AndreaDiGuardo
b,g,StefaniaColombini
c,GianpaoloBertoncini
d,MarcoCastelnuovo
d,
GuidoSali
a,MaurizioMoschini
e,MattiaSanna
a,AlessiaPerego
a,MarcoCarozzi
f,
MarcelloErmidoChiodini
a,MattiaFumagalli
a,∗aDepartmentofAgriculturalandEnvironmentalSciences–Production,Landscape,Agroenergy,UniversityofMilano,ViaCeloria2,20133Milano,Italy
bInformaticaambientaleS.r.l.,ViaTeodosio13,20131Milano,Italy
cDepartmentofAnimalScience,UniversityofMilano,ViaCeloria2,20133Milano,Italy
dRegioneLombardia,PiazzaCittàdiLombardia1,20124Milano,Italy
eInstituteofFoodScienceandNutrition,AgriculturalFaculty,CatholicUniversityofPiacenza,ViaEmiliaParmense84,29100Piacenza,Italy
fINRA,UMR1091,EnvironnementetGrandesCultures,F-78850Thiverval-Grignon,France
gDepartmentofEarthandEnvironmentalSciences,UniversityofMilanoBicocca,PiazzadellaScienza1,20126Milano,Italy
a r t i c l e i n f o
Articlehistory:
Received22October2013
Receivedinrevisedform19March2014 Accepted20May2014
Keywords:
Decisionsupportsystem GIS
Integratedevaluation Manuremanagement Multidisciplinaryindicators
a b s t r a c t
IntensiveagricultureandlivestockbreedingrepresentcriticalfactorsintheLombardyregionsincethe nitratevulnerablezonesare62%ofutilisedagriculturalplainarea.Theaimofreducingthe environmen-talriskcausedbyagricultureactivities(e.g.nitrogenlossesintogroundwaterandatmosphere)canbe onlyachievedthroughacriticalandscientificanalysisoflivestockmanuremanagementinawhole-farm perspective.Keepinginmindthisobjective,thedecisionsupportsystem(DSS)ValorEwasdeveloped.It canbedescribedasatoolabletoevaluatefromtheenvironmental,technical,agronomicandeconomic pointsofviewthemaincomponentsofmanuremanagement(production,storage,treatmentandland application)foravarietyoflivestocktypes(i.e.,cattle,swine,poultry,sheep,goatsandhorses),under differentscenariosadoptedatfarmandterritorialscale.ValorEconsistsofthreemaincomponents:data
managementsubsystem,modelmanagementsubsystemandtwoversionsofuser-interface,bothfor
farmandterritorialscale.MostoftheinputstotheDSScomesfromexternaldatabases,whileasoftware tooldevelopedinthe.NETenvironmentandimplementedusingobjectorientedprogramming(C# lan-guage),providesthelogictomanagethescenariosimulationofagronomicandenvironmentalfarm-scale models.Usersandstakeholderscancarryoutcomparativeanalysis,startingfromtheknowledgeofthe currentperspective,intermsofmanuremanagementsystematfarmorterritorialscalebyinterrogating theavailabledatabases.Moreover,theycangeneratedifferentalternativescenariosthankstodifferent optionsforthemanurehandlingandcroppingsystemsimulation.Thentheycanfinallyevaluateand comparedifferentscenariosthroughmultidisciplinaryandsyntheticindicatorsbutalsovisualisespatial effectsexploitingthecoupledwebGIS.ValorEisthereforeanattempttoofferacomprehensivetoolfor improvingbothfarmstrategyanddecisionmakingprocess,whichisparticularlyimportantinavery intensiveagriculturalarea,withoneofthehighestlivestockdensityintheworld,asLombardy.
©2014ElsevierLtd.Allrightsreserved.
∗ Correspondingauthor.Tel.:+390250316611;fax:+390250316575. E-mailaddress:[email protected](M.Fumagalli).
