaenel fondo(cargahidraúlica)quepermite ingresarbacteriasorgánicasporencimadelosrangospermitidosporOrganizaciónMundialdelaSalud(OMS).INSTALACION INCORRECTA

Foto: Casa en San Pablo. 3.Es importante realizar un analisis físico químico de la sustancias de color negroacumuladasenlasuperficiedelosbio-filtros,paradescartarlapresenciadeaceitesualgunosotrosagentescontaminantesqueestenalterandolosaltosnivelesdecoliformestotales.4.Esmuyimportantedescargarlasprimerasaguasdelluviafueradellosbio-filtrosporunosminutos,dependiendodeltamañodeltecho,paraevitarelingresodecontaminantesadheridosenlascalaminasquepudieranalterarelinteriordelfiltro.5.Cadafiltrodebetenercomomínimo40cmdearenalavadaderío,15cmdegravade1cmy10cmdepiedrasde5cm.6.Lasmuestrasrecogidasyanalizadasarrojaronaltosporcentajesdecoliformestotalesenlos filtros, los cuales no son muy alentadores, esto se puede mejorar si seguimos lasrecomendacionesdadasenteriormenteenla#5.7.Para monitorear la calidad del agua es necesario realizar análisis bactereologico comomínimounavezalaño.