Reducing Environmental Risks from Belt and Road Initiative …documents.worldbank.org/curated/en/700631548446492003/... · 2019-01-25 · Policy Research Working Paper 8718 Reducing

Policy Research Working Paper 8718

Reducing Environmental Risks from Belt and Road Initiative Investments

in Transportation InfrastructureElizabeth LososAlexander PfaffLydia OlanderSara MasonSeth Morgan

Macroeconomics, Trade and Investment Global Practice January 2019

WPS8718P

ublic

Dis

clos

ure

Aut

horiz

edP

ublic

Dis

clos

ure

Aut

horiz

edP

ublic

Dis

clos

ure

Aut

horiz

edP

ublic

Dis

clos

ure

Aut

horiz

Produced by the Research Support Team

Abstract

The Policy Research Working Paper Series disseminates the findings of work in progress to encourage the exchange of ideas about development issues. An objective of the series is to get the findings out quickly, even if the presentations are less than fully polished. The papers carry the names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the views of the International Bank for Reconstruction and Development/World Bank and its affiliated organizations, or those of the Executive Directors of the World Bank or the governments they represent.

Policy Research Working Paper 8718

The Belt and Road Initiative, due to its diverse and extensive infrastructure investments, poses a wide range of environ-mental risks. Some projects have easily identifiable and measurable impacts, such as energy projects’ greenhouse gas emissions. Others, such as transportation infrastructure, due to their vast geographic reach, generate more complex and potentially more extensive environmental risks. The proposed Belt and Road Initiative rail and road invest-ments have stimulated concerns because of the history of significant negative environmental impacts from large-scale

transportation projects across the globe. This paper studies environmental risks—direct and indirect—from Belt and Road Initiative transportation projects and the mitigation strategies and policies to address them. The paper concludes with a recommendation on how to take advantage of the scale of the Belt and Road Initiative to address these con-cerns in a way not typically available to stand-alone projects. In short, this scale motivates and permits early integrated development and conservation planning.

This paper is a product of the Macroeconomics, Trade and Investment Global Practice. It is part of a larger effort by the World Bank to provide open access to its research and make a contribution to development policy discussions around the world. Policy Research Working Papers are also posted on the Web at http://www.worldbank.org/research. The authors may be contacted at [emailprotected].

ReducingEnvironmentalRisksfromBeltandRoadInitiativeInvestmentsinTransportation

Infrastructure1

ElizabethLosos,AlexanderPfaff,LydiaOlander,SaraMason,andSethMorgan

DukeUniversity

1WethankMichaelToman,GlennMorgan,Sven‐UweMueller,GrahamWatkins,DanielTaras,BekeleDebele,MicheleRuta,andseminarparticipantsattheWorldBankfortheirhelpfulsuggestions.TristanReedandSashaTrubetskoywereespeciallygenerousinsharingtheirtransportinfrastructuredataandknowledge.Allerrorsremainourown.

INTRODUCTIONThelocationofRomanroadsbuiltalmosttwomillenniaagostillshapesmodernlandscapesandlocal

economies(Dalgaardetal.2018).Thatscholarscanstilldetecttheeffectofthistransportationfootprint

istestamenttothetransformationalandenduringinfluenceoftransportationnetworksonpeople,the

economy,andtheenvironment.Environmentallyspeaking,networksofroadsandrailwaysoftenhave

beenidentifiedasoneofthemostsignificantanthropogenicinterventionsoftheearth'secosystems(W.

Laurance2017;W.F.Lauranceetal.2015;W.F.LauranceandArrea2017;PoppandBoyle2017).

Historically,large‐scaletransportinfrastructureprojectshavehadsignificantnegativeenvironmental

impactsacrosstheglobe.GiventheenormityoftheBeltandRoadInitiative(BRI),whatenvironmental

risksdotheproposedBRItransportationprojectspose?InPart1ofthisworkingpaper,wereviewmany

ofthemostsignificantrisksfacingBRItransportationprojects.Weconsiderboththedirecteffectsfrom

roadsandrail–typicallyimmediateandlocalized–andindirecteffectsfromopeningupnewlandscape

frontiers,shiftinghumanpopulationsandmarkets,alteringdemandsfortransportation,andbalancing

energyefficiencywithinduceddemands.Next,inPart2,wereviewpotentialenvironmentalmitigation

strategiesthatcouldimprovetheenvironmentalperformanceofBRItransportationprojects.Finally,in

Part3wewilldescribepoliciesthatcanshapetheimplementationandenforcementofmitigation.

Thoughnotaddressedinthispaper,BRIprojectsinothersectorsalsohaveenvironmentalimpacts.In

particular,BRIenergyprojectssuchasfossilfuelandrenewableenergyprojectscanhaveaconsiderable

effectongreenhousegasemissions(GHG)andotherpollutants.Seriousconcernshavebeenraisedthat

thepromotionofBRIfossilfuelinvestments(especiallycoalplants)couldlockhostcountriesintofossil

fueldependencyforthecomingdecadesandhamperthemfromreachingtheirnationallydetermined

contributioncarbontargetsasestablishedundertheParisAgreementonClimateChange(Gallagherand

Qi2018;Sausmikatetal.2017;Zhouetal.2018).WhileBRItransportprojects–thefocusofthispaper–

havelessdirectimpactonGHGemissions,thevastgeographicreachofroadandrailnetworksresultsin

amorecomplexsetofenvironmentalrisks,whichcanincludesignificantimpactsuponGHGemissions.

Transportationinvestmentsalsoposeserioussocialrisks,manycloselyparallelingenvironmentalrisks.

Socialriskscanincludeeconomicandphysicaldisplacementfromtakingoflandandassets;impactson

disadvantaged,vulnerable,orgroupswithspecialrights;impactsonqualityoflife;degradationoflabor

andworkingconditions;andcommunityhealth,safetyandsecurity.Thesecanbeprofoundyet–like

environmentalrisks–theyvarybycontextsuchassettlementdensity,workingconditions,laborforce,

workerprotections,planningprovisions,andotherconditions.Socialissuesarenotaddressedhereyetit

shouldbenotedthatthereisacomplementaritybetweensocialandenvironmentalrisksandmitigation

strategies.Forinstance,ourrecommendationsareequallyviableforaddressingmanyofthesocialrisks.

Throughout,weprovideexamplesdrawnfromtheBRItransportationprojectswithintheBRICorridors,

illustratedintheappendices.Appendix1isanoverviewofenvironmentalrisksatthescaleoftheentire

BRI.Appendices2‐9focusonanumberofaspectsfromonespecificeconomiccorridor,theChina‐

IndochinaPeninsulaEconomicCorridor(CICPEC)–specificallythethreealternativetransportation

routesbetweenKunming,ChinaandBangkok,Thailand–toillustratemoregeneralpointsmade

throughoutthepaper.

Insum,ourworkingpaperidentifiesseriousenvironmentalrisksfromtheBRItransportationprojects

andthepotentialformitigationviaCorridorSEAplanningdoneasearlyaspossible.TheCICPECexample

willprovideamodelofhowaBRICorridorStrategicEnvironmentalAssessmentcouldmeettheseneeds.

A.DataBasis

OurworkingpaperandespeciallytheCICPECexamplesdrawontheWorldBankdatasetfrom

ReedandTrubetskoy(2018)ofBRIrailwayandroadprojects.Theseauthorshavegeolocatedalmost

30,000kmofneworupgradedrailwaysandroadsthathavebeenconstructedorareintheprocessof

beingconstructedsince2013andalmost15,000kmmorearecurrentlyintheplanningstages(seeTable

1;ReedandTrubetskoy2018).Themajorityoftheseinvestmentshavebeenorwillbeinrails,notroads:

morethan60%(bydistance)ofthoserecentlyconstructedorunderconstructionarerailprojectsand

90%oftheplannedBRIprojectsarerailways(ReedandTrubetskoy2018).2Whenpossible,wewill

distinguishbetweentransportcategories(newversusupgradedinfrastructure;trainversusrail;high

capacityversusconventionalrail;dividedhighwayversusundividedhighway).Whereresearchislacking,

wemustspeculatebasedonpriorsimilaritiesanddifferencesamongthesecategoriesinotherlocations.

Table1.BRIroadandrailprojects,completed,on‐going,andplanned

BRIProjectsKmunder

constructionoralreadybuilt

Additionalkmplanned,notbuilt

Kmtotal

km % km % km %

NewUndividedHighways

5,547 19%

378 3% 5,926 14%

NewDividedHighways

809 3%‐ 0% 809 2%

UpgradetoDividedHighways

4,723 16%

841 6% 5,564 13%

TotalRoadImprovement 11,079 1,220 12,299

NewConventionalRailways 8,649 30% 4,030 27% 12,680 29%

NewHighCapacityRailways

2,835 10%

3,975 27% 6,809 16%

UpgradetoHighCapacityRailways

6,228 22%

5,768 38% 11,997 27%

TotalRailImprovements 17,712 13,774 31,486

2BecausethereisnoofficiallistofBRIprojectspublishedbytheGovernmentofChina,analystshaveassembledtheirownlistsoftransport.WeutilizetheBRIprojectsidentifiedinReedandTrubetskoy(2018).Oneofthecriteriausedforinclusionisthattheendpointsoftheroadorrailinvestmentsinquestionarecitieswithatleast300,000inhabitants,consistentwiththeBRIvisionoftradecorridorsthatwouldconnectmajorhubsthroughoutEurasia.Notsurprisingly,then,theprojectsintheReedandTrubetskoy(2018)datasetaremostlylargehighwaysandhigh‐speedrailcorridors.Otheranalystshavechosenbroaderdefinitionsconcerning“BRItransportationprojects”(see,forinstance,analysesinReconnectingAsiaoftheCenterfromStrategicandInternationalStudies(Hillman2018)orbyAiddata(BenYishayetal.2016)).ThosebroaderBRIclassificationstendtocapturemoresmallerprojects.

Total

28,791 100%

14,993 100% 43,785 100%

Source:Reed,T.andS.Trubetskoy(2018)TheBeltandRoadInitiativeandtheValueofUrbanLand.WorldBankmimeo.

B.Road&RailTypologiesWesubdivideeachtypeoftransportinfrastructurefollowingReedandTrubetskoy(2018)toconsider:

undividedroads,dividedroads,conventionalrail,andhigh‐capacityrail.Thefourcategoriesare

describedbelowwithparticularreferencetoproximitiesofbothpopulationcentersandintactfrontier

landscapes–twofactorsthatareespeciallyimportantforunderstandinginvestments’environmental

risks.

HighCapacityRail:High‐capacityrailisdualtrackand/orelectrifiedrailthattransportfreightand/or

passengers,includinghigh‐speedrail(HSR).HSRaredefinedashighcapacityandfrequencyrailway

servicesachievinganaveragespeedofover200km/hr(Givoni2006).TheHSRsystemshave

restrictedgeographicaccessbecausepassengersand/orfreightarelimitedtoentry/exitatveryfew

stations.Further,thoseHSRstationsareusuallysitedeitherinoradjacenttopre‐existingurban

centers,thuslimitingnewaccesstofrontierlandscapes.3

ConventionalRail:Conventionalrailtravelisatspeedslessthan200km/hrandistypicallyfueledby

dieselorotherfossilfuels.Conventionalrailstopsatmanymorestationsthanthehigh‐speedrail,

thuscreatingmanymoreentrypointstoaccessfrontierlandscapes.Smallerstationsaresitedinless

populatedareas,potentiallyincreasingaccesstointactfrontierlandscapes.

DividedHighway:These“high‐quality”roads,classifiedasfour‐lanehighwaysorlarger,tendtobe

sitedbetweenlargeurbanareas,thoughtheymaypassthroughlessdevelopedareas.Theyare

restrictedtohighspeedtraffic.Thedividedhighwaysoftenhavelimitedorcontrolledaccess.The

degreeofaccessdependsonthefrequencyofoff‐ramps.Accessmayalsoberestrictedthrough

tolls.

UndividedHighway:Theseare“low‐quality”roads,i.e.,roadswhichhavenomorethantwolanes

butareserviceablefortransportingfreight.Theyconnectpopulationcentersofvariedsizesandare

lesslikelytofeaturerestrictionsonaccessthroughrampsand/ortolls.BRIundividedhighwaysmay

alsobeconstructedtoconnecttoothertransporthubs(forexample,tolargerBRIprojects)suchas

dryports,airports,ortrainstations.Becauseundividedhighwaystypicallyhaveunrestrictedaccess

forentryandexit,newroadsmayexpandgatewaysintoundevelopedfrontiers.

Thereareseveraladditionalcategoriesofminorroads–roadssmallerthanundividedhighways–that

alsohaveenvironmentalrisksbuttypicallyarenotafocusofBRIprojects.Thesecanrangefromsmaller

two‐lanehighwaystounpavedseasonalorloggingroads.MinorroadscouldbeincludedwithintheBRI

asaformofaccesstoprimaryBRIprojects.Additionally,unplannedspurroadsthatdivergefromlarger

roads(typicallyfromundividedhighways,duetotheirunrestrictedaccess)mayspontaneouslyresult

fromplannedBRIprojects.Spurroadsmightpenetrateintofrontierlandscapes,yielding,forinstance,a

gridof"fishbonedevelopment"ashasbeenextensivelydocumentedintheAmazon(Barberetal.2014).

3HSRisoftenflankedbyroadsformaintenanceandemergenciesthatmayprovidesomeaccessbetweenstations.

Part1:EnvironmentalRisksfromBRIInvestmentsinTransport

Transportationinfrastructurecanposemanyriskstotheenvironment.Inthissectionwe

distinguishdirecteffects–abiotic,ecosystems,wildlife–fromtheindirecteffectsinducedbychangesin

transportcostsandlanduse.Whiletheformertendtobelocalizedandunambiguouslylinkedtoaroad

orrailproject,thelatteroftenhavemorecomplexconnectionsbutpotentiallymorepervasiveimpacts.

Themagnitudesofvariouseffectsdifferbycontext.AsillustratedinAppendix1(ForestCoverChange

AcrosstheEntireBeltandRoad),deforestation–arelativelystraightforwardproxyforenvironmental

risks–hasshownheterogenousresponsestotransportationinvestments,dependinguponboththe

ecologicalanddevelopmentsettingsinwhichinvestmentsoccur.Cumulativerisksofmultiplefactors

areevenmorechallengingtoassessandpredict,asthesefactorsmayevolveandalsomayinteract.

Environmentalrisksaretypicallydividedbetweenthoseincurredinthedesignandconstructionphases

ofanyinfrastructureprojectandthoseexperiencedduringtheoperations.Thisworkingpaperwillonly

reviewthelatterbecausetheformertendtobeextremelysitespecificandcanbeconfidentlyidentified

duringaprojectfeasibilitystudyorenvironmentalimpactassessmentprocess(WorldBank2012).

I.DIRECTEFFECTS

A.AbioticEffects

i.Edaphic,Topographic,andHydrologicImpactsOnceoperational,roadsmodifytheirphysicalenvironmentbyalterationsofhydrologicsystems,

sedimenterosiondynamics,anddebrisdepositiondynamics.Theextentandintensityoftheseabiotic

effectscanvaryasaresultofthepositionoftheroadrelativetoslope,nearbywaterwaysanddrainage,

surroundinglandcover,andprevailingwinds(Coffin2007;Forman1998;Sidle,Ghestem,andStokes

2014).Mostabioticfactorshavelocalizedeffectsimmediatelyaroundtheroadandroadside.These

impactscanbesevere,asinthecaseofdestructivelandslidesthatdamagenearbycommunitiesand

ecosystems.Someabioticeffects,however,permeateacrossawiderlandscape.Forexample,poorly

constructedroadsonmountainousterraincangenerateincreasedsedimentationinriversandstreams,

creatinglong‐termeffectsondownstreamaquaticcommunitiesandproducingwidespreadflooding

risks(Forman1998;Renaud,Sudmeier‐Rieux,andEstrella2013;Sidle,Ghestem,andStokes2014).

BecausesomanyoftheBRIEconomicCorridorspassthroughsteepterrain,BRItransportationprojects

areespeciallyvulnerabletosuchrisks.Theconsequencescanbesevere:Twenty‐fivemillioninhabitants

livingdownslopefromtwoproposedBRIroadprojectsinMyanmar,forexample,arepotentiallyatrisk

duetovulnerabilitytoincreasedsedimentationandflooding(Helsingenetal.2018).

Thereislittleenvironmentalresearch(beyondengineeringfeasibilityplans)ontheeffectsofrailwayson

hydrologicsystems,sedimenterosion,anddebrisdeposition.However,itisexpectedthattheywillbe

similartoroads,withimpactsdependingontopography,hydrology,andclimate.Itispossiblethatrisks

relatedtouneventopographyandsteepslopesmaybemoresevereforrailprojects(especiallyHSRs),

whichforengineeringpurposesareconstrainedtorelativelystraighttrajectories.Lessabletoadjustto

theterrain,HSRmayfacegreatercomplicationsthanroads,andmayrequiremoretunnelsandbridges.

ii.PollutionInthecontextoftheBeltandRoadInitiative,neworupgradedhighwaysorrailwillbebuiltprimarily

withtheintentionofincreasingfreighttraffic,withthesecondaryeffectofpotentiallyincreasing

passengertraffic.Intermsofroads,itisexpectedthatincreasedhigh‐speedfreighttrucktrafficonnew

andupgradedBRIhighwayswillgeneratesignificantnoiseandairpollutionincludinggreenhousegases

(GHGs).Waterpollutionwillalsobecreatedfrommetalcorrosion,tirewear,exhaustemission,trash

andotherpollutantsthatarewashedintowaterways,especiallyduringheavyrainstormsorfloodswhen

drainagecapabilitiesofditchesandsoilareexceeded(EconomicCommissionforEurope1975).

Measuringpollutionrequiresdataonvolumeandcompositionoftrafficaswellastheweight,speed,

andefficiencyofthecarsandtrucks(JanicandJovanović2012).Assessingtheincreaseinpollutionfrom

theneworimprovedBRIismorecomplicated,however,asitalsorequiresidentifyinginduceddemand

forhighwaytravelandtransport,displacementoftrafficfromotherroutes,andsubstitutionbetween

differentmodesoftransportation.TheseissueswillbediscussedbelowintheIndirectEffectssection.

WithregardstopollutionfromBRIrailprojects,conventionalrailtypicallyhashigheremissionsper

passengerorpertonfreightthanhigh‐speedrailbecausethelatterispredominantlypoweredby

electricityratherthandiesel.HSRemissionsaredirectlyrelatedtothesourcesusedtogenerate

electricityandtheirenergyconsumption.Ifsignificantcoalisusedasthesourceoftheelectricity,then

HRScancontributesignificantlytolocalairpollutionaroundthecoalplantsthroughtheproductionof

SO2.However,ingeneralbothconventionalandHSRcontributesignificantlyfewerGHGsperpassenger

orfreightthanotherformsoftransportationsuchasroad,airtransport,andwaterways(Givoni2006;X.