Introduction
Livestockproduction,responsibleofabigpartofagricultural landuseforgrazingandfeedproduction,determineserious envi-ronmentalproblemssuchasgreenhousegasemissions(Steinfeld etal.,2006)andemissionsofreactivenitrogen(N)inatmosphere
and water (Oenema, 2006). These problems are getting much
http://dx.doi.org/10.1016/j.landusepol.2014.05.007
culturalpoliciesandregulationsforpreventingpollutionofland,air andwater.Thecoreofthelivestockproductionisthemanure man-agementfromtheanimalexcretiontothelandspreading,becauseit affectsboththequalityofsoil,air,waterandthecropgrowth,and consequentlyitbearsonthefarmincome.Theselectionof
live-stockmanuremanagement optionsis becomingastrategictask
thatfarmersandpublicpolicymakershavetohandleproperly.As presentedbyKarmakaretal.(2010),severaloptionsformanure collection,storageandlandapplicationareavailable.Moreover,as discussedbyPetersenetal.(2007)avarietyofmanuretreatments withaspecifictargethasbeendevelopedaswellasimprovements
inanimalnutrition tocontrolmanureproductionand
composi-tion.Consequently, beforeinvesting money, it is of paramount importancetogetasupporttoolthatcouldassiststakeholdersand farmersonidentification,evaluation,andselectionofthemore suit-ableoptionofthemanuremanagementforaspecificareaandaim. Infact,eachmanagementstrategyhasitsadvantagesand disad-vantageswhenconsideringenvironmental,agronomic,technical, energetic,costandlabourissues(Fumagallietal.,2012).
Adecision supportsystem(DSS) isaninteractive
computer-based system intended to help decision makers in using
communicationtechnologies,data,documents,knowledgeand/or
models to identify and solve problems, hence completing the
decisionprocesstaskswiththeoverallobjectiveofmaking
well-informed decisions (Power, 1997). Multiple examples of the
developmentandapplicationofDSSsinagricultureaddressinga varietyof domains, suchaspest management (Periniand Susi, 2004;Riparbellietal.,2008;Callieraetal.,2013),water manage-ment(Fassioetal.,2005;Pallottinoetal.,2003;Giupponi,2007; Acutisetal.,2010),agriculturallandmanagement(Mazzocchietal., 2013) and nutrientmanagement (Djodjicet al., 2002;Forsman etal.,2003;Deetal.,2004),areavailable.AsreviewedbyKarmakar etal.(2007)DSSsformanuremanagementareavailablebutmost
ofthemareaddressedtothenutrientmanagementinthe
agro-nomicplanningwithregardonlytotiming,amountandspreading method(Deetal.,2004;DeandBezuglov,2006).OnlyfewDSSs
con-siderthewhole-farmmanuremanagementfromtheproductionto
thelandapplicationprovidingsupporttowardsthechoiceofthe moresuitableoption.AmongtheseKarmakaretal.(2010) devel-opedaspecificDSSforswinefarmsoftheCanadianPrairesregion:
multiplecombinationsofmanagementoptionscanbeevaluated
consideringdifferentdecisioncriteriasuchasenvironmental, agro-nomic,socialandhealth,greenhousegasemission,andeconomic factors,whilstthesoftwareMLCONE4(Ogilvieetal.,2000)allowsto evaluatemanure-handlingsystemsofagreaternumberoflivestock types(i.e.,swine,dairyandpoultry)anditwasspecificallydesigned forOntarioProvince’sconditions.Similarly,Sørensenetal.(2003)
developedamodeltoevaluatedifferentmanurehandlingsystems forpiganddairyfarms.
TheuseofDSSsconsideringmanuremanagementina
whole-farmperspectivebecomesapriorityinareaswithnutrientsurplus
and where farmers shoulddefine optimalstrategies to reduce
environmental impact at a sustainable cost. In fact, in these
conditionssolutions ofteninclude theimplementationof treat-menttechnologiestoremove nutrient surplusthat entailshigh investmentandoperatingcosts.Agoodexampleofthiscondition isrepresentedbytheplainareaoftheLombardyRegion(northern
Italy)in which the Governmenthave developed regional
legis-lationincludingimplementationoftherequirementsofNitrates (91/676/EEC)andWaterFramework(2000/60/EC)Directivesand ofItalianRegulations,suchastheMinisterialDecreeof19April 1999approvingtheCodeofgoodagriculturalpracticesandthatof 7April2006regardingcriteriaformanuremanagement.Specific
ActionProgrammesfor both nitrateand non-nitratevulnerable
zones (D.g.r. VIII/5868/2007 and D.g.r. IX/2208/2011) together
Programme (RDP) have been implemented to control nutrient
pollutionofwaterfromagriculturalsources.Moreoverfrom2011 is in force the nitratederogation (EC, 2011) for which eligible
farmers who want to obtain its benefit have to respect some
requirementsaboutmanureandlandmanagement.
Thisterritoryinwhichthenitratevulnerablezonesrepresent 62%ofutilisedagriculturalareaischaracterisedbyanintensively managedagriculturewithhighlivestockdensityaccountingfora bigpartoftheItalianlivestock,inparticularmorethan27%of cat-tleand51%ofpigs.Recentstudiesconfirmedthepotentialimpacts oftheagriculturalandlivestockactivities.Fumagallietal.(2011, 2012)highlightedthehighuseofproduction factorssuchasN, fossilenergyandplantprotectionproductstosustainanimaland cropproductions.Peregoetal.(2012)reportedhowtheintensive
maize-basedcroppingsystemsbasedontheuseoforganicand
inorganicfertiliserscoulddeterminehighriskofnitratepollutionas wellasCarozzietal.(2012,2013,2013a)showedhowalternative
low-ammoniaemissiontechniqueshavetobeprescribedduring
manuredistributiononfields.Provolo(2005)showedthenegative
environmentalimpactofsomemanuremanagementsystemsby
mappingsomeindicatorresultssuchasthelivestockmanure pro-duction,theratiobetweennutrientsbroughttothelandandthe uptakeofthecropandtheamountofNappliedperhectare.
The awareness of the environmental concerns related to
livestockactivitieswithwhole-farmperspectiveledtothe devel-opmentofaDSSabletoprovidethestakeholders,suchaspolicy makers,farmers andtheirconsultants,withanassessment tool toevaluatetheintroductionof differentlivestockmanure man-agementsystems.Thedesignandevaluationofdifferentscenarios couldallowtheidentificationofthebestmanagementwhichcould becharacterisedbyavailabletechniquesandtechnologies.
Anintegrateddecisionsupportsystemisherepresentedtobe usedintheLombardyregiontoaddressallthemajorcomponents
ofmanuremanagement(production,storage,treatmentandland
application)foravarietyoflivestocktypes.Itwasdevelopedon thebasisofthepreviousexperiencecarriedoutbyProvolo(2005)
whoevaluateddifferentlivestockmanuremanagements.TheDSS
allowsanintegratedassessmentatfarmandterritorialscaleusing twodifferenttoolsaimedattwodifferentstakeholders.
TheobjectiveofthisworkistopresenttheDSSValorE,which helpsstakeholders(i)tofindthebestoptionforminimisingtherisk ofenvironmentalpollution(mainlyN),(ii)improvingthevalueof manurefromdifferentlivestockinenvironmental,technical, agro-nomicandeconomicterms,(iii)planningmanuretreatmentplants, and(iv)evaluatingtheeffectsofnewtechnologiesonfarm man-agementaswellaschecking,antefactum,possibleimpactsofnew policies.
ValorE:aDSStoenhancelivestockmanuremanagement
ValorE(Valorisation of Effluents) is a user-friendly software developedtocopewithdifferentlivestock(i.e.,cattle,swine, poul-try,sheep,goatsandhorses)andtosuggestandanalysealternative manuremanagementoptionsatfarmandterritorialscale.SuchDSS
consistsofthreemaincomponents:datamanagementsubsystem,
modelmanagementsubsystemanduser-interface.Asimple repre-sentationoftheDSSstructureisreportedinFig.1.Severalexternal databasesaredirectlylinkedandperiodicallyinterrogatedinorder tosupplytheDSSdatabasemanagementsystemwiththerelevant input,whileasoftwaretooldeveloped inthe.NETenvironment
andimplementedusingobjectorientedprogramming(OOP–C#
language),providesthelogictomanagethescenariosimulation linkingagronomicandenvironmentalfarm-scalemodels.Thetwo interfacesallowmanagingthesimulationatfarmandterritorial
Fig.1.SchemashowingthegeneralstructureoftheValorEDSS.
scale respectively.The territorialinterfaceis a web portal
con-nectedtoaWebGIS(geographicalinformationsystem)handling
thespatiallydistributedinputsandoutputsoftheDSS.Allthemaps andtablesproducedbythesoftwareareinItalianlanguagesince anEnglishversionhasnotyetbeenreleased.
Databasesandreferenceinformation
All information needed to run the system are stored on
databasesprovidedbytheLombardyRegionalGovernment.Such
datainclude(i)farmstructure,(ii)meteorologicaldataatdailytime step,and(iii)pedologicalcharacterizationofthewholeregion.
Anotherdatabasecreatedbytheteamgroupcontainsseveral tablesofdefaultdatacalledthereafter“referencetables”. Farmstructure
ThedatabaseoftheAgriculturalInformative Systemof Lom-bardiaRegion(SIARL)containsdatarelatedtothefarmstructure forthewholeregion.Allinformationareperiodicallyupdatedby farmers.Inparticular,farmershavetoprovidedetailsaboutthe regulatorycomplianceonthematterofNmanagement(Provolo, 2005).Thisdatabasecollectsinformationof87%offarmssurveyed by the Italian institute of statistics during the 6th Agricultural
censuslaunchedin 2010.Thedatabaseincludesinformation on
distributionoftheherdaccordingtoanimalagecategories,animals housing,manureandslurrystorageandtreatment.Moreover,land usedataofeverycadastralplotarestoredforeachfarmproviding informationontheareaallocatedtothedifferentcropsoverthe years.