Li,Fan,andWu2017;NationalRail2009;VianaandCenamo2008).Conventionalfreightrailhastwoto

fivetimeshigherfuelefficiencythantrucktransport(Dorsey,Olsson,andRew2015),aswellasuses

21%lessenergyperpassengermile(FederalTransitAdministration2009).InEuropealongtheMega‐

TransEuropeantransportcorridor,thesubstitutionfromtruckstofreighttrainsresultedin64‐70%

savingsinGHGs(JanicandJovanović2012).Trainareevenmoreefficientthanairplanes.Ascompared

toairtransport,HSRhaveloweremissionsofCO2andNOx.Moreover,NOxemissionsathighaltitude

affectclimatechangemuchmorethanground‐levelemissions,makingairtransportevenmore

damaging(Givoni2006).Aswithroads,theultimatepollutionimpactfromaBRIrailprojectwilldepend

notonlyontherelativefuelefficiencyandpollutioncontrols,butalsotheinducedtrafficdemandthat

couldincreaseoverallpollution,whichwillbeaddressedintheIndirectEffectssectionbelow.

Noisepollutionfromtrains(freightandpassenger)hasalsobeenrecognizedasharmfultothehealthof

humansandwildlifepopulations.Forfreighttrains,thelonger,heavier,andfastertrainsgeneratehigher

levelofnoisebutexposuresareshorterduetohigherspeeds(Givoni2006;JanicandJovanović2012).

B.EcosystemEffects

i.HabitatLossThemostbasicenvironmentalimpactfromaroadorrailisthedestructionofhabitattocreatea

transportationcorridor.Thedirectlossinhabitatduetothefootprintofroadsandroadsideistypically

1‐2%ofthelandcoverofmostcountries,withagreaterproportionnearurbancentersandthelowest

percentageinthelessdevelopedlandscapefrontier(Forman1998).Habitatlossiscloselycorrelated

withmanyaccompanyingenvironmentalrisksthatarerelatedtolossofecosystemservicessuchas

biodiversityhabitats,carbonstorageandsequestration,waterprovisionandquality,soilstabilization

anderosionprotection(A.Dobsonetal.2006;Song,Huang,andTownshend2014).(SeeAppendix3:

EnvironmentalIndicator:EcosystemEffectsasMeasuredbyForestCoverChangeinCICPEC.)

Roadsareconsideredoneoftheleadingproximatecausesofhabitatdestruction,especiallytropical

deforestation.4Forexample,ameta‐analysisbyGeistandLambin(2001)oftheexistingliteratureonthe

causalfactorsfortropicaldeforestationfoundthattheextensionofroadinfrastructurewasaproximate

causeofdeforestationin61%ofthecasesstudied(93of152)whilerailexpansionwasaproximate

causein11%(17or152).Suchevidencehasledtothedevelopmentofarelativelynewdiscipline,“road

ecology,”whichaddressessolelytheimpactsofroadconstructionandoperationsandtheirmitigation

(Coffin2007;Forman1998;Formanetal.2003;vanderRee,Smith,andGrilo2015).Morerecently,

“railwayecology”hasalsobeenproposed(PoppandBoyle2017;Wingardetal.2014),althoughthe

generalassumptionhasbeenthattheenvironmentalimpactsfromrail,whilesimilarinnature,areless

severeinintensitytothosefromroads.Perhapsforthisreason,thegreatmajorityoftheresearchon

environmentalimpactshasfocusedonroadsratherthanrailways.Forexample,asurveyoftransport‐

relatedimpactsonwildlifefoundthat94%ofpeer‐reviewedarticleswereaboutimpactsofroads,with

only6%onrailimpacts(PoppandBoyle2017).

ii.EdgeEffectsThe"roadeffect"(alsoknownasthe“edgeeffect”)istheareaoverwhichtheecologicaleffects

ofaroadandtrafficextendintotheadjacentlandscapeduetohabitatdisturbanceeffects(vanderRee,

Smith,andGrilo2015).Thisisamoresubtlebutpervasiveecosystemeffectofaroadthatextendswell

beyondtheinitiallossinhabitatfromthetransportationcorridor.Thenewlycreatededgesalonga

roadsideallowpenetrationoflight,wind,andchemicalpollutionandmodifymicroclimaticconditions.

Suchalterationsaffectthedistributionandabundanceofplantandanimalspecies.Specialized“habitat‐

interior”speciesofplantsandanimalsareoftenoutcompetedby“edge‐adapted”generalistorweedy

species(Bruschietal.2015).Theedgeeffectcanalsoleadtocascadingabioticeffects.Forexample,

becausetheedgesaredesiccatedfromlightandwindexposure,theyareespeciallypronetoforestfires.

IntheAmazon,edgesareparticularlyvulnerabletofiresinitiatedinadjacentburnedpasturesand/or

selectiveloggedforestswiththeirbuilt‐upfuelload.Oncetheforestfireentersalongtheroadside,

forestfirescanburndeepintotheinterioroftheforest(CochraneandLaurance2002).

Thepenetrationoftheroad‐effectzoneintotheadjacentlandscapeisdeterminedbythecharacteristics

ofthe(i)road(dividedversusundivided,pavedversusunpaved,elevationrelativetoadjacent

landscape);(ii)traffic(vehicletype,volume,speed);(iii)adjacentlandscape(topography,hydrography,

vegetationtype,habitatquality);(iv)prevailingwindspeedanddirection;and(v)speciestraitsandtheir

sensitivitytotheimpact(R.vanderRee,Smith,andGrilo2015).Theroadeffectzonecanbefeltasfar

as1500mfromahighway(Bruschietal.2015).Itisestimatedthat15‐20%oftheearth’slandcoveris

influencedbythisroadeffect(Forman1998).GiventhatmostBRIroadprojectsarerelativelylarge,we

canestimatethatthe“roadeffect”willextendonaverageatleastakilometerintotheadjacent

roadsidehabitat(Benítez‐López,Alkemade,andVerweij2010;Ibischetal.2016).TheChina‐Indochina

PeninsulaEconomicCorridorintropicalSoutheastAsiaisespeciallyvulnerabletoedgeeffectsbecause

researchhasshownthatthiseffectisespeciallypronouncedintropicalecosystems(Goosem2015).

4ItshouldbenotedthatmuchoftheresearchfocusedonroadsanddeforestationinLatinAmerica,especiallytheAmazon,andthatmorerecentrigorousresearchinAsiaalsoverystronglysupportsthatthecontextsarecritical.

Oftenrailwaysarelumpedintothesamecategoryasroadsandotherlinearinfrastructurefortheir

contributiontotransforminglandscapes(Bruschietal.2015).Themagnitudesofedgeeffectsfrom

railwayclearingsarenotwellunderstoodbecausethefieldofrailwayecologyisstillnascent(Poppand

Boyle2017).However,thereisstrongreasontobelievethattheecosystemseffectsforBRIrailprojects

wouldbe,onaverage,lessthanthosefromBRIhighwayprojects.Theedgeeffect–thatis,theareaover

whichtheecologicaleffectsofrailextendintotheadjacentlandscape–islikelytobelessseverefor

severalreasons:First,thewidthofclearingisusuallynarrowerforrailways.Thelossofhabitatfromthe

railright‐of‐waycorridor–typically15moneithersideofthetracks–isonaveragesmallerthanthe

corridordestroyedforroadsandroadsides(andsubstantiallysmallerthanthatofdividedfour‐lane

highways)(Dorsey,Olsson,andRew2015).Morenarrowcorridorsleadtolesschangeinlightexposure,

windpatterns,andmicroclimates,andthuslessopportunityforinvasionbyexoticandweedyspecies.

Second,traintrafficislessfrequentthancarandtrucktraffic,thuscreatinglessfrequentthoughlouder

behavioraldisruptionsofwildlife.

iii.HabitatFragmentationTheedgeeffectisexacerbatedbyanotherattributeoftheroadnetwork:fragmentationofthe

landscape.Roadsoftenfragmentlargehabitatexpansesintosmallerpatches,leadingtodramatic

landscapetransformationandlossoftheabilitytosupporthealthyecosystems,populationsofplants

andanimals,andotherecosystemservices(Bruschietal.2015;Ibischetal.2016;Potapovetal.2017).A

reviewofalmostfourdecadesoffragmentationexperimentsaroundtheworldfoundthathabitat

fragmentationreducesbiodiversityby13to75%,decreasesbiomassandcarbonstorage,andalters

nutrientcycles.Theseimpactsweremostsevereinthesmallestandmostisolatedfragmentsand

continuedtogrowinmagnitudeovertime(Haddadetal.2015).

Theedgeeffect–describedabove–furtherexacerbatestheimpactofhabitatfragmentation.The

smallestandmostirregularly‐shapedpatcheshavethelargestratioofperimetertovolumeand

consequentlythegreatestedgeeffects.

FromaBRIperspective,itisimportanttodistinguishtheimpactoffragmentationfordifferentclassesof

roads,sinceapproximatelyhalftheBRIroadprojectsarelargedividedhighwaysandmostoftherestare

relativelylarge,pavedtwo‐laneroads.Manyoftheseroadorrailprojectsrepresentupgradesrather

thannewtransportationcorridors,somaynotbefragmentingadditionalhabitat.Bycomparison,much

ofthefragmentationresearchreportedintheliteraturehasfocusedonsmallerroads–oftenbuiltfor

loggingaccess–thatabutintactfrontierlandscapesandareaswithlittlepriordevelopment.Mostbut

notallofthisresearchhastakenplaceinLatinAmerica(ChomitzandGray1996;Pfaff1999;Deininger

andMinten2002;Barberetal.2014).Thefewstudiesoffragmentationthatcomparedifferentsize

classesofroadshavefoundthatthesmallerroadsareresponsibleformostofthefragmentation.For

example,astudyoftheimpactsoftheexpansionofroadnetworkfrom1970to2008inXishuangbanna

inYunnanProvince,Chinafoundthatincreasesinroaddensityfortheminorroadsresultedin

significantlygreaterlevelsoflandscapefragmentationthanforthelargerroadsandhighways(S.Liuet

al.2011).WhilethiswouldimplythatthelargeBRIhighwaysdonotcreateasmuchconcernabout

forestfragmentation,intheIndirectEffectssectionbelowwewilldiscusstheconditionsunderwhich

BRIhighwayprojectsmayleadtoaccesstoundevelopedlandscapesandfurthergenerationofadditional

roadsandthusfragmentation.

Itislikelythatraillinesresultinlesshabitatfragmentationduetothefactthat,forengineering

purposes,raillinesareconstrainedtofollowstraighterpathways–especiallyforhigh‐speedrail–and

thuscreatefewerirregularly‐shapedpatches.Perhapsmostsignificantly,railwaysallowgreatercontrol

ofaccesstotheadjacentlandbylimitingtheconstructionofnewsecondaryroads,whichdrive

fragmentation(VianaandCenamo2008),aswillbediscussedintheIndirectEffectssectionbelow.

iv.IntactFrontierLandscapes

Acomplementaryconcernisthat,asmoreofthelandscapeispartitionedintosmallerandsmaller

fragmentsbytheroadnetwork,fewerlarge,undisturbedroadlesstractsremain.Astheselargetracts

diminishinnumberandsize,theecologicalimportanceoftheremainingtractsincreasesbecausethey

createcriticalrefugiaformanyplantandanimalspeciesthatcannotsurviveinsmallerfragments.They

alsoprovidesignificantecosystemservicessuchascarbonstorageandsequestration,climate

stabilization,waterprovision,indigenousculture,andthemaintenanceofhumanhealth(Ibischetal.

2016;Selvaetal.2015;Watsonetal.2018).Theselargeareasofundisturbedwildernessorfrontier

landscapescanbemeasuredas“intactforestlandscapes”(IFL),thatis,connectedmosaicsofforestand

naturallytreelessecosystemswithnoremotelydetectedsignsofhumanactivityandaminimumareaof

500km2.Asof2013,ILFsrepresentedlessthan21%ofonceforestedlandcoverglobally.TropicalSouth

AmericaandtheborealregionscontainedthegreatestpercentageoftheIFLs.(SeeFigure1.)In

SoutheastAsiaonly8%ofIFLsstillremainwhilethenorthernEurasia/temperateandsouthernboreal

regioncontainsjust9%(Figure1;Potapovetal.2017).BecauseoftheoutsizedrolethatIFLsservein

protectingbiodiversityandgeneratingecosystemservices,encroachmentonIFLareasbytransportation

projectscreatesstraightforwardriskstotheseenvironmentallycriticalareas.

AlmostnoneoftheBRIprojectsidentifiedbyReedandTrubetskoy(2018)crossorevenapproachintact

frontierlandscapes,exceptforaproposedroadandarailprojectsinthenorthernreachesoftheChina‐

Figure1.IntactFrontierLandscapeextentfortheyear2013,IFLareareductionfrom

2000to2013,andregionalboundaries.FromPotapovetal.(2017)

MongoliaRussiaEconomicCorridor(CMREC)andseveralsegmentsofprojectswithinCICPECSoutheast

Asiapass.Thismaynotbesurprising,asBRIprojectsaredesignedtoconnectpopulationcenters.Given

howfewIFLsremainintheworld,itisespeciallycriticaltosafeguardtheintegrityofthefewremaining

IFLsinthosetwoBRIEconomicCorridors.(SeeAppendix4:EnvironmentalIndicator:EcosystemEffects

asMeasuredbyIntactFrontierLandscapes.)ItshouldbenotedthattheIceSilkRoadinthearcticregion,

however,traversesaregionwithamuchgreaterproportionoflargetractswithinwhichIFLsarestill

relativelyextensiveinnorthernEurasiaandtheNorthernborealregion(45%)(Potapovetal.2017).

C.WildlifeEffects

i.HabitatandBehavioralDisruptionSincethe1970s,researchershavefocusedoneffectsofroadsonwildlifemorethananyotherrisks.

Extensiveresearchhasbeencarriedoutconcerningtheriskofroadstomammals,birds,herpetofauna,

pollinators,non‐pollinatinginsectsandotherinvertebrates(Ahmedetal.2014;Coffin2007;Poppand

Boyle2017;R.vanderRee,Smith,andGrilo2015).Forsomeanimalspecies,road‐relatedmortality

ranksasoneofthecauses,ifnotthemostsignificantcause,ofspeciesendangerment(SeilerandHelldin

2006).

Wildlifeimpactsaremediatedthrougharangeofmechanisms,thefirstofwhichwascoveredabove,

i.e.,risksduetohabitatloss,fragmentation,andtheedgeeffects.Ameta‐analysisoffragmentation

investigationsshowsthatreducedfragmentareaandincreasedfragmentisolationhassignificantly

reducedtheabundanceofmammals,birds,andinsects(Haddadetal.2015).Understorybirdsin

Amazoniaareparticularlyvulnerabletoedgeeffects,road‐inducedfragmentation,fire,selectivelogging,

hunting,andtrafficdisturbance(Ahmedetal.2014).Whilepopulationsofmanyhabitat‐interiorspecies

maybediminishedintheroadsidebuffer,otheranimalspeciesareattractedtotheopencanopyand

thickunderstory.Thesetendtobeedge‐adaptedspecies–secondaryforestspecieswithlifehistory

traitstomaximizegrowthrates,oftenreferredtoas“weedyspecies.”Manyoftheseweedyspeciesare

notnativetothearea(Bruschietal.2015).Becauseedge‐adaptedanimalspeciesaregeneralistsand

haveverylargerangesandhighreproductiverates,theycontributelittletothebiologicaldiversitylevels

ofthebroaderregion.Meanwhile,thelossofspeciesadaptedfortheintacthabitatsdiminishesthe

speciesdiversityandecosystemresilience.(SeeAppendix5:EnvironmentalIndicator:WildlifeEffectsas

MeasuredbyBiodiversityhotspotsandUmbrellaSpeciesinCICPEC.)

Wildlifearevulnerabletoarangeofadditionalroad‐relatedconditionsbeyondecosystemeffects.

Wildlife‐vehiclecollision(WVC)isperhapsthemostvisibleandwell‐studied.Vehiclesareinherently

dangeroustowildlife.High‐volumeandhigh‐speedtraffic–whichtypifiesBRIroadprojects–

contributetothehighestratesofWVC(Gunson,Mountrakis,andQuackenbush2011;LitvaitisandTash

2008).Thesituationisexacerbatedforwildlifethatareattractedtoopenflywaysandpaths,further

increasingtheirriskofWVC(Bruschietal.2015;Coffin2007;Morellietal.2014;Myczkoetal.2017).

Thisphenomenonhasbeendocumentedespeciallyforpredatorsandlargeherbivoresandexplainsthe

highlevelofcollisionmortalityamongFloridapanthers(Felisconcolor),Iberianlynx(Felispardina),wolf

(Canislupus),andbrownbear(Ursusarctos)(SeilerandHelldin2006).Thedecreasedwildlifeabundance

duetovehiclecollisionshasbeenshowninsomestudiestoposeagreaterriskthanthenegativeeffects

ofhabitatfragmentation(JacksonandFahrig2011).

Withinrailwayecology,environmentalresearchhasalsofocusedprimarilyonwildlifeeffects.Wildlife‐

traincollisions(WTC)maybesimilartothatfromroads,buttheirmagnitudeoftheimpactisuncertain

(Dorsey,Olsson,andRew2015;Formanetal.2003;PoppandBoyle2017;SeilerandHelldin2006;

Wingardetal.2014).Whererailwayspassthroughhigh‐qualitywildlifehabitat,collisionswithungulates

andcarnivoresarecommon.Somecharacteristicsofrailascomparedtoroadsmakethemmoreprone

tocollisions,suchasdensevegetation(habitat)neartracksthatobscuresvisibility.Railwaycorridors–

lackingthehardasphaltpavementofroads–arealsomorelikelytoserveascorridorsforanimalsto

travelon.Forspeciesattractedbytherail“corridoreffect,”WTCcanbeespeciallydangerous.Inareaof

Alaska,forexample,moosepopulationsdeclinedby70%duetoWTC(Dorsey,Olsson,andRew2015).

FormostBRIroadandrailprojects–whichtendtobelarge–themoreseriousriskisnottheattraction

ofopencorridorsbutratherthebarrierscreatedbythelinearinfrastructureproject.Largedivided

highwaysmayserveasbarrierstoallspeciesexceptthosethatarehabituatedtoopen‐airflight.These

largeroadsarenotonlywidebutmosthaveadditionalbarrierssuchasmediansinthemiddleand

fencingoneithersideoftheroad(Bruschietal.2015).Inadditiontophysicalobstacles,frequentand

loudtrafficcancreatebehavioralmodificationsthatleadsomewildlifespeciessuchaselkduringthe

huntingseasontoavoidareaswithroads(Patonetal.2017).Attheotherextreme,narrowroadswith

lowtrafficvolumeandsomeregrowthcreatefarlessofabarrier,thougheveninthesecasestheroad

caninhibitsensitivespeciesthattendtoavoidbothedge‐affectedhabitatandtheroadclearingitself

(W.F.Lauranceetal.2014).Likeroads,raillinescancreateabarrierforsomespecies.Raillines,

especiallyHSR,oftenarelinedbyfencesthatcreateaddedbarrierstobothresidentandmigratory

wildlifespecies(Bruschietal.2015).

Thebarrierscreatedbyroadsandrailendangeranimalpopulationsbydisruptingmigration,splitting

populations,andthusreducinggeneticvariability(Sawyeretal.2013;Wingardetal.2014).Roadsare

responsiblefordecreasedmovementandgeneticvariabilityamongpopulationsofunderstorybirdsin

theAmazon(S.G.W.Laurance,Stouffer,andLaurance2004).Formigratoryandnomadicmammal

species,thesebarrierscancreateanevenmoreseriousobstacle.TheQinghai‐TibetRailway,for

example,createdabarriertothemigrationoftheendangeredPrzewalski’sGazellewhichhasreduced

thegeneticviabilityofthatspecies(Yuetal.2017).

Railwaycreatessomeuniquehazardstowildlifenotfoundinroads.Forexample,trainembankments,

whenhigh,canpresentasignificantbarrier(Wingardetal.2014).Similarly,wildlifecanbeelectrocuted

bythethirdrailofHSR.Smallanimals,especiallyturtles,canbecometrappedbetweenthetworails

(Dorsey,Olsson,andRew2015;SeilerandHelldin2006).

ii.WildlifePoachingandIllegalTradeRoadsandrailwaysnotonlycreateconduitsforwildlifetravel,buttheyalsooftenincrease

accessforhunters.Roads,andinparticularminorroadsthatpenetratingfrontierlandscapes,can

increasebothlegalandillegalhuntingofwildlife(Clementsetal.2014;Grayetal.2018;W.F.Laurance

2015;Wingardetal.2014).Thisislikelyagreaterthreatfromroadsthanrailways,duetothelimited

humanaccesspointsalongraillines.Trainstationsarebuiltinhumansettlements,sorarelyoffernew

accessforpoachinginfrontierlandscapes.Thatsaid,asnotedearliersomeHRSlineshaveaccessroads

flankingtheraillineformaintenanceandemergencieswhichcouldbeusedbypoachers.

Transportationinfrastructure–bothroadsandrail–posesasecondandperhapsmoreseriousthreatto

wildlifetrade:Improvedtransportationnetworkscanfacilitatethemovementofillegalwildlifetraffic.

ThisisanespeciallyseriousconcernfortheCICPECinSoutheastAsia,oneoftheworld’smostactive

centersforillegalwildlifetrafficking(BenYishayetal.2016;Broussard2017;Felbab‐Brown2013a,

2013b).Ultimately,itisthisimprovedtransportationefficiency–ironically,akeygoaloftheBRI

EconomicCorridors–thatpresentsthegreatestthreatforsomeofwildlifespecieshighlyprizedinEast

Asianmarkets.(SeeAppendix5:IllegalloggingandwildlifetradeinCICPEC.)

II.INDIRECTEFFECTSOnceroadandrailprojectsbeginoperating,theysetintomotionadditional,unintended

changesthatcanaffecttheenvironmentevenmoreprofoundlythandirecteffects.Twoparticularly

notableindirecteffectsarethechangesinforestcoverandGHGemissionsthatoftenresultfromnew

frontieraccessandchangeswithinmarkets.Inthissection,wewillreviewtheconditionsunderwhich

thelanduseandGHGchangearemostprevalentandtheimplicationsforBRIroadandrailprojects.

A.Land‐UseChange&DeforestationDeforestationcausedbyland‐usechangesfollowingchangesintransportcostscandramatically

exacerbateenvironmentalrisks.Roadbuildingopensupfrontierstosettlement,whileincreasingmarket

accessforfarmersandrancherstoforests.StudiesofdeforestationinBrazil(A.Pfaffetal.2007),Mexico

(G.C.NelsonandHellerstein1997),Belize(ChomitzandGray1996),andDemocraticRepublicofCongo

(ManLietal.2014)indicatesomecloserelationshipsbetweenroadbuildinganddeforestation.More

generally,infact,roadbuildinghasbeencorrelatedwithdeforestationrates−notingthattheresearchhasbeenheavilyfocusedontheimpactswithinmanysettingsinLatinAmericainthe1980sand1990s

(ChomitzandGray1996;Cropper,Griffiths,andMani1999;DeiningerandMinten2002;Ferretti‐Gallon

andBusch2014;GeistandLambin2002;G.C.NelsonandHellerstein1997;A.S.P.Pfaff1999).Nearly

95%ofalldeforestationintheAmazon,occurswithin5.5kmofroadsor1kmofrivers(Barberetal.

2014).Thiscangenerateaspiralinglossofforestandincreasingdamagetobiodiversityandecosystems

fromforestloss,habitatfragmentation,andedgeeffectsdiscussedintheDirectEffectssectionabove.

However,spatially‐explicitstudieswithattentiontodiversecontextsrevealsignificantheterogeneityin

thedeforestationgeneratedbyroads−bythelocationsofroads(andperhapsalsothetypes,althoughevidenceislimited).Thedegreetowhichroadsleadtodeforestationvarywiththetopography(Freitas,

Hawbaker,andMetzger2010;Hoyos,Cabido,andCingolani2018)andagriculturalsuitability(Ruan,Qiu,

andDyck2016),localwealthlevels(DeiningerandMinten2002;M.Li,Wu,andDeng2013),protection

enforcement(BenYishayetal.2016;Bhattarai,Conway,andYousef2009;HargraveandKis‐Katos2013),

andtenurearrangements(GeistandLambin2001;Mena,Bilsborrow,andMcClain2006).Cruciallyfor

forestdynamics,ithasbeenfoundthatpriorlandclearingandpriordevelopmentaffectroads’impacts

(Andersen1996;Cropper,Puri,andGriffiths2001;A.Pfaffetal.2018).Insum,foranumberofreasons

thedeforestationthatapolicymakershouldexpectduetoanewroadvariesgreatlybythesetting.

i.EvidenceonRoadImpactsGivenPriorDevelopmentandDeforestationChomitz(2007)delineatesthreemaintypesofforestsettings:forest‐agriculturemosaicland

wherepopulationdensitiesarerelativelyhigh,marketsarenear,anddeforestationrateishighforthe

sparseremainingnaturalforests;frontieranddisputedareaswherepressuresfordeforestationand

degradationarehighorincreasing,andcontrolisofteninsecureandinconflict;andareasbeyondthe

agriculturalfrontierwheresubstantialforestremainswithfewandlargelyindigenousinhabitants,and

somepressureontimberresources.ThesecategoriesaccordwiththoseusedbyPfaffetal.(2018)in

lookingatheterogeneousimpactsofnewroads,distinguishingbetweensettingswithhigh,medium,and

lowlevelsofpriorroadsanddeforestation.Forthosesettingswithmediumpriordeforestation,new

roadsareexpectedtoleadtosignificantdeforestation,asthoseareastendtobewherethefrontierof

developmentislocatedandexpanding.Bycomparison,moredevelopedforest‐agriculturemosaiclands

mayhavealreadyexperiencedmoredeforestationgiventhattheyalreadyhadrelativelylowtransport

costsandhighereconomicactivity.Suchforest‐agriculturemosaiclandareasmayevenexperiencenet

reforestationasaresultofnewroadconstructionifthefurtherimprovementsintransportcostleadto

outmigrationoratransitionawayfromagricultureastheprimaryeconomicactivity.Finally,forregions

beyondtheagriculturalfrontierwithlowpriordeforestationitisimportanttodistinguishbetweenthe

short‐runandlong‐runimpacts.Intheshortrun,inputsforrapidexpansionofproduction,suchaslabor

andcapital,arelimited.Thus,deforestationresponsescanbeminimal.However,ifnewroadsgenerate

profitopportunities,inputsandlaborwillenterthearea,increasingproduction—anddeforestation—in

thelongrun.Thatcanlead,inturn,tospatiallypath‐dependentdynamicsinwhichearlyarrivalscreate

theconditionsforothers,includingbylobbyingforotherinfrastructure,sothatthefrontierexpands.

Theimplicationsofthisrelationshipbetweenpriordevelopmentanddeforestationforenvironmental

impactsaremanifold.Newroadaccesstoforestwithmediumpriordevelopmentascomparedtonew

accesstoareaswithhighpriordevelopment,onaverage,leadstogreaterabsolutedeforestationandits

consequenthabitatloss,fragmentation,edgeeffects,andbiodiversityloss.Newroadaccesstointact

frontierlandscapes(areaswithlowpriordevelopment)coulddoeitherlittleoragreatdealtoforests,

dependingonthetimeperiodbeingconsideredaswellasanycomplementarymitigationpoliciessuch

asprotectedareasthatcouldflankinvestments(seePart2).Ifsuchaccesssetsinmotionasignificant

economicdevelopmentdynamic,theimpactonforestscanbehighestofall.Thisimpactisfurther

compoundedbythefactthathabitatlossanddegradationtointactfrontierlandscapesisrelatively

moreharmfulthantosecondaryordisturbedhabitatsbecauseoftheecosystemfunctionsoflarge

intactareasofforest,includinghigherlevelsofbiodiversity,highlevelsofecosystemservicessuchas

carbonsequestration,androleasarefugia,asdescribedaboveintheDirectEffectssectiononIFLs.

ii.TheoryaboutShiftsinSuchRisksasEconomicDevelopmentProceeds

Twowell‐knowntheoreticalframeworks–notmutuallyexclusive−provide“macro‐scale”rationalesfor

expectingheterogeneitiesacrosssitesintheimpactsofinvestmentsintransportationinfrastructure.In

consideringaverageimpactsofnewroads,givenotherfactors,andmarginalimpactforeachcondition,

bothEnvironmentalKuznetsCurve(EKC)andForestTransition(FT)frameworksconsiderevolutionsof

wholeeconomies.TheyarepotentiallyrelevantifBRIinvestmentsunleashsignificanteconomicshifts.

TheliteratureconsideringEKCsfocusesonchangesinpercapitaincomeandtheirimplicationsforthe

changeinenvironmentalquality.Asincomesriseforanygivenpopulation,consumptiontendstorise

overtimewithit,whileenvironmentalqualitytendstofallduetotheeffectsofbothproductionand

consumption.Asaconsequence,eventuallythemarginalutilityofconsumptiontendstodiminishwhile,

incontrast,risingscarcityofenvironmentalqualityraisesthemarginalutilityofpreservingecosystem

servicesandenvironmentalpublicgoods.Thiscouldmotivatechangesinindividualbehaviorsandcould

leadtosupportforpublicenvironmentalregulationthatwouldincreaseenvironmentalqualityevenifat

theexpenseofconsumption.EmpiricalevidenceforanEKCintermsofdeforestationremainsmixed

(CropperandGriffiths1994;FosterandRosenzweig2003;KoopandTole1999).Nevertheless,thepoint

holdsthatawealthiersocietymightfinditworthwhiletousemorecostlyroutestolowerenvironmental

damageorputprotectedareasalongsideroadstolowerimpacts.(SeePart2onmitigationstrategies.)

MorerecentstudiessuggestthatEKCpatternscouldbemoreN‐shaped,withareturntohighlevelsof

deforestationaswealthincreasesdemandforland‐intensiveproducts(JoshiandBeck2017).DeFries

andcolleagues(2010)provideevidencethatconcordswiththeproposedN‐shape:inatimeseriesfrom

2000‐2005theyfindthaturbanizationandtradeareassociatedwithincreaseddemandforagricultural

products,whichleadstoincreaseddeforestation.ItseemsthereforethatanypositedEKCislikelytobe

acontext‐dependentdepictionoftherelationshipbetweendemandfordifferentcommoditiesandtheir

correspondinglanduses.Atlowlevelsofincome,themostimportantfactorsmaybelocaldemandfor

localagriculturalgoodsvs.localdemandforforestproducts.Athigherlevelsofincome,acountry’s

forestedregionsarelikelytobemoreintegratedintotheurbaneconomyandglobaltradenetworks,so

therelevantvariableswillincludedemandforexportableagriculturalcommoditiesversusthedemand

forsustainablyharvestedtimberandothercommoditiesoramenitiesprovidedbyforestcover.

Thiscompetitionbetweenlandusesappearstobesignificantlyrelatedtotradeopennessand

comparativeadvantage.Lebloisandcolleagues(2017)findthattradeinagriculturalcommoditiesand

tradecompetitivenessareassociatedwithdeforestation.However,thiseffectislesspronouncedwhere

agriculturallandisalreadyextensive(Leblois,Damette,andWolfersberger2017).Itappearsthatregions

whereagricultureisstillexpandingintoforestareaswillexperiencemoredeforestationpressureas

developmentprogresses,whileregionswhosecomparativeadvantageliesinsustainabletimberor

agroforestryproductionsystemsmayseenetreforestation.Thispatternisborneoutbytheexperience

oftheAmericanSouth,wherelandusehastransitionedfromagriculturetowardforestryinthe20th

century(Carter,Kellison,andWallinger2015).Similarly,Kenyahasseensignificantsecondaryforest

regenerationasaresultofpost‐colonialdemandfortreeproductsonsmall‐holdingfarms(Holmgren,

Masakha,andSjöholm1994),whileinIndonesiacomparativeadvantageinannualagricultureversusin

treeplantationsissignificantlylinkedtodeforestationoutcomes(Baylis,Fullerton,andShah2016).

ForestTransition(FT)theory,whichhasfocuseduponforestsinparticular,positsthatatleastsome

countriesmayundergoalarge‐scaletransitionfromnetdeforestationtonetreforestation.FTtheory

depictsaprogressionovertimethatstartsfromwidespreadintactforestlandscape,movesthrougha

periodofhighdeforestation,towardastateofstabilizedforestcoverandforest‐agriculturemosaics,

andeventuallygeneratesnetreforestation(notingthathistoricallythishasincludedrisingplantations).

ShiftsfromnetforestlossestonetforestgainshavebeenobservedinEuropeandtheEasternUnited

States(Rudel,Schneider,andUriarte2010),aswellasinsomedevelopingnationssuchasVietnam

(MeyfroidtandLambin2009),CostaRica,PuertoRico,andregionsofIndonesiaandIndia.However,

therearemultiplepossiblemechanismsforsuch“foresttransitions”andnotallofthemclearlyinvolve

netforestgainswhenlookingmorebroadlyacrossspace,forexample,atthegloballevel.Tradeclearly

canallowforonecountrytoriseinforestbysimplydisplacingdeforestationelsewhereviainputsof

timberandagriculturalproducts(Meyfroidt,Rudel,andLambin2010;A.PfaffandWalker2010).

Anothermechanismthatcouldsupportaforesttransitionviamovementswithinonecountrymightbe

laborreallocationfromagriculturaltourbanareas−whereforestimpactperpersonmaybelower.If

laborisscarceinanarea,andthenon‐agriculturaleconomyexpands,thenagriculturallandusecould

decreaseandforestscouldincrease.Thisaccordswithpatternsofstructuraltransformationandcould

becausedbyimprovedtransportationinfrastructure(AsherandNovosad2015;D.J.Kaczan2016).This

canhappenwhenthemanufacturingsectorexpandsandmaycontinueastheservicessectorsexpand.

Thinkingattheregionallevel,employingageneralequilibriumperspective,transportimprovements

mightthemselvesincreasethepaceofurbanization,facilitatingmigrationtourbanareas.Thiscanalso

bepartofincreasingtherelativeproductivityofmanufacturingandservicesinurbanareasversus

agriculture,solandandpopulationexitagriculturalproductioninruralareas(Dengetal.2011).

Movingtogeneralequilibriumeffectsfortheeconomyasawhole,moretransportationinfrastructure

couldincreasetherateofeconomicgrowth.Country‐widegrowthcouldleadtoincreaseddemandfor

agriculturalcommoditiesortimber,raisingdeforestation.Alternatively,growthcouldcontributetoan

EKCpattern,andapossiblenetreforestation,sincethemarginalimpactofincreasedincomeisnotthe

sameatallpointsofeconomicdevelopmentbutinsteadappearstovarywithfoodandnaturalscarcity.

Whichofthesemechanismsdominatesinaparticularsettingandtimeperiodisanempiricalquestion.

Withmoreinformationaboutasetting,wecanspeculateaboutthelikelyimpactsofvariousformsof

investmentsintransportinfrastructure,includingnewtransportationcorridorsandvariousupgrades.

iii.EmpiricalExamplesTheempiricalliteratureonroadsandforestcoveronthewhole,then,providestheopportunity

tomatchcasestotheoreticalpredictions.Thefollowingsummariesreviewempiricalstudiesbyregion,

thenraisethequestion(notsufficientlyexploredintheempiricalliterature)ofthetypeofinvestment.

Brazil:TheBrazilianAmazoncontainsasignificantshareoftheworld’sprimaryforestandhas

attractedahighdegreeofattentionfromconservationists,includingduetoitsextensiveintactfrontier

landscape(Potapovetal.2017).TheliteratureaboutroadsanddeforestationinLatinAmericaindicates

thatroadbuildingtendstoraisedeforestationinsuchsettingsandhigherroaddensityifanythingraises

landclearing,particularlyforcattleproduction.Asstudieshavesuggestedthatdeforestationimpacts

differbypre‐existingclearingpatterns(Pfaffetal.2018),androad‐linkeddeforestationinsomeareas

waslowerthansomeexpected(ChomitzandThomas2003),avoidingdeforestationfrominvestments

seemslikelytorequirebothtargetingofparticularsettingsandcomplementarymitigationpolicies.

India:RelativetoBrazil,muchofIndiaisalreadycharacterizedbyforest‐agriculturemosaicland,

withvirtuallynoremainingintactforestlandscape(Potapovetal.2017).StudiesofthePrimeMinister’s

RuralRoadprojecthaveconcludedthatinthistypeofsettingincreasingtheroadnetworkinruralIndia

hasresulted,onaverage,inroughlyzeronetdeforestation(Asher,Garg,andNovosad2017;D.J.Kaczan

2016).Heterogeneityanalysissuggestsnetreforestationfordenselypopulatedagriculturalareas(D.J.