Meteorologicaldatabase
TheLombardia Region hasmadeavailabletwenty-year time
seriesofdailymeteorologicaldatasuchasmaximumandminimum
temperature(◦C)andprecipitation(mm)in14stations represen-tativeoftheregionalclimatezones.
Soildata
Avectorialsoilmapatscale1:50,000isavailable,where1038 soilscapesaredefinedandcharacterisedbyatleastonesoilprofile. Soilphysicalandchemicalproperties,suchastexture,structure, organicmatter,pH,soilcationexchangecapacity,derivedfromfield andlaboratoryanalysisareavailableforeachhorizonofthesoil profiledownto2mdepth.Thesoilsareclassifiedaccordingtothe WRBclassification(FAO,1998).
Technologicalandagronomicmanagementdata
OnlyapartoftheinformationneededtoruntheDSSisdirectly available from the SIARL database (Regione Lombardia, 2010),
therefore another database containing five reference tables of
defaultdatawasproduced.Defaultdataderivedfromexisting lit-erature,expertsknowledgeandfarmers’interviewsare:
-thetechnique,functionalandeconomicfeaturesofavailable tech-nologiesusedforthemanuretreatment;
-theanimalsration forvariouslivestockcategoriesin termsof proteinandphosphorouscontent;
-themaincropsgrownintheregionalarablelandandtherelated
agronomicmanagement, suchassowing andharvestingtime,
organicandmineralNsupply;
-theirrigationtechniques,thefrequencyandthewatervolumes typicalofthedifferentareasoftheregion;
-thecurrent regulationonthematterof (i)Nitrate Vulnerable Zonesdefinition,(ii)allowedtimingofmanureapplication,(iii) restrictionon manurefertilisationin particularareas suchas riparianzones and protected areas,(iv) guidance for manure incorporation(RegioneLombardia,2007).
Fig.2. Schemashowingthesimulatednitrogenflowsatfarmlevel(modifiedfromBerntsenetal.(2003)).
DSSdevelopment
TheDSShastomeetaseriesofrequirementstobeusefulfor differentkindofstakeholders(e.g.,farmersandtheirconsultants, PublicAuthorities,producersorganisations,scientists,etc.)andfor aneasyupdatingandmaintenance.TheterritorialpartoftheDSSis awebportal,whereasthefarmsimulatorcanbeinstalledandrun onanycomputerrunningwindowsXPOSorlaterversionswithout
specifichardwarerequirement.Moreover,thedevelopmentofan
easywayofoperatingwasamainobjective(nomorethan5clicks togettoacompleteanalysisfollowingthesuggestionofthe“three clickrule”foruserfriendlyandmoreimpactfulwebdesign)with reportsimulationresultseitherinmapsandtabularform.
Theintendedpurposeofthesoftwareistosimulateatfarmscale eachstageoflivestockexcretacyclefromproductionbytheherd tothecropNuptakeaswellastheNcycleandlossesoccurringvia leaching,andgaseousemission(volatilisationanddenitrification).
Fig.2showsthesimulatedNflowsatfarmlevel.Thesoftware con-sistsofdifferentmodularcomponentsrelatingaspecificstageof themanureproductionprocess.Eachcomponentallowsfor selec-tionofstrategiestosimulateaspecificprocessandeachmodule resultsrepresenttheinputdataforthesubsequentone(Fig.3).
Excretionmodule
Inordertoevaluatetheimpactofthedifferentlivestockrations onurineandfaecesproducedbycattleandswine,theexcretionof NandPcontentissimulatedasafunctionoffeedintakeandanimal performance.Inthisanalysis,dairycattle,beefandpigsfarmsare consideredasthemainsourceofproductionofslurryinLombardy. Withregardtocattle,themodelallowsestimatingseparately forurineandfaeces,theamountofNandPexcretedby quanti-fyingtheamountofmanure.Instead,theamountofKexcretedis estimatedasafixedpercentageofliveweight,asrecommendedby existinglegislation.Fordairycattle,theexcretioniscomputedbya sub-modelfromthefollowinginputvariables:(i)thebodyweight oflactatingdairycows,drycows,heifersandcalves(ii)themilk productionlevel,(iii)themilkfatandproteincontent,(iv)thedry matterintake,(v)andtheproteincontentoffeed.Inparticular,the drymatterintakeiscalculatedbyusingtheequationproposedby theNationalResearchCouncilofUSA(2001).Themodelproduces thefollowingoutputdata:(i)theexcretedproductsasfreshmatter (kgFMd−1),calculatedaccordingtoNennichetal.(2005),(ii)urine anditsNcontent(kgd−1),calculatedaccordingtoFoxetal.(2004),
(iii)theamountoffaeces,calculatedasdifferencebetweenthetotal excretedproductsandurine(kgd−1),(iv)theNfaecescontentand, (v)themilkNcontent(kgd−1).