Kaczan2016)withnetdeforestationformoredistantandheavilyforestedregions.Onemechanismthat

seemstoatleastinpartexplaintheresultfordenselypopulatedareasislaborreallocation,sinceroad

buildingisfoundtobelinkedtoanincreaseinnon‐agriculturalemployment(AsherandNovosad2015).

China:Chinaalsohasasignificantruralpopulationlivinginwhatmightbecharacterizedas

forest‐agriculturemosaiclandandlittleintactforestlandscape(Potapovetal.2017).Dengetal.(2011)

findthatroadconnectionsinruralChina,controllingforeconomicvariablesatthewatershedlevel,are

associatedwithnetreforestation.Themechanismpositedislabormigration,asroadsinChinatendto

allowmigrationintoswiftlygrowingindustrialcentersofproduction.Anotherelementhereisthehigh

degreeofeconomicgrowth.Inadditiontocausingrural‐urbanmigration,agrowingeconomycreates

moredemandforcommoditieswhichmaybeproducedattheexpenseofforests.Generalequilibrium

effectslikethesearemoredifficulttoempiricallyverifybutmayhavesignificantimpactinreality.

Cross‐countryComparisons:DasguptaandWheeler(2016)andDanyo,DasguptaandWheeler

(2018)providespatiallyexplicitestimatesofforestlossassociatedwithroadupgradinginCameroon,

Bolivia,MyanmarandtheLaoPeople’sDemocraticRepublic.Theirestimatesindicateafootprint

between5and10kmwhereforestlossislikelyfollowingroadimprovements.Butwithintheirestimates

significantheterogeneitysupportstheframeworkpresentedinthisreport.ForLaoPDR,themost

significantforestlossisexpectedinthenorthernregionwherelowpre‐existingclearingcoincideswith

relativelyhighexpectedreturnsfromagriculture.Similarly,theless‐developednorthernhighlandsof

Cameroonexperiencesignificantforestlossassociatedwithroadimprovements.Myanmar’scentraland

east‐centralforestsandBolivia’sYungasregionarealsoseriouslythreatened.Protectedareasappearto

haveapositiveeffectonforestcoverinallcountriescoveredbythesestudies,pointingtowardaneed

forprotectioninareaswithbothlowpriorclearingandhighpotentialforagriculturalprofitability.This

cancomplementroadsandrails.

iv.ApplicationtoBRIBRIroadprojectsareproposedforawiderangeofsettings,soitisnothelpfultoofferasingle

viewconcerningtheindirectdeforestationeffectsfromnewconstructionandupgradingofBRIroads.

Yetseveralfactorsleadustobelievethatforlargefractionsofpotentiallyaffectedareas,thepatterns

identifiedinIndiaandChinatodatemayonaveragebemoreapplicablethanthosedocumentedforthe

BrazilianAmazon−althoughwehastentoemphasizethatevenwithinregionsofIndiaandChinait

appearsthattheheterogeneityofimpactscanincludefallingversusrisingforestarea.(Appendix7:

EnvironmentalIndicator:IndirectEffectsasMeasuredbyForestCoverChangeinCICPECexploresinmore

detailindirecteffecttrendsinSoutheastAsiaandtheirrelevancetothreeBRItransportationroutes.)

Thesingleclearestreasonfortheapplicabilityofsomemorethanotherpastresultsforroads’impactsis

thatmostBRIroadprojects–includingeventhenewconstructionprojects–arebuiltorintendedtobe

builteitheronorveryneartotheexistingfootprintsofsmallerroads.Thus,theyarenotprovidingany

entirelynewaccesstolandscapefrontierseveniftheyareimprovingaccess(perhapstocriticallevels).

Mostsuchplannedinvestmentsarealsoconnectingpopulationcentersand,assuch,notcreatingaccess

tofrontiersattheirendpoints−thoughpotentiallycreatingaccesstointactfrontierforestsintheareasbetweenthepopulationcenters(itiswithinthoselattertypesofareasthatthetypeofinfrastructure

andcomplementarymitigationpoliciesarelikelytomatteralotforforestoutcomes).ForBRIdivided

highwayprojectswithfourlanesormore,accesstointactforestmayberestrictedbyexitrampsortolls.

Insuchcases,alimited‐accesshighwaythatreplacesasmaller,unrestrictedroadcouldactuallyreduce

accesstofrontierforests,atleastinprinciple.Withinthealreadymoredenselydevelopedareas,effects

fromincreasedcommerceandexpansionofnewmanufacturingorservices,urbanization,changesin

landprices,andotherequilibriumeffectsarelikelytodominate.Finally,whilemostBRItransportation

projectshavenotbeensitedorplannedneartoIntactFrontierLandscapes–remoteandundeveloped

wildernessareaswouldbedisturbedbyanydeforestationfromnewroadconstruction–afewplanned

projectsalreadyhavebeen.(SeemapofIFLsinFigure1andtheexampleofaThaiprojectinAppendix4:

EnvironmentalIndicator:EcosystemEffectsasMeasuredbyIntactFrontierLandscapes.)

v.RoadsversusRails MostBRIprojectsarerailnotroadprojects,yettheliteratureonindirecteffectsfromrailson

landuseisfarlessextensive.Donaldson(2010)findsthatrailconnectionsinIndiadecreasetradecosts

andpricegapsbetweenregions,integratingmarketsandtransmittingpricesacrossthesub‐continent.In

theUScontext,Vessali(1996)findsthatrapidtransitinfrastructuretendstoincreaseurbandensitynear

stations,ifurbanzoningregulationisalignedwiththisoutcome.Thus,itappearsthatrailinvestments

exhibitsimilareconomicimpactsasdoroads,creatingopportunitiesfortheexpansionofa*gricultural

commoditieswhilealsopotentiallyincreasingurbanizationandoutmigration(deSoyresetal.2018).

Yetonemightnonethelessconceptualizeacontinuumintermsofthedegreeoflocalaccesstotransport

networksandconsequentlocalenvironmentalrisks.Railprovidesaccesstofreightandpassengersonly

atstations.High‐speedraillinesservicefewerstationsthanconventionalrailways.Assuch,railallows

accessatfewerpointsandthustheymayopenlessaccesstofrontiers.Thishasledsometoadvocatefor

raillinksasanalternativetoroadsinsensitiveareasliketheAmazon(VianaandCenamo2008).Risks

couldstillbehighifstationsarenearforestfrontiersorserveasnewnodesfortransportationnetwork

growthwhichgeneratesexpansionintoforestfrontierareas.Moreover,manyhigh‐speedraillinesare

oftenaccompaniedbyminoraccessroadsnearthetrackstofacilitatemaintenanceandemergency

repairs.Thus,BRIrailinvestmentsarelikelytoallowsomeaccesstofrontierforests,evenifperhapsin

generalthatislessthaniscreatedbyroads.Limited‐accesshighwayssimilarlyprovideaccessonlyat

designatedexits,whiletheconventionalroadinfrastructuremightallowaccessatanypointalongits

length.Thus,itmightbeexpectedthatbothrailroadsandlimited‐accesshighwayscauselessindirect

environmentalriskonaveragethandohighwayswhichprovidegreateraccess.Moreover,onecould

envisionthepossibilitythatsuchinvestmentsinrailsystemsorlimited‐accessdividedhighwayscould

potentiallyreducepressureonhabitatsiftheydrawtrafficfromsmallerroadsthatoffergreateraccess.

B.CarbonandOtherGHGEmissionsThepotentialforindirecteffectsofland‐usechangeanddeforestationfromBRIroadandrail

construction,asdescribedabove,couldnotonlyprofoundlyaffectforestcoverandecosystemhealth

butalsogenerateasignificantimpactontheglobalclimate.Deforestation,especiallyinthetropics,

representsasignificantsourceofcarbonemission,tothepointofcontributingclosetoone‐fifthofall

anthropogenically‐producedemissions(Harrisetal.2012).DeforestationwithintropicalIFLscouldbe

especiallyharmfulfortheirimpactoncarbonemissionsfromdenseforests(Potapovetal.2017).

BRIprojectscanhaveindirecteffectsonGHGemissionsthroughthesecondaryimpactviashiftsintrade

andresultantchangesinvehicletrafficandlanduse(MaliszewskaandvanderMensbruggheinprep.).

Anassumptionhasbeenthatfuel‐efficientmodesoftransportationshouldreduceenergyconsumption

andGHGemissions,asabove.High‐speedrailhashigherenergyefficiencyandlowerGHGemissionson

aperpassengerorpertonfreightbasisthanconventionalrail,whichhashigherefficiencythanfreight

trucksandcars,whichinturnhavehigherefficiencythanairtransport.However,whethertheupgrade

toamoreefficientmodetranslatesintoanabsolutedecreaseinGHGsorairpollutionemitteddepends

onseveralfactorsincludingsubstitution(howmanypassengersorfreightshiftfromairtransport,cars,

orconventionalrail)andtrafficgeneration(howmuchnewdemandfortransportisgeneratedbythe

construction).Emissionssavedalsodependsonhowfreedcapacity(ontheroadandrunway)isused.If

newlyavailablerunwaycapacityisusedformoreenergy‐intensivelong‐haulflights,forinstance,then

modesubstitutioncouldraiseenvironmentalimpact(Givoni2006).Theresultsvarydependingonthe

settingandthecompetingmodesoftransportationaswellasuncertaintyinfuturetransportdemand,

patternsofurbanization,technology,andsourcesofenergyproduction(e.g.Chaietal.2016;National

Rail2009;WestinandKågeson2012).Forexample,whencomparingHSRandairtransportusinga

duopolymodel,D'Alfonsoandcolleagues(2016)foundHSRaremoreenergyefficientperpassenger/km

butbecauseoftheinduceddemand,thenewrailwaysystemtendstoincreaseenergyconsumptionand

GHGswhencompetingwithairtransport.Similarly,modelingbyWestinandKågeson(2012)foundthat

substitutionwouldhavetobeverylarge,andmostlyfromairtransport,tobalanceGHGemissionsfrom

newHRSconstructionandincreasedtraffic.AgeneralequilibriumstudybyChenandcolleagues(2016)

ofhigh‐speedrailinChinauseddatafromthelast15yearsofHRSconstruction.Theauthorsfoundthat

HSRconstructionhasresultedinincreaseddemandfortransportservices,therebyresultinginmore

GHGemissionsthanweredisplaced.Studiesliketheseindicatethatinduceddemandisanimportant

elementoftransportationinfrastructure’simpactthatmayaffectlanduseanddeforestationaswell.

Part2:MitigationofEnvironmentalRisksfromBRIInvestmentsin

TransportationInfrastructure

WhileenvironmentalrisksgeneratedbyBRIroadandrailprojectscanbesignificantandextensive,arangeofchoicesexistthatcouldlowerboththechancesandthemagnitudesofsuchpotentialharms.WithinPart2,wepresenttheseactivitiesintermsofthe“mitigationhierarchy,”asensibleframeworktoorganizetheconsiderationofwaystolimitvariedenvironmentalrisksfromBRIroadandrailprojectstobiodiversityandecosystemservices.Suchahierarchyiscommonlycomprisedoffourtypesofactions–avoidrisks,reducerisks,restore,andoffsetdamages–thatoftenareusedincombination(Figure1).

FortheBRI,suchactivitiescouldbalanceeconomicgainsfromtransportinvestmentwithgainsfromsustainingnaturalcapital.Bothsupporthumanwell‐beinglocallyandaroundtheglobe.MitigationactivitiescanhelptoreducethedirectandtheindirecteffectsfromBRIprojects(describedinPart1).Suchactivitiescanhelpmeetstandardsorguidelinesestablishedbymultilateraldevelopmentbanks,Chinesebanks,industryassociations,andvariousotherfinancialstakeholders.TheycanalsohelptomeetenvironmentalpoliciesandregulationssetbyhostcountrygovernmentsandbyChinesegovernmentagenciesoverseeingBRIandotheroverseasinvestments.Environmentalpoliciesincludinglaws,regulations,standards,andguidelineswillbediscusseddirectlyinPart3.Takentogether,mitigationactivitiescanreducetherisks,costs,anddelaysofBRIprojectsandimprovestakeholder

supportinhostcountries,inChina,andinternationally(Ekstrom,Bennun,andMitchell2015).

AppropriatemitigationactionsforanyBRIprojectmaybeidentifiedthroughenvironmentalassessment

processesiftheyareappliedatsufficientlyearlyplanningstagesandatsufficientlybroadspatialscales.

Forthemostimportantmitigationstep–avoidance–strategicenvironmentalassessment(SEA)from

Figure2.TheMitigationHierarchy.www.thebiodiversityconsultancy.com/approaches/mitigation‐hierarchy/

theverystartofplanningcanbeacriticallyusefultool.SEAaretypicallyadistinctregionalorsectoral

assessmentconductedbygovernmentagenciesatthestageofprioritizationwithinplans,programs,or

policiespriortothedevelopmentofanyindividualproject(UNEnvironment2018a).Environmental

assessmentsthatareconductedatearlystagesintheplanningprocessandwithabroadspatialscope–

forexample,acrossanentire“BRIEconomicCorridor”–canhelptoavoidsubstantialenvironmentalrisk

and,further,allowthecoordinationofmitigationchoicestobroadenprojectbenefits.Thisallowsall

decisionmakersthemostpossibleflexibilityinplacingnewroadsandrailswheretheyarelikelytocause

theleastenvironmentalandsocialharm,whilemaintainingdesiredeconomicandconnectivitybenefits.

Forotherstepswithinthemitigationhierarchy,well‐establishedenvironmentalimpactassessment(EIA)

processescanusuallyguideindividualprojectplanningandimplementation,especiallyforthelocalized

directeffectsgeneratedduringconstructionphases.Suchmitigationstrategiesmightincludereducing

impactthroughalteringprojectdesign,restoringimpactedareas,ordevelopingoffsetstocompensate

forunavoidableimpactsanddamages.BothSEAsandEIAswillbediscussedingreaterdetailinPart3.

Table2.PotentialmitigationactionsforBRItransportationInfrastructureinvestmentsusingmitigationhierarchy

Avoid Integratedplanningofroutechoicetoavoidvulnerableenvironmentsandmaximizenetgains:[1]

Identifyalternativerouteoptionsthatcansimplyavoidtheenvironmentallyandsociallymostsensitiveareas,i.e.,wherethereareclearlyhighenvironmentalorsocialdamages;

Conductanintegratedcost‐benefitanalysis(consideringeconomic,environmental,andsocialimpacts)toguideselectionbetweenalternativeroutes,giventhatallofthebenefitsandcostsareheterogeneousacrossroutesandsomeroutesmaybebetteronmanycounts.

Thisshouldbedoneatthemicrolevel,forBRIcorridors,andtheBRIasawhole,toaccountfortheinterdependencyoflocations,investmentimpacts,andaffectedenvironments.

Reduce Mitigateimpactsthroughenvironmentally‐consciousengineeringandcomplementarypolicy:

Choicesoftechniques:Wildlifecrossings(bridgesandunderpasses,withmechanismsto‘funnel’wildlifetocrossing),soundbarriers,lightsdownwardstoreducelightpollution,retentionoftrees,timingconstructiontoavoidimportanttimesforanimalmigrationormating;tunnel‐bridge‐tunnelengineeringinordertoreducelandslideanderosionrisks.

Choicesofcomplementarypolicies:Regulations,theirenforcement,andincentivestoreducedeforestation,poaching,andvulnerablespeciestrade.Thismayincludethecreationofprotectedareas(PAs)nearbytotransportinvestments,addressingweaknessesinenforcementcapacity,and/orincentive/compensatorypaymentstolandlordsorlocalgovernmentsinreturnformaintainingforestsandecosystems.PAsshouldbecoordinatedalongtheBRI,toensurethesereduce,ratherthandisplace,harmfulactivity.

Applyintegratedcost‐benefitanalysesinselectingacrosstransportoptions(roadcategories,railversusroads,electricversusstandardrail)aswellasregulationonvehicleemissionsandmaintenance,etc.Theseconsiderationsraisethecasefor(particularlyhigh‐speedelectric)railoverroads–duetolowerpollution,andreducedencroachmentonfrontierlandscapesduetofeweraccesspointsandtheirconcentrationinalreadydensecities.

[1]SuchasIntactFrontierLandscapes(IFLs),biodiversityendemismhotspots,protectedareas,forestsliabletodeforestation,landscapeswithtopographicalorearthquakehazards,andothervulnerablelandscapes.

Restore Takeremedialactiontorepairdamageinflictedbytheconstructionprocess:Stabilizedamaged

slopes;replantvegetation;andrepairwaterwaysorwetlandsdisruptedbynewinvestments.

Offset Compensateforenvironmentaldamagethatcannotbeavoided,reduced,orrestoredbyinvestingin

off‐sitelocationstoachieveoverallnetneutralornetpositiveenvironmentaloutcomes:

carbonoffsetting,or,followingharmtobiodiverseareas,theenhancementofalternativecomparablesiteselsewhere(e.g.withsimilarendemicspeciesorecosystemfunctions).

mechanismsforachievingsuchimpactselsewherecouldincludeprotectedareas(thattargetpressure),“PaymentforEcosystemServices”(PES),biodiversitycompensationfundsintowhichprojectspay,biodiversitybankssellingoff‐settingcredits,andmoreadhocproject‐by‐projectsolutions,allsupportedbynationalorlocaloff‐settinglaws.

Off‐settingisproposed,however,onlyasa’lastresort’,becauseatleastinlightofpastevidenceit

appearstosufferbothefficiency(impact)andequity(distribution)challenges.

FortherestofPart2wedescribeindetailthefourcategoriesofthemitigationhierarchyinrelationto

BRItransportationinfrastructure.Table2describesexamplesofmitigationactionsandAppendices7

and8describehowthemitigationhierarchyframeworkcanbeusedtodevelopasuiteofmitigation

actionsfortheCICPECinSoutheastAsia.

I.AVOID:SelectionofProjectsandRoutesAvoidanceisgenerallyconsideredthemosteffectivewayofreducingpotentialnegativeimpactsfromtransportationinvestments(ICMMandIUCN2013;McKenneyandWilkinson2015;PilgrimandEkstrom2014).AvoidingtheriskstotheenvironmentfromBRItransportationprojectscouldentailrelocatingorrethinkingaprojectorprioritizingotherprojectswhenresourcesarelimited.SuchspatialplanningandprioritizationacrossallpossibleBRItransportationprojects(notjustsingleprojects)wouldideallybepartofanearly‐stageSEA‐typeregionalassessmentthatwouldconsiderentireorlargesectionsofBRIEconomicCorridors.Itisimportanttoconductsuchplanningduringtheearlystagesoftheplanningprocesswhencost‐effectiveavoidanceoptionsarestillavailableandfeasible.Asaproposedprojectmovestowardapprovalitismorelikelytojustproceedasinitiallyconceptualized,regardlessofthegainsfromalternativeoptions(Ekstrom,Bennun,andMitchell2015;UNEnvironment2018b).Itisimportanttoconductsuchearlyplanningoverawideregion,withallpotentialprojectsandlocationsconsidered,sothatindividualprojectsarenotsimplydisplacedtootherlocationswithsimilarrisks.

Projectplanningforinfrastructureinvestmentsoftenidentifiesenvironmentallysensitiveareasthatshouldbeavoided.Atthelevelofanindividualproject,afeasibilitystudyisconductedearlyintheprocesstoidentifydisqualifyingriskssuchashighvulnerabilitytolandslides,flooding,hydrologicdisruptions,orotherrisksthatcannotbereasonablyminimizedandmanaged.Sensitiveareasforbiodiversityandecosystemservicesshouldalsobeaddressedinprojectfeasibilitystudies,thoughthatisnotalwaysthecase(W.F.Laurance2015).Protectedareas(PAs)andlargeexpansesofundevelopedwildernessareoftenusedtodesignatetheenvironmentallysensitiveareasthatshouldbeavoided.

AsdescribedinthePart1andinAppendix4,PAsandintactfrontierlandscapes(IFLs)typicallyfunctionasimportantrefugiaforbiodiversityandsafeguardmanyecosystemservices.Tobeeffective,though,theseprotectedareasshouldbeestablishedas“offlimits”oravoidanceareasforenvironmental

objectivesearlywithinaplanningprocess(Bruschietal2015;Mahmoudetal.2017).5Appendix8(MitigationStrategies:ProtectedAreas)illustratestheopportunitiesthatmightbeavailablewithinCICPECforlocatingnewprotectedareasorincreasingtheenforcementofexistingPAswithinrelativelyundisturbedforestedareasthataretargetedforplannedBRIhighwayprojects.Ensuringadequateenforcementisanimportantcomponentofthisconservationstrategy,asisillustratedbytheencroachmentinSnuolWildlifeReserve,Cambodia,whereenforcementwaslacking(Appendix8).ItisrelativelyeasytousePAsandIFLsforplanningpurposes–againleavingimplementationascriticalforachievingactualimpacts–becausetheyarerelativelyeasytodesignatespatially.However,someoftheareasatriskthatcontainsomeoftherichest,mostuniqueandthreatenedbiodiversityorecosystemservicesareoutsideprotectedareas.Theseareparticularlyimportanttoavoidwhensitingroadsandrailwaysbecauseoftheirlackofprotectedstatus.Appendix5(EnvironmentalIndicator:WildlifeEffectsasMeasuredbyBiodiversityHotspotsandUmbrellaSpecies)providesanexampleofhotspotsforbiodiversitywhichconsidersspeciesrichness,endemism,andvulnerabilitywithinCICPEC.Identifyingandmappingsuchhotspotsforbiodiversity,waterprovision,carbonstorageandsequestration,andotherecosystemservicesismorechallengingthanmappingPAsandIFLsbutattentiontothesenaturalassetsearlyintheplanningprocesscanprecludecostlyactionsatalaterstageofprojectdevelopment.Manyexamplesexistinwhichgeospatialmodelinganddataanalysishavebeenusedtoassistwiththemappingofavoidanceareasforlinearinfrastructure.Mostfocusonroadnetworks,suchasstudiesofalternativeroutesinMyanmar,Nigeria,andTanzania(Caroetal.2014;Dobsonetal.2010;Fyumagwaetal.2013;Helsingenetal.2018;Mahmoudetal.2017;butseeFyumagwaetal.2013),showinghowroutechoicecanavoiddamagetoprotectedareas,ecosystemservices,andwildlifemigrationpatternswhilestillgeneratinghighsocioeconomicbenefitsbyconnectingagriculturalcenterswithmarkets.DasguptaandWheeler(2016)andDanyo,DasguptaandWheeler(2018)spatiallyexplicitlyestimatethelossesthatcouldbeavoidedwithinupgradingofroadsinCameroon,Bolivia,MyanmarandLaoPDR.Similarbalancingofobjectivescouldbedoneforrailprojectstooptimizeenvironmental,engineering,economic,andotherfactors.Forexample,Dongandcolleagues(2018)developedanintegratedriskevaluationmodelusingecological,social,andeconomicdatafromthehigh‐speedrailwayproposedfortheChina‐Mongolia‐RussiaEconomicCorridor.Usingdataonbiodiversity,forestcover,protectedareas,pronenesstofireandearthquakedisasters,theauthorsidentifiedregionswithinHeilongjiang,RepublicofBuryatia,IrkutskOblast,ZabaykalskyKrai,andtheLakeBaikalareathatfacehighecologicalriskfromHSR.Toavoidtheseecologicallysensitiveareasyetallowsocialandeconomicbenefits,theyproposedalternateroutesandpoliciestomitigateconstructionrisks.InKenya,publicattentiontoenvironmentalrisksfromtheStandardGaugeRailwayledtothemodelingandmappingofrisksandalternativeroutesbetweenNairobiandbothMombasaandMalaba.Thegoalsofmappingandmodelingincludedavoidingwildlife‐traincollisions,pollution,barrierstowildebeestmigrationandmore(Ambani2017).Appendix9(MitigationStrategies:UsingUmbrellaSpeciesforMitigationPlanninginCICPC)providessomeexamplesofmappingandplanningintheChina‐IndochinaPeninsulaCorridortoidentifykeyareasformitigation.

Chinahasexperienceswithenvironmentalsafeguardswithinthedevelopmentofitsrailnetwork−over30years−workinginclosecoordinationwithinternationalinstitutionssuchastheWorldBank(WorldBankTechnicalAssistanceProgram2009).Chinaroutinelydoesnationalandproject‐levelrailwayEIAs,

5EstablishingPAsorIFLsasofflimitstoroadandraildevelopmentcanbepoliticallychallengingbecause,insomecases,theveryfactthattheyhavenopreexistingdevelopmentcouldmakethemattractiveforlinearinfrastructuresinceagovernmentwouldnothavetopurchaseorexpropriateprivatelandand/orresettlelandowners.

whichincludeanalysesofalternativeroutestoavoidenvironmentallysensitivesites.Forexample,initsplanningoftherecentlycompletedHSRGuiyang‐GuangzhouRailwayProject,theChinesegovernmentidentified47environmentallysensitivesitesthatincludedprotectedareas,forestparks,scenicareas,watersheds,andculturalrelicssites.Intheprojectfeasibilitystudy,plannerscreatedmanyalternativeroutesandtherouteselectedavoided40ofthosesensitiveareas(Wang,Yang,andQuintero2012).

II.REDUCE:AfterProjects/RoutesSelected

Ifavoidingrisksisimpossible,thenreducing6environmentalrisksshouldbeconsidered,assuggestedbyEkstromandcolleagues(2015):“Measurestakentoreducetheduration,intensity,significanceand/orextentofimpacts(includingdirect,indirectandcumulativeimpacts,asappropriate)thatcannotbecompletelyavoided,asfarasispracticallyfeasible.”Reductionsintheenvironmentalrisksofaprojectbelowthresholdsofharmmaybepossible.Ifnot,actionsmayreducethescaleandtheexpenseoftheremediativemeasures,suchasrestorationoroffsets,thatmayberequiredlaterforunmitigatedloss.Riskreductionviaprojectandrouteselectionhasbeenintroducedabove,withintheAvoidancesection.Forsomeinfrastructureprojects,suchastheHSRGuiyang‐GuangzhouRailwayProject,China,describedabove,therecanbeanoverlapbetweendecisionstofullyavoidimpactsversuspartiallyavoidimpacts,i.e.,reduceimpacts.Inthatexample,railwayplanningcompletelyavoided40sensitivesitesyetbroadlyitwouldalsoberighttoclaimriskreductionby85%(40ofthe47sensitivesites).Forthiscase,actionscouldbeconsidered“avoidance”or“reduction”butwewillnotrepeattheavoidancediscussionhere.Instead,fortherestofthissectiononreducingrisks,wewillconsiderseveralotherreductiondecisionsforBRItransportationprojects.Firstistheselectionoftechniquesduringinfrastructureconstructionandoperations.Secondistheuseofcomplementarypoliciesthatcouldchangetheimpactsofinvestments.

Justasforavoidance,reductionsshouldbeproposedduringearlyprojectfeasibilitystudies(additionaloptionswilllikelybeidentifiedthroughEIAs).Therearemanystrategiestoreducedirectenvironmentalimpactduringconstructionand,toalesserextent,operationsofroadsandrailways.Manyconcentrateonadjustingthedesignoftheinfrastructure,butothersfocusonlocationandtiming.Herewelistjustafewexamplesthataddressreducingtheimpactsfromarangeofdifferentenvironmentaldirecteffects.AnexhaustivereviewofstrategiestoreduceimpactsisfoundinvanderReeetal.(2015).PrescriptivegoodengineeringdesignstrategiesarealsodetailedinvoluntarysustainableinfrastructurestandardsdevelopedbyorganizationssuchasSuRe®,CEEQUAL,Envision®,andGreenroads®(seePart3andAppendix15).

A. LocalStrategiesforReducingDirectEffects:GoodEngineeringDesign

Strategiestoreducelocaldirecteffectssuchasabioticandwildlifeimpactsfrominvestmentsintransportoftenhavebeenfocuseduponchangestothedesignandtheconstructionofinfrastructure.Goodengineeringdesignsolutionsatboththeprojectandlandscapelevelareamongthemostfrequentstrategiestoaddresshydrologic,sedimentation,anderosiondynamics.Forexample,tunnel‐bridge‐tunnelengineeringschemescanbeinstitutedtoavoiddangerouslandslidesorerosionalongsteep

6Theterm“minimize”alsoisoftenusedalongsideorinsteadof“reduce,”forthesecondelementofthismitigationhierarchy.

Asitishardtominimizemorethanoneobjective,andastrulyminimizingmeansavoiding,weusereduceasourdefault.

terrain(Wangetal.2012).Similarly,engineerscaninstallwildlifecrossingstructuressuchasbridgesandunderpassestoreducewildlifecollisionsandbarrierstomigrationwhileincreasinghabitatconnectivity(Iuelletal.2003;LitvaitisandTash2008).Underpassesincludetunnels,culverts,andviaductsbeneathraisedroadwayswhileoverpassescanrangefromsmallcanopyropecrossingsforarborealprimatestolargevegetatedlandscapebridgesupto50mwide(Iuelletal.2003;vanderRee,Smith,andGrilo2015).Suchstructuresmustbecoordinatedwiththefunnelingofanimalstotheappropriatecrossingpoints.Themostcommonfunnelingtechniqueisafence(Iuelletal.2003;R.vanderRee,Gagnoy,andSmith2015;D.J.Smith,vanderRee,andRosell2015)preventingwildlifefromenteringaroadorraillineanywhereotherthanatsafecrossingpoints.Thisisespeciallyimportantforreducingwildlifeaccessto“roadkillhotspots”alongroadsorrails–wheresomephysical,biological,ortopographicalfactorincreasesthelikelihoodofwildlifecollisions(LitvaitisandTash2008).

Arangeofdesignoptionsusingbehavioralmodificationsexisttoreducenegativeimpactsoftransportcorridors.Focusingartificiallightingdownwardreducesbirdattraction(Blackwell,DeVault,andSeamens2015)andconstructingsoundbarriersreducesstressorinterferencewithanimalcommunication(Parris2015).Retentionoftreesretainshabitatfeatures(SoanesandvanderRee2015;Weller2015),whilespecializednoiseandlightspectracanbeusedtodetercertaintargetedwildlifespeciesfromaroadorrailwaytoavoidcollisions(D’AngeloandvanderRee2015).Recordingsofnaturalanimalwarningcallstriggeredbyoncomingtrainshavebeenshowntoelicitfasterreactiontimesandincreasedescapebycertainwildlifespecies(Babińska‐Werkaetal.2015).Incontrast,commonlyusedroadsidereflectorshavehadlimitedsuccessindeterringwildlifefromenteringroads(Angeloetal.2006;Iuelletal.2003).Timingofconstructioncanalsohelpreduceimpacts.Forinstance,Constructioncouldbescheduledtoavoidtimesoftheyearwhencertainspeciesareespeciallymobile.Amphibiansareespeciallyvulnerablewhentheymoveacrossthelandscapeduringtheirbreedingseasonandbirdswhentheymigrateforwinteringandnesting(Andrews,Langen,andStruijk2015;Hamer,Langton,andLesbarreres2015).Wildlife‐vehiclecollisionrateshavebeenshowntobehigherduringmigrations(LitvaitisandTash2008).Timingconstructionortraffictoavoidtimesofhighwildlifemigrationcanreducecollisionincidence.Somedirectimpactsresultfromhumanbehaviorandarebestreducedthroughenvironmentalpolicies,regulations,andenforcement.Forexample,roadsandrailprovideaccessforillegalloggingandwildlifepoachingandprovidetransportationforillegaltraffickingofthiscontraband.EnvironmentalpoliciesandenforcementtoaddresstheseactivitiescanbeimplementedintandemwiththeBRIinvestment.Environmentalpoliciesmaybemoreeffectiveinfluencingimpactsfromsometransportationmodesversusothers.Forinstance,airborneparticulatepollutionandGHGemissionsfromincreasedfreighttrafficarerelativelyeasytoreduceinrailprojects–viahighefficiencyengines,efficientscheduling,andappropriatemaintenancewhichlimitsemissions–whileemissionsfromroadprojectsarelesseasilyregulatedbecauseofthedispersednatureofrelatedpollutionsources.EIAstypicallycallfornationalmeasuresregarding,forinstance,regulationsonfueltypeandpurificationofexhaustgases,vehicleemissionregulationsandstandards,andthelocalenforcementoflawsconcerningvehiclecondition.

B. BroaderStrategiesforReducingIndirectEffects:ComplementaryPolicy

Reducingindirectimpactsfromtransportationprojects–suchasinducedlandusechanges–islessstraightforwardthantheinfrastructuredesignmodificationstrategiesemployedtoreducedirectimpacts.Complementarypoliciesarethepreferredstrategyformodifyinghumandecisionmaking.

i.ComplementaryPoliciesInteractwithTransportationInvestments

Muchastheheterogeneityofimpactsfrominvestmentsintransportationinfrastructurealreadyleadstogainsfromroutechoice,thetargetingofpoliciesalsoisraisedbyinteractionsbetweentypesofpoliciesand,inparticular,interactionsbetweendevelopmentinvestmentsintransportandconservationpolicy.Theoptimalpolicyblendcangeneratea‘win‐win’outcomefromintegratedearlyplanningasitcanraisetheconservationimpactsofconservationpolicyresources,whileloweringtheenvironmentalrisksfromdevelopmentinvestmentsandstillpreservingtheirraisond’être,i.e.,benefitsofeconomicconnectivity.

Considerthatprotectedareasorpaymentsforecosystemservicesoftenhavenoteffectivelyaddressedenvironmentalthreatsandthushavelowimpact(seetheliteraturediscussionwithinIV.Offsetsbelow).Inbothcases,theproblemisthatimplementationavoidsthreats.Forprotectedareas,thisiscommonwheneconomicinterestsdeflectintendedprotection,sothatprotectedareasendupfarfrompressure,whereevenperfectenforcementhaszeroimpact.Forpaymentforecosystemservices,thisoccurswhenprivateactorsarepaidtomaintainforestonunprofitablelandstheywouldhavekeptinforestanyway.

Incontrasttothosescenarios,withgoodintegratedearlyplanningpublicactorsmayconsciouslyplaceinvestmentsintransportandconservationneareachother,e.g.,protectedareasneartoroadsorrails.Forthelongerrun,thathasimpacteveniftheroadorrailinvestmentsaretraversingrelativelypristineareas(e.g.,betweendistantbigcities)notpreviouslyorcurrentlyfacinghighthreat.Thefutureclearlycouldholdhigherthreats,asinvestmentscreateprofitableopportunitiesthatlurelaborandcapitalaswellaspublicinvestmentsasinschoolsorhealth.Assuch,previouslylow‐pressurefrontiersmayhavehighenvironmentalvaluewhichcouldbevulnerabletoimpactafterarailorroadlineisestablished.

ii.PublicandPrivateActions,Interactions,andFurtherResponses

Whileanysuchpolicyinteractions(abovewithexplicitjointtargeting)alwaysinvolvesomepublicroles,herewealsowanttohighlightthevalueofconsideringpossibleprivateresponsestopublicpolicies.Thoseprivatereactionsmayalso,inturn,affecttheoptimalfollow‐onpublicresponses.Generally,moresuchinteractionsagainsuggestthattherearegainsfromintegrationofdevelopmentandconservation.

Thelogicaboveconcerningprotectedareas,forinstance,couldbeappliedtopaymentstosuggestsitingpaymentsforecosystemservicesalongsideinvestmentsinroadsorrails.However,forlocallandownerslivingalongsidetheproposedBRIroadortracktovoluntarilyforgocuttingdownforestsonceanewrailorroadisinplace,wemustaskwhethertherearesufficientprivateincentivestojoinanysuchprogram.Forinstance,landsthatmighthavebeenreadilyofferedforconservationevenforalowpaymentbeforearoadorrailinvestmentoccursmaynolongerbevolunteeredaftertheinvestment(asissuggestedbythepastpurchasingoflandsasspeculativeinvestmentswhenpossibleplansforroadsaremadepublic).Yet,nonetheless,theremightexistproductivecombinationsoftransportinvestmentsandprotectionandpayments,atripleintegration,iftheplacingofprotectedareasimmediatelyaroundnewtransportlowerstheeconomicexpectationsforthoseparcelsnearbywhichareprivateandnotbeingprotected.

Wemightsimilarlyconsiderprivateincentivesifprotectionisnotstrictbut,instead,permitssomeuse–asinextractivereservesthatexplicitlyallowsomesmallholderproduction(andresultingforestlosses).Theyclearlychangethepoliticaleconomybymakingiteasiertohavebothprotectionanddevelopment.Thatcanmakeprotectionmorepoliticallyfeasible.Italsocouldgeneratelocalincentiveformonitoring.Thus,wemayconsiderwhattypesofconservationpoliciesarebestforinteractionswithroadsandrail.Further,inprinciplemuchthesamecouldbesaidforhavingprivaterightstologgingwithinconcessions.

Whileincasessuchrightshaveledtoratesofforestlosseshigherthaninthebackgroundfortheregion,rightscreateincentiveforactorstodefendforestassets–muchlikesmallholdersinextractivereserves.Forconcessions,actorsmaybefirms,whomayhaveincentivesandtheresourcestostopillegallogging.Generally,combinationsofdevelopmentandconservationtoolscouldinteractusefullyonthefrontier.