Themodeldevelopedforpigsestimatetheexcretedamountof N,Pand Kaccordingtoseveralstudies(Pomaretal.,1991a,b,c; LeBellegoetal.,2001;vanMilgenandNoblet,2003).In particu-lar,theestimateiscarriedoutforphysiologicalstagesofgrowth andproductionoftheanimal.Themodelquantifiesthefeedintake basedontheanimalgrowth(kgd−1)andfeedconversionefficiency fortheconsideredgrowingphasesandfornumberoffarrowsand litterssizeforthesow.Thenitrogen,PandKintakes(kgd−1)are estimatedbasedonfeedintake(kgd−1)anddietcontents,while excretionsaredeterminedfromdietandproteindigestibilityand mineralabsorption(%)fortheconsideredphysiologicalstages.The modelallowstocalculatethemanureproduction(i.e.,dry mat-terandvolume)andtheN,PandKexcretioninfaecesandurine. Forotheranimalspeciessuchaspoultry,sheep,goatsandhorses,
Fig.3. ModularcomponentoftheDSSrelatingtoeachspecificstageofthemanure productionprocesses.Eachmoduleimplementsitsspecificsimulationmodel.
recommendedbyexistinglegislation(RegioneLombardia,2007).
Housing,treatmentandstoragemodules
Slurryissubjectedtochemicalandphysicalmodificationswith relativegaseouslossestotheatmosphere.Foreach stageofthe storageandtreatmentprocessthemodulesimulatestheamountof slurrymassanditsN,PandKcontenttogetherwiththeinvestment andoperatingnetcostsofanyjointproduction(energy,compost, fertilisers,etc.).Moreover,itallowstheassessmentofthefeasibility andsuitabilityofalternativetechniquesinplantmanagement.
Theinputdataoftheslurrystorageandtreatmentmoduleare: (i)thechemicalandphysicalcompositionoftheexcretedproducts expressedaskgof dry matter,kgFM,faeces TKN (Total Kjeld-halnitrogen)content,urineTKNandP2O5contentinfaecesand urine),(ii)thelitterfractionofthemanure,and(iii)therainfall.The effectofthetypologyoflivestockhousingandtheeffectofdifferent typesofslurrystoragearesimulatedaccordingtoIPCC(2006)and
EEA(2009),consideringalsotheexperienceofAmonetal.(2006)
andWebband Misselbrook(2004).A widerange oftreatments is considered in the module: solid–liquid separation (Dinuccio etal.,2008;Fangueiroetal.,2008;Cocoloetal.,2012),anaerobic digestionwithbiogasandenergyproduction(Amonetal.,2007; Biswasetal.,2007),ammoniastripping(BonmatìandFlotats,2003), nitrificationanddenitrification(Rousseauetal.,2008),aerobic sta-bilisation(Loyonetal.,2006;Belineetal.,2007)andcomposting (Paillatetal.,2005;Szantoetal.,2007).
Theslurrymodulecalculates:(i)thefinalvolumeofthestored slurry,(ii)thefinalchemicalandphysicalcomposition,(iii)thesolid andliquidfraction,(iv)thegaseouslossestotheatmosphere,and (v)thepossibleproductionofbiogasfortheanaerobicdigestion plantsandotherjointproductsoftreatments.
Economicaspectsareinvolvedintheestimationoftheweight ofmanuremanagementoptionsonfarmincome,sinceithasbeen recognisedtheimportanceofcrosscomplianceontheeconomyof agriculturalsector(Bezlepkinaetal.,2008;DeRoestetal.,2011). Foreachphaseofmanagingslurryandmanure(housingtype, treat-ments,storage,distribution),themodulecalculatesinvestmentand
operatingcosts (Berglund and Börjesson,2006;Gourmelen and
Rieu,2006).Forthehousingsystems,whiletheinvestmentcost isrelatedtothecostofconstruction(e.g.,rawmaterial,facilities) theoperatingcostdependsonbeddingmaterials,energy consump-tionandcostoffacilitiesmaintenanceandlabour.Inthecaseof manurestorageanditscoverandofplantformanuretreatment