Sometimes,however,developmentandconservationwouldnotappeartoworkwellneartoeachother.Forinstance,ifastrongfederalgovernmentenvironmentalagencyestablishesastrictPAnearbyaroadorrail,itmaywellfunctionasasignalthat,inthefuture,moredevelopmentinvestmentisunlikelyinthisarea.Suchconservationsignals–perhapsfederal,thoughlocalenforcementisveryimportant–canencouragenetprivateoutmigrationdecisions.Inturn,thosecoulddiscourageanyfurtherpublicactions.Forinstance,whenin‐migrationslows,lesspayoffmaybeperceivedfrommaintainingoldersmallroads.Thus,publicactionscan,viaprivateresponses,leadtocomplementarypublicactions(Herrera2015)andinthisscenario,thepublic‐private‐publicresponsedynamicservestoimprovethelocalforestoutcomes.

Inothersettings,localpublicactorssuchasstateormunicipalgovernmentsarefocusedmoreuponthelocaleconomicgainsfrompoliciesthantheenvironmentaloutcomes.Thatcanremoveenvironmentalcomplementarityacrosspublicactions.Federalactionsmightconstrainlocalagencies’choices−whicheffectivelymaximizelocalinterestssubjecttofederalpolicies.Forinstance,whenafederal“blacklist”intheBrazilianAmazonconstrainedhigh‐deforestationmunicipalities,morethanonelocalprogramwithinternationalfunding(throughTheAmazonFund,e.g.)didnotachievestatedobjectivestolowerlocaldeforestation.Thatisunsurprisingiflocalprogramswereessentiallytohelpthelocalityinquestiontogetoffofthecostlyfederalblacklist,i.e.,effectivelythelocalprogramswereaimingsimplytoachievethefederalrequirementsyetatalowerlocalcosttotheeconomy(Correaetal.2018;Sillsetal.2015).Steppingback,though,eventhatcanhelpintegratedplanningbyincreasinglocalpoliticalfeasibility.

III.RESTORE:ActivitiestoNeutralizeNetTransportImpactsLocally

Restorationisconductedatprojectsitestorepairdirectorindirectimpactstobiodiversityorecosystemservices.Restorationisonlyforwhenavoidanceandreductionarenotfeasible(Ekstrom,Bennun,andMitchell2015).RestorationforBRIroadandrailprojectswouldprimarilyoccurduringtheconstructionphaseandactionswouldlikelybeidentifiedduringtheinitialfeasibilitystudywithrestorationstrategiesoutlinedintheEIA.Ifextensivelandscapetransformationoccurs,forexample,throughdiggingtunnels,stabilizingslopes,orbuildingembankments,thenvegetativerestorationwilllikelyberequired.Restorationmayalsobeneededifroadsorrailconstructiondisruptwetlandsandwaterways.

IV.OFFSET:CompensationElsewhereforUnavoidableImpacts

Offsetshavebeendefinedas:“Measurableconservationoutcomes,resultingfromactionsappliedtoareasnotimpactedbytheproject,thatcompensateforsignificant,adverseprojectimpactsthatcannotbeavoided,minimizedand/orrehabilitated/restored”(Ekstrom,Bennun,andMitchell2015).Offsetsarealastresort,afterallotherstepsinthemitigationhierarchy.Theyinvolverestoration,rehabilitation,orprotectionofbiodiversity,carbon,water,orotherecosystemservicescomparabletotheprojectlosses.Offsetsmayaimfornonetloss.Recently,somenetgainhasbeenpromotedtoaddressglobaldeclinesinbiodiversityandecosystemservices.Inprojectfinance,theInternationalFinanceCorporation(IFC)specifiestheneedforanetgaininbiodiversityincriticalhabitatasarequirementforlending(IFC2012).ThenewWorldBankframeworknotesapreferenceforanetgainofbiodiversityaswell(TheWorld

Bank2017).Over75majorfinancialinstitutionshaveexpressedthesamepreferencesviaadoptionofTheEquatorPrinciplesforprojectfinanceofoverUS$10million(Raineyetal.2015).Yettherearegoodreasonstoconsideroffsetsalastresort.Theycanbecomplexandexpensiveandyetofuncertainbenefit(Bulletal.2013;IUCN2016;PilgrimandEkstrom2014).Somearguethatoffsetsgivedevelopersafreepasstobothdestroyanddegradebiodiversity(Ledec,Campos,andReay2016)and,further,thatimplementationchallengesmayinfactpreventeffectivenetmitigation.Anothersignificantissueiscompensationofthosewhosufferloss.Ecosystemservicessuchaswatershedservicesarelocalandthusoffsetsmaybehardpressedto“makewhole”theaffectedcommunitiesifgainsarebeingrealizedinotherlocations.Projectlossesfollowedbyoffsetgainscouldinvolvetrade‐offswheresomeservicesarereducedthenothersareenhanced(e.g.,bysoilstabilizationaftertimberproductionortherelocationofwateraccess)and,asaresult,someportionsofacommunitycouldgainonnetwhileotherscouldstillloseoutonnet.

Belowweconsideranumberofapproachestogeneratingoffsetswhichcould,inprinciple,balanceoutthenetlossesthatmayoccurastheresultofatransportationinfrastructureproject–oncetherailorroadprojecthasbeenselectedandroutedanditsdamagesreducedandrestoredasmuchaspossible.

A. BiodiversityandHabitatOffsetsWhenaneworupgradedrailorroadprojectresultsinthedestructionofordamagesto

biodiversity,offsettingcanbeaccomplishedbyrestoringorenhancingcomparablebiodiversityelsewhere.Anoffsetshould“achievenonetlossandpreferablyanetgainofbiodiversityonthegroundwithrespecttospeciescomposition,habitatstructure,ecosystemfunctionandpeople’suseandculturalvaluesassociatedwithbiodiversity”(BBOP2009).Biodiversityoffsetshavetakentheformofnewprotectedareas,restorationofdegradedordestroyedhabitat,orevenincreasedenforcementofexistingprotectedareas(G.Bennettetal.2017;Ledec,Campos,andReay2016).Biodiversityandhabitatoffsetsmayfocusonparticularspecies(e.g.,endangeredorthreatened)orhabitattypes(e.g.,thosehostingparticularspeciescommunities)orfunctions(e.g.,wetlands/streams).

Biodiversityoffsetsarerelativelyunproven,thusitwillbeimportanttodesignsuchprogramswelltoensureimpactandmonitortheirperformance.Thereisnobestwaytodesignabiodiversityoffsetbecausetheyaresodependentoncontext,yettherearebestpractices(IUCNandICMM2013).Forinstance,twointerrelatedelementsthatmanybelievewillenhancesuccessinoffsetsare:planningforoffsetswithinthelandscapecontext(BBOP2012;McKenneyandWilkinson2015);andsettingupaggregatedbiodiversityoffsets(Ekstrom,Bennun,andMitchell2015;ICMMandIUCN2013;Ledec,Campos,andReay2016).TheseprinciplessuggestthataBRIEconomicCorridormayallowforplanningattheappropriatescale,withbroadandearlystageenvironmentalassessmentstheappropriatetooltoplaneffectiveoffsets.Aggregatedoffsetscanreducetransactioncostsandconsiderinteractionsbetweenprojectstoenhancebenefitslikelandscapeconnectivity(Ekstrom,Bennun,andMitchell2015;Ledec,Campos,andReay2016)(PilgrimandEkstrom2014;BBOP2009).Onecouldenvisionaroadmapforanoffsetscorridor,eveniftheBRInetworkpassesthroughmultiplecountries.Fewnationalexamplesexist,however.SuchroadmapscanbeseenforLiberia(TheWorldBankGroup2015),Mozambique(BechtelandNazerali2016),andMongolia(TNC2016)thoughthesearenewenoughthattheirsuccesscannotbedetermined.Amajorblockageseemstobetheseedfundingrequired(Ekstrom,Bennun,andMitchell2015).

FinancinginstitutionssuchastheWorldBankandIFChavemandatorybiodiversityoffsetpolices.Thesepolicieshavebeena“smallbutsignificantdriverinoffsetdemand.”Asof2017,9IFC‐financedprojectscreatedbiodiversityoffsets,with12moreprojectsplanningtoimplementoffsetsinthefuture(G.Bennettetal.2017).IftheWorldBankorIFCweretogetmoreinvolvedinco‐financingBRIprojects,theirbiodiversityoffsettingstandardswouldbeappliedtoanyprojecttowhichtheywerelending.

Offsetscanbeonlyaseffectiveastheirdesign,implementation,andactualenforcementallows.Clearanddetailedpolicyguidanceisrequired.Alackofoversightisaprimaryreasonoffsetsfailtoachievetheirgoals(Bulletal.2013;PilgrimandEkstrom2014).Sucheffectiveguidanceandoversightcould,inprinciple,beprovidedbyhostcountriesorbyBRIoversightorganizationsintheChinesegovernment.However,totheextentglobalfinanceisinvolved,theremaybegainsfromincludingoutsideparties.

Itishardtogeneralizethesuccessofmitigationforbiodiversityandhabitatbecauseprogramsmeasure

successusingdifferentmetrics,whichmakecomparisonsortheanalysisoftrendsdifficult.Analysesof

large,standardizedprogramsliketheUSstreamandwetlandmitigationprogramhavefoundthatonly

30%ofoffsetsaremeetingfullprojectobjectives(MatthewsandEndress2008;n=76sites)and74%of

offsetsareachievingnonetloss(BrownandLant1999;n=68banks).Withthisinmind,itisclearthat

thereishighuncertaintyofthesuccessofbiodiversityoffsets.Manyprojectssimplydefinesuccessas

establishingaprotectedarea;asdiscussedelsewhereinPart2,thathasawidevarietyofoutcomes.

B. OffsetInstruments:ProtectedAreas(PAs)

Themostcommonconservationpolicyintermsofareaaffectedistheestablishmentandmaintenanceofprotectedareas(PAs).Theycontainover13%oftheworld’slandsand,inprinciple,couldgenerateoffsetsbypreservingforestcover,biodiversity,andecosystemservices.WhilePAstendtopreventforestlossonaverage,however,theirimpactvariessubstantiallybylocation(Andametal.2008;L.N.JoppaandPfaff2011;Miteva,Murray,andPattanayak2015;A.Pfaffetal.2009;A.Pfaff,Robalino,Herrera,etal.2015).Further,aPA’simpactoftenislessthanitseems,sincesitesarebiasedtowardslow‐pressurelocations.7Thismattersintermsofappropriatequantitativeestimatesofoffsets.

TheoffsetsaPAcangeneratearelimitedbythedeforestationthatwouldhaveoccurredwithoutthePA.Thus,evenaperfectlyimplementedandenforcedPAsimplywouldnotavoiddeforestationifitislocatedinanisolatedareawherenodeforestationwasexpectedtooccur.Generalizing,manyfactorsaffectthenetbenefitsfromnon‐PAlanduses,i.e.,aPA’sopportunitycosts(foregoneprofits),whichcreateshighvariationinthecounterfactualsorbaselinelandusesthatthePAcouldblockwhencreatinganyoffset.

Oftenthemostbeneficialnon‐PAuseofafrontierparcelisagriculturaluse,inwhichcasethelandswithsteeperslopes,poorersoils,andsitesfartherfromtheimportantmarketshaveloweropportunitycosts.

7Forlow‐pressurelocations,itiseasytooverestimatePAimpact.Forinstance,ifcomparingPAswithrandomlychosencontrols

oraverageunprotectedland,low‐pressurePAsarebeingcomparedtohigherpressure.Andametal.(2008)findforsomeCosta

RicanPAs,“applestoapples”(usingobservablelandcharacteristics)comparisonsreducetheestimatedgainsfrom44%to11%.

Globally,JoppaandPfaff(2010)findcontrolsforlandcharacteristicsreduceestimatedimpactonaveragebyaboutone‐half.

WewouldexpectthelobbyingeffortsagainstPAsinthoselocationstobelower.8However,itispossiblethatthemostbeneficialnon‐PAuseofthelandis,instead,forhydropower–inwhichcasesteeperslopecouldincreasetheopportunitycostsofmaintaininganundisturbedPA.Indeed,weobservemanycasesinwhichcreationofhydropowerdamshasledtothedegazettementofexistingPAs(Symesetal.2015).

Thus,understandingtheland‐usedynamicsinagivensettingiscriticalforgoodestimatesofabaselineorcounterfactualland‐usetrajectorythat,comparedwiththeoutcomeswithinPAs,determinesoffsets(inthesenseofcomputinghowmuchareathatwasgoingtobedeforestedwasconservedbythePA).Offsetcalculationsarelikelytoalsoincludecomputationsconcerningthedensityofadesiredservice–e.g.,ahectareofforestinafavoredhabitatmaybeweightedmorehighlythanotherforesthectaresoravoideddeforestationthataffectswaterqualityupstreamoflargecitiesmaybeespeciallyappreciated.

DistinguishingacrossPAs,thepoliticalcalculusinvolvedinsitechoicefornewPAs,aswellaschoicesonbothmonitoringandenforcement,islikelytovarywiththelevelofgovernmenttakingthePAdecisions.Variousstudiesindicatethatfederaldecisionsontheenvironmentcanbeexpectedtodifferfromlocaldecisions,sincelocalactorsareexpectedtoputmoreweightonlocalcosts(A.PfaffandRobalino2012).OnecanimaginethatdecisionswouldalsodifferamongcountriesacrossaBRIcorridor.ThereislimitedsuchevidenceforPAs,ingeneral,thoughfortheBrazilianAmazonthispointappearstobesupported.

Further,thelocalandglobalbenefitsandcostsfordifferentrelevantactorsinoffsetsvarywithPAtype.StrictPAsthatdonotpermitanyentryorproductionvaryintheirnetbenefitsfrommultiple‐usePAsorsustainablereservesthatpermitsomesmallholderproduction.Offeringsomelocalbenefitscouldeasilyinfluencelobbyingpositions.Consequently,multiple‐usePAsmaybemorelikelytobeinlocationswithrelativelyhighforestpressure(forglobalevidence,seeNelsonandChomitz(2011);foraparticularcaseinBrazil,seePfaffetal.(2014);andforanintermediateregionalresultforPeru,seeRicoetal.(2018)).WhetherthatimplieshigherimpactsforonetypeofPAvaries,evenifmixed‐usePAsareinareaswhicharemorethreatened(Blackman2015;Ferraroetal.2013).Notsurprisinglythedynamicsvary(see,e.g.,Pfaffetal.(2015)andPfaff,Santiago‐ÁvilaandJoppa(2017)).Understandingland‐usedynamicsiskey.

Thesamelogicconcerninglocalbenefitsandcostscanalsoaffectmonitoringandenforcement.AswasraisedbyAlbers(2010),localactorswithincentivestokeeptheforestsstandingcancontributenotonlythroughtheirownland‐usechoicesbutalsobycontributingtomonitoringofothers.Thus,forexample,themultiple‐usePAswhoselocationsarehighlightedabovemightalsobebetterlocallymonitored.Onthethemeofprivateincentivestokeepforestsstanding,itisevenpossiblethatloggingconcessions–whoserightscreateincentivetodefendforestassets–couldoutperformPAsintermsofavoidinglossshouldthosePAshavefewerresources,lessincentives,andpoorenforcement(Panlasiguietal.2017).Consequently,onboththepositiveandthenegativeside,inestimatingtheoffsetsthatarelikelytobegeneratedbyanygivenconservationinvestmentofthistype,theprivateincentivesclearlydomatter.

C. OffsetInstruments:PaymentforEcosystemServices

PaymentsforEcosystemServices(PES)areavoluntaryinterventionforpreservingforestcover(Engel,Pagiola,andWunder2008;FerraroandSimpson2002;Wunder2007).Theirtypicaleffectivenessintermsofavoidedforestlosshas,todate,oftenbeenrelativelylow.Thus,attheleastdependingupon

8GivenvariationinthebenefitsandcostsofPAs,andwhichgroupsinasocietyincurthem,weexpectthatpoliticaleconomy

willaffectthelocationaswellasmonitoringandenforcementofPAs.Ifweholdbenefitsconstant,wepredictthatPAsshould

gotothelocationswiththelowestopportunitycosts.Globally,ithasbeenconfirmedthatPAsaredisproportionately“high,

steepandfar”(L.N.JoppaandPfaff2009).Moregenerally,characteristicsthatloweragriculturalopportunitycostsappearto

attractsitingofPAs.

programtargetingandimplementation,onemaynotbeabletocountuponPESforsignificantoffsets.Paymentforwatershedservices(PWS)–aspecifictypeofPES–offsetserviceslikesedimentretention,pollutantfiltration,andwaterretention(toreducefloods)thatareaffectedbyconstructionandoperationofroadsandrail.PWSor“waterfunds”offerpaymentsfromwateruserssuchaswaterutilitiestolandownersorusersinexchangeforimprovingagriculturallandmanagement,includingriparianbufferstoimprovewatershedmanagement(LocatelliandVignola2009).

EvenmorethanforPAs,additionalforestbeyondbaselineishardtoguaranteewhenPESparticipationisvoluntary.Individualswhowerenotgoingtocleartheforestonaparceloflandanywaywouldrationallybethefirstinlinetovolunteerthatparcelforacontractthatpaysmoneyiftheforestremainsstanding.ThatcanexplainlowimpactestimatesforPES(J.RobalinoandPfaff2013;Sanchez‐Azofeifaetal.2007).

Estimatesofgainstendtobemoreencouragingiftargetingisnotentirelyuptoindividualsbutratherisdeterminedbyotherfactors,e.g.:biologicalfactors,asintheprotectionofMonarchbutterflyhabitatsinMexico(Honey‐Rosés,Baylis,andRamírez2011);orbygovernmentactorsorcivilsociety,dependingontheirformaloreffectivedecisionrules;orrandomly,asinatrialinUgandawherethelocationfeaturedhighdeforestationforlowopportunitycostsandindividualshadtoputalltheirparcelsintothePES(Jayachandranetal.2017).China’sSlopingLandConversionProgramissaidtobeimpactful,althoughnotquitevoluntaryevenifadequatelycoveringlocalopportunitycosts(yetdistortionsmayinterferewithtargeting,dependingonincentivesoflocaladministrators(J.Xuetal.2010;Z.Xuetal.2004)).Underthisprogram,paymentsforremovingslopedlandsfromagriculturalproductionalsoresultedinmorelaborersswitchingtonon‐farmemployment(M.T.Bennettetal.2014;KellyandHuo2013).

ContractdetailsofferotherdimensionsofPESdesign.PEScontractsinvolveasymmetricinformation,asland‐holdersknowtheirownopportunitycostsandthereforethepriceatwhichtheywouldbewillingtoacceptaPEScontract.AgovernmentorNGOdoesnothavethisinformation(Ferraro2008),whichraisescostandthechancethatsomeofthefundingdoesnotleadtoadditionalforest.Thatcandiscouragethecreationofnewprogramsandcouldbeaddressedbyauctions.PESdesignsthatpaymoretothefarmerswhosepreviouswillingnesstocontributetoconservationwaslow(suggestinghighopportunitycostsofconservationandtheneedforincentives),however,canintroduceperceptionsofunfairnessandlowercontributionbyparticipantswhowereoriginallymorewillingtocontribute(Alpízaretal.2017b,2017a).

Collectivecontractsshowsomepromise,ifdesignedwell,usinggroupenforcementandsocialpressuretopotentiallyincreasecompliancewhileloweringthecostsofadministration(D.Kaczanetal.2017).InformationonbaselinelanduseandbehaviorunderthePESislowercostatahigher,collectivescale.Forinstance,whilenotlabeledas“PES”,theBrazilianAmazon“blacklist”mentionedaboveprovidedaconditionalincentivefromthefederalgovernmentatthecollectivescaleofthemunicipality–someofwhicharehugeandallofwhichhavepre‐existingpoliticalinstitutions–tobemonitoredbysatellite.

Despitechallenges,PEScouldofferaneffectiveintervention,especiallyinconcertwithotherpolicies.FundstransferswithinPESmaybeoneelementinbundlesofinterventions,helpingtomakeelementsthatarelocallycostly,suchasstrictPAs,morepalatable.Forgeneratingoffsetstoo,then,combinationsofconservationpoliciesmightusefullybeconsideredattheleveloftheentireBRIEconomicCorridor.

D. CarbonOffsets

Carbonstorageandsequestrationprovidesanimportantclimate‐stabilizationecosystemservice(TheWorldBank2017).Lossofforests,wetlands,andotherhabitatsduetoconstructionandoperation

oftransportationinfrastructure,though,canreleasecarbonandreducesequestration.Thus,offsettinghereshouldincludeclimatestabilization,i.e.,compensatingfortheemissionsandlossofsequestration.Todate,thereisoneexampleofoffsetsrequiredforlossofGHGsinksorincreasesinemissionsdueto

infrastructure,agreenhousegasregulatoryregimeintheStateofCaliforniaunderstateenvironmental

protection(CEQA),whereonemustassess,minimizeandoffsetshiftsincarbonsinksandemissions.

However,itisnotenoughtodoactivitiesandclaimoffsetsifactivitiesdidnotactuallyloweremissions.

StudiesofcarbonoffsetssuchasfromprojectsintheKyotoProtocol’sCleanDevelopmentMechanism–theworld’slargestcarbonoffsettingprogram,withmanyenergyandindustrialtypesofoffsets–showa

lackofimpact(Haya2010).Clearly,somemethodsforcalculatingcarbonimpactsandcarbonoffsetsare

imperfect(HarangozoandSzigeti2017;Johnson,Edwards,andMasera2010).Itisstillthecase,though,

thatwell‐designedandstringentlyenforcedprogramscouldloweremissionsandevenhaveco‐benefits

(e.g.,someforest‐carboneffortsintheU.S.fortheCaliforniamarket(Anderson,Field,andMach2017)).

E. SpilloversandNetImpacts

Asemphasized,itisachallengetoachieveimpactsviainterventionsofanytypeandpurpose–avoiding,reducing,oroffsetting–giventhatsimplyannouncingaprogramdoesnotchangebehaviors.

Simplycreatingaprotectedareadoesnotguaranteetherewillbenoillegalinvasionanddeforestation;

plus,evenifthetreesremainstanding,itdoesnotguaranteeanyimpactiftherewerenotanythreats.

Simplycreatinganincentivesprogramdoesnotguaranteethatanybodywillsignuporwillchangeany

behaviors.Ifoneispaidtokeeptreesstanding,onemightsimplyvolunteeragriculturallyuselessland.

However,sometimesinterventionswillinfactchangebehaviorsonthoselandswithintheirboundaries.

Unfortunately,eventhatdoesnotguaranteeanetimpact,sinceinterventionsmightgeneratespillovers.

Ifaprotectedareadoesinfactlowerclearingrelativetowhatwouldhaveoccurredwithoutprotection,

whichmanyhavedoneonaverageacrosstheworld,unfortunatelyitcouldstilleasilybethecasethat

theindividualswhowouldhaveclearedinsidetheprotectedareajustdisplacetheirdeforestation

elsewhere.Withinpaymentsprograms,evenifoneparcelthatwouldhavebeenclearedisnot,dueto

payments,thereisoftennothingpreventingalandownerfromjustshiftingtheclearingtoanothersuch

parcel.Shouldsuchspillovers– inthosecases“leakage”– occur,theycancanceloutanypositiveimpacts.

Conceptualmodelsofhowmarketdemandsforlandclearingplayoutacrossanylandscapecertainly

suggestthatPAscouldinducespilloverdeforestationintonearbylocations(Armsworthetal.2006).

EmpiricalestimatesofforestspilloversfromPAsfindheterogeneities(A.PfaffandRobalino2017;J.

Robalino,Pfaff,andVillalobos2015).Thatisinpartbecausewhenlackingimpact,PAswillnotspillover,

e.g.,ifthreatsarenottakenonthentheycannotbeblocked,forimpact,butalsowillnotbedisplaced.

Spilloverstendtobehigherdeforestationelsewhereifinterventionsarenearertoroads(J.Robalino,

Pfaff,andVillalobos2017)–whichisobviouslyrelevantwhenconsideringintegratedpoliciesthatplaceprotectedareasnearBRIprojects−andifpricesforcommoditieswhoseproductionisdisplacedarehigh

(Baylis,Fullerton,andShah2016).SpilloversmaybelessnegativenearPAentrances,whereecotourism

tendstoconcentrateandtoraisethevalueofintactforestparcels(J.Robalino,Pfaff,andVillalobos

2017),aswellaswhenanydisplacedcommodities’pricesarelow(Baylis,Fullerton,andShah2016).

Spillovereffectscanevenbepositive,environmentallyspeaking,forinstanceinthesensethatPAslower

netforestlossesnearby.Forinstance,ifPESsupportintroductionoftreesintoagriculture(agroforestry),

itmaysometimesbethecasethatlocalprofitsareinfactenhanced(e.g.,byshadeforcoffeeandcows).

Uponseeingsuchanincreaseinprofit,neighborsmightvoluntarilyimitatethoseagroforestrypractices.

Theexistingempiricalresultsfordeforestationspilloversfromprotectedareasindicatethatnotonlythe

size(magnitude)butalsoeventhesign(direction)ofspillovereffectswilldependoncomparativerents

(Angelsen2007)andpoliticalsignalingdependingonwhichagenciestakewhichactions(Herrera2015).

Forinstance,thecreationofastrictprotectedareabyastrongfederalagencythatisknowntocurrently

featureanenvironmentalagendacansignalthattheprotectedregionwillnotbeadevelopmentfocus.

That,inturn,shouldinfluenceprivatemigrationchoices,publicinvestments,andotherpublicactions.

Spilloversoutsideinterventions’boundaries(distinctfrommultiple‐usePAs)canalsobesocioeconomic,

emphasizingagainthattheycanvaryeveninsignbycontextandthuslocation.Anumberofstudiesfind

PAshaveweaklypositiveeconomicwelfareimpacts(Alix‐Garciaetal.2013;Sims2010;SimsandAlix‐

Garcia2015).Sims(2010)findsthatdecreasesinpovertymayresultfromPAs–primarilythosePAs

foundatanintermediatedistancefromurbanareassuchthattheyarefarenoughthatthePAisnot

substitutingforurbanlandusebutcloseenoughthattheseareasareaccessibletopayingtourists.In

othercontexts,thecreationofPAshasbeentiedtosignificantnegativesocioeconomicimpacts,duefor

instancetodisplacementofvulnerablepopulations(GeisleranddeSousa2001)andreducedwages(J.

A.Robalino2007)−anditisimportanttoconsiderthedistributionoftheassociatedcostsandbenefits.

Part3:PoliciestoImproveEnvironmentalNetImpactsofBRICorridors

Becausetransportationinfrastructuredramaticallytransformslandscapes,policieshavebeendeveloped

overdecadestoimproveenvironmentalandsocialperformance.InPart3,wereviewtypesofpolices−includinglaws,regulations,guidelines,agreements,standards,andsafeguards−thatcanaffecthowBRItransportationprojectsaddresstheenvironment.Totheextentpossible,wewillfocusonenvironmental

policiesthataredistinctlyrelevanttoBRItransportationprojects.Wefocusfirstonnationalpoliciesand

thenoninternationalpolicies,includingChinesepoliciesrelatedtodirectforeigninvestments.Finally,

wediscusshowstrategicenvironmentalassessmentscustomizedtoBRIEconomicCorridorsarewell

suitedtoeffectivelyandefficientlyaddresstransboundaryenvironmentalimpactsrelatedtoBRI

projectsandcouldimprovebothenvironmentalandsocialoutcomes.Whileourfocusison

environmentalpolicies,itshouldbenotedthatmanyofthesepolicies–especiallyinrecentyears–also

addresssocialissues.

I.NATIONALANDPROJECT‐LEVELENVIRONMENTALPOLICIES

A. OverviewofNationalEnvironmentalPolicies

Thefirstlineofenvironmentalprotectionsisofcoursechoicesbythehostcountries.Chinahaspledgedtofollowhost‐countrystandardsandnormsforallBeltandRoadprojects,soinprinciplehost‐countrypoliciesprovidetheminimalprotectionsexpectedforeveryinfrastructureproject(Dollar2018).

Oftheroughly80countriesinvolvedintheBRI,eachonehasitsownsetofenvironmentalpoliciesthatmustbeunderstoodtograspthefullregulatorylandscape.Thereisawiderangeofnational,provincial,andlocalregulatorypoliciesthatcouldhaveabearingonenvironmentalrisksfromtransportationinfrastructureprojectsincluding:environmentalassessmentlaws;environmentalprotectionlawsforpubliclands;environmentalprotectionlawsfromprivatelandsincludingeasem*ntlaws;primaryprotectedareaslegislation;criminalcodesaddressingviolationofenvironmentalprotectionlawsandrelatedobligation;wildlifelaws;constructionpermitting;wetlandsprotectionandmitigation;vehicleemissionsstandards;damageslawforhazardousmaterialsspillsandleaks;indigenouspeoplesrightslaws/territoryrights;zoninglaws;imminentdomainlaws;andnationalprocurementrules.Inadditiontonationallegislation,countriesmayhavepolicies,guidelines,andstrategiesforaddressingmanyoftheseissues.Itisalsoimportanttotrackvariouscomplementarylaws,regulations,andzoningordinancesfromprovincialorlocalgovernments,sincethesecansometimeshaveasmuchifnotmoreinfluenceoverBRIprojectsthanacentralgovernment(H.LiuandLim2018).Thereisagrowingbodyofresourcesthatdocument,archive,andanalyzeenvironmentallaws.TheGreenGrowthKnowledgePlatforms(GGKP)providesanexcellentmeta‐analysisof24databasesofenvironmentalregulations,financialincentives,overarchingpoliciesandframeworks,andvoluntaryapproaches(Booth2017).Overallthesedatabasesprovidecomprehensivecoverageoftheelectricity,agriculture,forestry,andfishingsectorsinOECDcountries.Coverageofothersectorsandregionsoftheworld–includingmostBRIcountries–islessexhaustive,unfortunately.LegalAtlasisbeginningtodevelopaninteractiveknowledgeplatformforallenvironmentalregulationsrelatedtoinfrastructureprojectsineveryBRIcountry,initiallyfocusingonlawsforlinearinfrastructure.Appendix10describesseveraldatabasesthatareparticularlyvaluableforassessingenvironmentalpoliciesthatmayhaveabearingonBRItransportationinfrastructure.Itshouldbenotedthat,whilehostcountrylawsmayconstrainalltheBRIinvestments,significantelementsoftheenvironmentalrisksthatareassociatedwithBRItransportcorridorsaretransboundaryrisks,includingimpactsuponcarbonemissions,forestcover,andbiodiversity.Limitingthesetransboundaryimpactscanfairlybeconsideredtobeglobalorregionalpublicgoodsprovision,soweshouldnotexpecthost

countriestofullytaketheseimpactsintoaccountintheirownlaws(Kaul,Grunberg,andStern1999).

B.EnvironmentalAssessments

Oneelementrepeatedlyfeaturedinmanyofthe

nationalregulationsandpoliciesistheenvironmental

assessment,amongthemostcommonlyusedplanningand

managementtoolsforaddressingenvironmentalrisksfor

manydecades.“Theabilityofcountriesandcommunities

toachievesustainabledevelopmentdependsinnosmall

measureonrobustandeffectiveEIA/SEAlegislationand

implementationasamajorcatalystforovercomingcurrent

implementationgapsandachievingbetterenvironmental

outcomes.”(UNEnvironment2018,pvi)Inthenexttwo

sectionswewilldescribetheuseofenvironmentalimpact

assessments(EIA)andstrategicenvironmentalassessment

(SEA)forBRItransportationprojects.

EIAsaretypicallyfocusedonaspecificdevelopment

project,suchasanewhighwaysegmentorrailway.The

EIAdictatesaformaladministrativeprocessthatisaimed

atpreventingharmtotheenvironment,oftenalsotaking

intoaccountrelatedsocialconsiderations(Figure3).EIAs

shouldbeginintheprojectplanningphaseandthuspriortoanactivitytakingplace.TheEIAidentifies

potentialdirect(andlessoftenindirect)environmentalimpactsofaprojectandtrade‐offsbetween

policygoals.EIAsareaimedatinformingdecision‐makingonhowtoproceed(UNEnvironment2018a).

Inrecentyears,ithasbecomeincreasinglyevidentthatEIAshavemanyshortcomings.BecauseEIAs

oftenoccur“downstream,”wellintotheplanningprocess,mostdecisionsandcommitmentshave

alreadybeenmadeandpartiestendtobevestedinparticularoutcomes(OECD2006).Moreover,the

continualratchetingupofEIAstandardsbytheInternationalFinanceCorporation(IFC),WorldBank,and

othermultilateraldevelopmentbankshashadtheperverseeffectofpromotingborrowingfromother

sourceswithlowerEIAstandards,evenifitmeanshigherlendingrates(Humphrey2016).

Toaddresstheseshortcomings,governmentsandmultilateraldevelopmentbankshaveincreasingly

beguntoencourageorrequirestrategicenvironmentalassessment(SEAs)–early‐stageanalysesthat

integrateenvironmental(andoftensocial)considerationsintopolicies,plans,andprograms–rather

thansolelyrelyingonthesingleproject‐basedEIAs(WorldBank2012).Bycomparisontotheproject‐

focusedEIA,SEAsaimtointegrateenvironmentalconsiderationsintopolicies,plans,andprograms(Dusi

andXi2009;OECD2006).SEAsareoftencarriedoutattheregionalornationalscaleorforanentire

sectorsuchastransportation.Theycanalsofocusontransboundaryprojectsintwoormorecountries.

Mostsignificantly,SEAsarecarriedoutearlyintheplanningprocesswhenalternativetransportation

routingisstillfeasible(Figure4),andthefullmitigationhierarchycanbefollowed.(SeePart2fora

discussionofthemitigationhierarchy.)SEAshavebeencharacterized,tosomedegree,asforward‐

looking,“sustainabilitydriven”instruments,whereasEIAsaremore“reactive”(Wingardetal.2014).

Figure3:EIAprocessflowchart.Source:

UNEP2018.

WhileSEAsareincreasinglybeingadoptedand

strengthened,ascomparedtoEIAstheyarestill

inanascentstageofdevelopment.Atleast40

countrieshaveSEAsystemsinplaceincluding

China,Vietnam,Indonesia,Mongolia,Kenya,and

Tanzania;however,onlysomeofthese40

countriesincludeaformallegalrequirementto

conductSEA(UNEnvironment2018a).These

assessmentsaredesignedtobeflexibletoallow

analysistomeettheneedsofdecisionmakers,

butasaresultthereisuncertaintyand

inconsistencyinapplication(UNEnvironment

2018a).InChina,SEAimplementationhasbeen

limitedbythesharingofinformationanddata,

decision‐makingprocess(especiallythetimingof

SEAplanning),andlegislativeandpolitical

context(T.Lietal.2016).GunnandNoble(2011)

notethatSEAmethodologiesarestill

inadequatelydevelopedanddefinedandarenot

capableofaddressingregionalcumulativeeffects

adequately.Chakeretal.(2006)foundchallengesidentifyingappropriateleveragepointsinplanning

anddecision‐makingprocessesforSEAinput,limitingintegrationofSEAfindingsintodecisionmaking.

DespitetheselimitationstheWorldBankfoundsufficientevidencethatSEAscaninfluenceandinform

decisionstoencouragetheircontinueduseinbankpractice(WorldBank2012).

ForSEAstofulfilltheirpromiseasaformatforstrategicplanning,theywillneedtobeconductedwithin

anintegratedeconomicplanningprocess.TheInter‐AmericanDevelopmentBank’sguidanceconcerning

sustainableinfrastructurecontainstwokeyobservationsrelatedtoSEAimplementation:sustainability

shouldbeconsideredearly,“upstream”inthepolicyandplanningstageofinfrastructuredevelopment,

andenvironmentalconsiderationsshouldbeincorporatedintothemainstreamoftheplanningprocess

sothateconomicdevelopmentandenvironmentalassessmentarenotconsideredwithinseparatesilos

(Serebriskyetal.2018;UNEnvironment2018b;Watkins2014).Strategicintegrationofenvironmental

assessmentwithinfrastructureplanninghasthepotentialtodecreasethecoststhatareassociatedwith

mitigationofenvironmentalharmbyaddressingconflictsbetweenenvironmentalgoalsandeconomic

objectivesatanearlierstage,whenmoreoptionsremainonthetable.InsofarastheSEAframework

encouragesearly,integratedapproachestoplanningforsustainabledevelopmentitappearstobea

promisingvehicleforimprovingenvironmentaloutcomesassociatedwithinfrastructureinvestment.

C. EnvironmentalPolicyCompliance

Theadoptionofcomprehensivepolicieswithappropriateguidanceonimplementationand

monitoringarecriticalfirststepstowardpositiveenvironmentaloutcomes.However,thesepoliciesare

onlyaseffectiveastheirenforcementmechanisms,andthecapabilityandpoliticalwilltoenforcethem

(Leungetal.2013;UNEnvironment2018a).BRIcountriesnotonlydifferintheirenvironmentalpolicy

frameworks,buttherearealsoenormousdifferencesintheircapabilitiesandpoliticalwilltoenforce

Figure4:StrategicEnvironmentalAssessment

processflowc4art.Source:UNEP2018.

environmentalpolicies.Limitedinstitutionalcapacityandgenerallyalackoflegalrequirementsoften

limitfollowuponbothEIAsandSEAs(UNEnvironment2018a).GiventhatEIAsoccurlateinthe

planningprocess,thelikelihoodofcancellingaprojectormakingdramaticalterationsasaresultof

findingsintheenvironmentalassessmentreportislow.Therearetypicallynoprovisionsinnational

legislationregardingtheprocessforcasesofnon‐compliancewithanenvironmentalreportorinacase

whenmitigationmeasuresprovetobeineffective(UNEnvironment2018a).

D. ComparisonsofNationalEnvironmentalPoliciesandCompliance

Countriesvarytremendouslyinthecoverageandstrengthoftheirenvironmentalandsocial

policiesthataddresstransportationinfrastructure.Cross‐countrycomparisonsmostlyevaluatethe

differencesamongnationalenvironmentalassessmentpolicies,inpartduetotheirwidespreaduse.A

globalreviewbyUNEnvironmentofEIAsandSEAsfoundthat,thoughvirtuallyeverycountrystudiedin

thereport9hasadoptedEIAsystemsbasedonlegislationandanincreasingnumberofcountriesnow

haveSEAsystemsinplace,thedepthandapplicationvarysignificantlyamongcountries.Thereport

identifiedtrendsinEIAnationallegislation,including:amovementbysomecountriestowards

decentralizedoversightandimplementationofEIA,thoughasmallgroupofcountrieshavemovedin

theoppositedirectionbyestablishingdedicatedcentralauthoritiestodealspecificallywithEIAs;

increasedpublicparticipationrequirements(butmostlylimitedtothescopingandreviewstages);

increasedfocusonclimatechangeandhumanhealthand,thoughnon‐binding,ecosystemservices;and

ashiftfromEIAstoEnvironmentalandSocialImpactAssessments(ESIAs).EmergingtrendsamongSEA

legislationinclude:theincreasedadoptionofnon‐bindingproceduralSEAguidelinesbutrarelylegal

requirements;littleattentiontopublicparticipation;andregionalratherthannationallevelguidelines

tofacilitateassessingimpactsrelatedtoclimatechange,biodiversity,ecosystemservices,andhuman

health(UNEnvironment2018a).

Inamoretargetedlegalanalysisaimedatdevelopingguidelinesformitigatingtheimpactoflinear

infrastructureonlargemigratorymammalsinCentralAsia,Wingardetal.(2014)comparedcomponents

ofEIAandSEAlegislationasitrelatestolinearinfrastructureineightCentralAsiancountries.10Overall,

theyfoundthat,althoughalleightcountrieshadEIAlegislation,therewerefewdirectreferencesto

linearinfrastructure,migratoryspecies,ortransboundaryimpactswithinanyofthelegislation.EIAsin

fivecountriesmentionedwildlifeimpacts,butonlyKazakhstan’slegislationexplicitlymentioned

migratoryspeciesandrequirestheirconsiderationduringtheconstructionoflinearinfrastructure.

Kazakhstan,Kyrgyzstan,Tajikistan,andtheRussianFederationallrequiredtheassessmentof

transboundaryimpacts.IntermsofSEAs,allCentralAsiacountriesexceptRussiaandChinahave

nationallegislationforbothplansandprogram,thoughnotallcoveredpolicies.China’sSEAprovisions

arelimitedonlytotheassessmentofplansandtheRussianFederationassessmentlegislationdoesnot

currentlyaddressSEAs.Clearly,atleastalongtheChina‐CentralAsia‐WesternAsiaBRIEconomic

Corridor,thereissignificantvariationinnationalenvironmentallegislationincludingsomeseriousgaps

9Thereportfocusedonthefollowingcasestudycountries:Australia,Austria,Bhutan,Brazil,Cameroon,Canada,China,Colombia,Denmark,theEuropeanUnion,Fiji,France,Germany,India,Indonesia,Kenya,Lebanon,Mexico,Mongolia,Oman,Panama,Peru,SouthAfrica,Tanzania,theUnitedStates,andVanuatu(UNEnvironment2018a).10ThereportfocusedonKazakhstan,Kyrgyzstan,Tajikistan,Turkmenistan,Uzbekistan,Mongolia,China,andtheRussianFederation(Wingardetal.2014).

incoverage.Whilethepresenceofcomprehensivenationallegislationisnottheonlyfactorresponsible

forenvironmentalperformanceofacountry–EstyandPorter(2002)foundthatotherimportantfactors

includeregulatorystructuresresponsibleforcomplianceandenforcement;availabilityofenvironmental

information;civilsocietypresenceandempowerment;administrative,scientific,andtechnical

infrastructure–itisclearthatcross‐countrydifferencesinenvironmentalperformanceareassociated

withtherigorandstructureofenvironmentalregulationsaswellasacountry’swillingnessandabilityto

enforcetheseregulations(EstyandPorter2002).

Theenormousvariationincountries’environmentalpoliciesandenforcementcapabilitieshasreal

implicationsforBRItransportationinvestments.Oneoftherepeatedconcernsofenvironmentalcritics

isthatBRIinvestorswillshowapreferenceforinfrastructureprojectsandroutesincountrieswithlower

standardsandenforcementinordertoavoidtheupfrontcostsofcarryingoutacomprehensiveEIAor

SEA(W.Laurance2017;Moranetal.2018;Shinn2016).AsChinatightensitsownenvironmental

standards,thedifferencebetweenstandardsathomeandthoseinmanyBRIcountriesarebecoming

increasinglystark.WillChinafollowitslongstandingcommitmenttorelyonhost‐countrylegislationand

normstodictatehowBRIinvestmentsaddressenvironmentalimpacts?Orwillitrequireitsinvestment

toadheretomorestringentinternationalorChinesestandardsinordertoachieveitsvisionofa

sustainableandgreenBeltandRoad?Inthenextsectionwewillexploresomeoftheinternational“best

practice”standardsthatcouldhelpattainasustainableandgreenBRI.

II.GLOBALENVIRONMENTALPOLICIESTheGoodIndustryInternationalPractice(GIIP)standardsestablishedbyinternationalentities–

whetherinternationalagreements,internationallendingsafeguards,orvoluntarystandardsfromnonprofitorganizationsorindustry–areoftenmorerestrictivethanthoseofindividualnations,especiallycountrieswithweakenvironmentalpolicies.Inthissectionwereviewarangeofthesestandards,frombindingrequirementsfrominternationallendinginstitutionstovoluntarystandardsthatmaybeadvantageoustofollowwhenconsideringneteconomicbenefitsorreputationalenhancement.ThissectionalsoreviewstheenvironmentalstandardsofChinesepolicybanks.Critically,whileasjustnotedofteninternationalpracticesgobeyondthestandardsofhostcountries,ifthosepracticesarenotimplementedinwaysthatactuallyconstraintheimpactsofallBRIinvestments,includingwithsignificantconsequencesforfailingtocarryoutimprovedandagreedpractices,thentheirbenefitsarenotrealized.Consequentialmonitoringandtheenforcementofa*greedconditionsisanecessarypartofallprojects.

A. InternationalAgreementsThoughtherearemanyfewerinternationalenvironmentalpoliciesthanthemyriadnational

laws,therearestillfartoomanythatpotentiallybearonBRItransportationinfrastructuretodescribeindividuallyhere(andseeMonteiroandTrachtmaninreviewforprovisionswithintradeagreements).TheseinternationalagreementsarenotlikelytoaffectBRItransportationprojectsdirectlybutthroughtheirworkwithgovernments,internationalinstitutions,andNGOstocreateappropriaterelevantpolicies.Appendix11providesanoverviewofsomeofthecriticalagreementswiththegreatestoverlap.Formoredetailedinformation,theGreenGrowthKnowledgePlatformmeta‐analysisofenvironmentaldatabasesisagoodsource(Booth2017).TheEcolex(www.ecolex.org/),EnvironmentalTreatiesandResourceIndicators(http://sedac.ciesin.columbia.edu/entri/index.jsp),andInternationalEnvironmentalAgreementsDatabase(https://iea.uoregon.edu/)alsoprovideusefulreferencesalongthesesamelines.

B. LendingInstitutionEnvironmentalPolicies

AneffectivemeansofinfluencingBRItransportprojectshasbeenconditionssetbyfundingsources.

i.MultilateralDevelopmentBanks

WhilethemajorityofBRItransportationprojectloansarecomingfromChina,sometimesmultilateraldevelopmentbanks(MBDs)provideco‐financingfortheseprojects.AllMDBsnowhavetheirownsetofenvironmentalandsocialstandards,whichplacespecificrequirementsonloanrecipientsregardingtheminimumsocialandenvironmentalstandardsrequiredforinvestmentstobeallowedtoprogress,andBRIprojectswithMDBco‐financingadoptthestandardsoftheMDBpartner(Caderetal.inpreparation).EvenifthereisnofinancingfromMBDs,environmentalandsocialstandardsestablishedbyMDBs–and,inparticular,theIFC’sPerformanceStandardsortheWorldBank’snewEnvironmentalandSocialFramework(ESF)–areoftenconsideredtheGoodIndustryInternationalPractice(GIIP)orbestpracticestandardformultilateral,bilateral,orcommercialloans.Appendix12providesanoverviewoftheseenvironmentalandsocialsafeguardsforMDBsengagedinBRIinfrastructureloans.Still,clearly,theywouldhavetobestronglymonitoredandenforced,withconsequences,toprovidegoodincentives.

TheWorldBank,whichwasamongthefirstMDBstoaddressenvironmentalandsocialsafeguardsinthe1990s,adoptedrevisedenvironmentalandsocialsafeguardsin2018tostrengthentheirprotectionsfortheenvironmentandforpeoplewhilemakingiteasierforborrowerstocomplywithrelevantstandards.Thesereformswerearesponse,inpart,togrowingdemandforlendingoperationstobemoreefficient,forduediligencetobemoreflexible,andforgreaterrelianceonhostcountryenvironmentalandsocialstandards(Dollar2018),whichagainarelikelytobelessstrict.Thesestandardsparallel,tosomedegree,thefoundingvisionofthetworecentlyestablishedMDBs,theNewDevelopmentBank(NDB)ofBRICScountriesandtheAsianInfrastructureInvestmentBank(AIIB).BothNDBandAIIBclearlyhavepledgedtheircommitmenttosustainabledevelopmentaswellasefficiencygainsinprojectprocessing–whichmightconflict.AIIB,forexample,hascommittedtobe“lean,clean,andgreen,”meaningcosteffectiveandefficient,havezerotoleranceforcorruption,andarespectforsustainabilityandtheenvironment.NDB,ascomparedtotheothermultilateraldevelopmentbanks,dependsmoreheavilyuponthehostcountrypolicies,withtheaimofstrengtheningthosepolicieswhenfeasibleinsteadofimposingbankstandards(NewDevelopmentBank2016).Thispolicyof“countrysystemplus”allowsforhostcountryregulationstobeusedforenvironmentalandsocialsafeguardsexceptinprojectsorcountriesdeemedhighrisk,inwhichcaseNDBwoulddrawonIFCPerformanceStandards.SomecriticsworrythatbothAIIBandNDBeffortsfocussomuchonimprovingefficiencyandrelyingextensivelyonexistingcorporateorhost‐countrystandardsthatenvironmentalandsocialsafeguardsmaywellprovetobeinadequateinpractice.Criticsofthesepoliciesrecommendimprovingthedefinitionofsustainability,increasingclarityabouthowadverseimpactswillbemitigatedonceassessed,andincludingamoretransparentpolicyabouthowcountryandclientsystemswillbeassessedandincorporatedintotheplan(Weiss2017).

ii.ChinesePolicyBanksThetwolargestlendersforBRIprojectsarebothChinesepolicybanks:Export‐ImportBankof

China(ChinaExim)andChinaDevelopmentBank(CDB),whichhaveissued$100blnand$170blnrespectivelytowardsBRIprojects(GallagherandQi2018;Shinn2016).Chinesepolicybanksarestate‐ownedbanksthatareresponsibleforinfrastructurelendingaswellaspromotingforeigntrade.

TheenvironmentalandsocialsafeguardsofthesetwoChinesepolicybanksarestillnascentandlessrigorousthanthoseoftheMDBs(Appendices12and13;Dollar2018;FOE2016;Renetal.2017)).CDBwasoneofthefirstbanksinChinatohaveenvironmentalandsocialpolicies,though.By2005,e.g.,theCDBwasrequiringEIAstobeconductedinaccordancewiththelaw(Ren,Liu,andZhang2017).Specificenvironmentalregulationsarenotpubliclyavailable,althoughsomeofthebasicenvironmentalrulesissuedbytheCDBcanbeinferredfrompastlending,suchastherequirementtocompleteEIAsbyanindependentevaluatorandobtainapprovalthroughtheEIAprocess(FOE2016).TheChinesefinancialinstitutionsinvestinginoverseasinfrastructurearealsosubjecttoenvironmentalguidelinesputinplacethroughtheChinaBankingandRegulatoryCommission(CBRC,nowtheChinaBankingandInsuranceCommission—CBIRC).In2012,theCBRCissuedstrengthenedGreenCreditGuidelines(GCG)thatcompelChinesebanks’overseaslendingtofollowhostcountrylawsbutalsointernationalnorms,thoughthisisassessedattheinstitutionalnottheprojectlevelandimplementationisstillakeychallenge(FOE2017).Again,attheprojectlevel,monitoringandconsequentialenforcementarelikelytobeneededforimpact.

Overall,environmentalpoliciesofChinesepolicybanksarelessmaturethanthoseofotherinternationalfinancialinstitutionssuchastheWorldBankandIFC.Thepolicybankshavenotdevelopedcustomizedandspecificpoliciesforenvironmentalandsocialissueswhiletheadoptedenvironmentalpoliciesarenotstrictenough;significantly,theylackenvironmentaldepartmentstooverseeenvironmentalissues(Renetal.2017).Criticshaveurgedthepolicybankstostrengthentheentireprocess,includingthepre‐loanreview,thecomplaintmechanism,publicconsultancywithaffectedcommunities,andinformationdisclosuresystem(GEI2018).Whileallbanksrequireclientstofollowhostcountrystandards,IFCandotherMDBsgenerallyhavetheirownstandardsthat,whenstricterthanthoseofthehostcountry,mustalsobefollowedtohavetheirexpectedimpacts.InthecaseoftheChinaEximBank,onlywhennohostcountrystandardsexistwouldthebankconsiderapplyingChineseandinternationalstandards(Ren,Liu,andZhang2017).CDB'senvironmentalandsocialstandardsarealsomainlybasedonthehostcountrylawsandpolicies(Ren,Liu,andZhang2017).IFCandotherinternationalinstituteshaveworkedwithChineseofficialsrecentlytobuildcapacityinthesepolicybankstomanageenvironmentalandsocialrisks(Leungetal.2013).Thepolicybankshavemadeprogressinseveralareas,includingdevelopingrequirementsforenvironmentalandsocialriskcontrol,morestringentpunishmentsforviolatingenvironmentalregulations,11andmeasurestopromotegreeninvestment(Ren,Liu,andZhang2017).

C. ForeignDirectInvestmentPoliciesChinesestate‐ownedcompaniesaresofartheprimaryfirmsexecutingBRIprojects.Besides

environmentalregulationsimposedbytheirfunders,whatotherenvironmentalrestrictionsareChinesefirmssubjectto?Inadditiontohostcountrylawsandpolicies,Chinesefirms–especiallystate‐ownedfirmsbutalsoprivatecompanies–arepotentiallysubjecttoregulationsbytheirowngovernment.

Overthelastdecade‐‐andparticularlysince2015–domesticenvironmentalpolicieshavebeengreatlyexpandedandtheirenforcementstrengthenedwithinChina.InMarch2018thegovernmentannouncedinstitutionalreformthatprovidedexpandedresponsibilitiestotheMinistryofEcologyandEnvironment(MEE)andtheMinistryofNaturalResources(MNR).Appendix14,whichprovidesanabbreviatedlistofthemajorChinesenationalenvironmentalpolicies.Twoofthemostsignificantlaws,EnvironmentalProtectionLawofthePeople’sRepublicofChinaandLawofthePeople'sRepublicofChinaon

11Forexample,ifanenvironmentalimpactisdiscoveredandnotrectifiedduringmonitoring,theChinaEximBankwillstop

providingcreditandconsidertheprojectasanon‐performingloanandbaddebt.IftheCDBdiscoversaprojectisviolatingenvironmentalregulations,itdowngradestheproject’sassetlevelsandlimitsorstopsprovidingloans(FOE2016;Renetal.2017).

EnvironmentalImpactAssessment,requireenvironmentalimpactassessmentsandplanenvironmentalimpactassessment(otherwiseknownasstrategicenvironmentalassessments)forplanswhichaffecttheenvironmentwithinChina.TheEnvironmentalProtectionLawwasstrengthenedin2014andenforcementandmonitoringhasbeensubstantiallyincreasedsincethen(Leungetal.2013).However,theselawsdonotapplytoforeigninvestmentsoutsideChina.Followingtheexampleofmanydevelopedcountriesthathaveextendedtheirnationalenvironmentalstandardstoforeigninvestments,theNGOcommunityhasbeenpressingforChinesefirmstoberequiredtooperateunderthesameregulationsforforeignprojectsthattheymustfollowwithinChina.Ataminimum,someanalystsfeelhabituationofthesefirmstoregulationathomewillaffecttheirworkabroad.Othercriticssuggestthat,withoutstrictregulations,Chinesefirmsarelikelyto“outsource”theirpollutingindustriestoothercountriesasdomesticenvironmentallawstighten(Moranetal.2018).

ThoughgenerallyChinesedomesticenvironmentalregulationsdonotapplytooverseasinvestments,environmentalregulationsspecifictoChinesefirmsinvestingabroadhavebeenexpandedinrecentyears.ThekeypolicythatrelatestoChineseoverseasinfrastructureconstructionprojects,GuidelinesonEnvironmentalProtectionforOverseasInvestmentandCooperation,wasreleasedin2013bytheMinistryofCommerce(MOFCOM)andtheformerMinistryofEnvironmentalProtection(nowtheMinistryofEcologyandEnvironment—MEE).TheseguidelinesencouragebutdonotrequireChinesecompaniestocompleteenvironmentalimpactassessments(EIAs),developenvironmentalmitigationmeasures,andworkwithlocalcommunitiestoidentifypotentialnegativeenvironmentalandsocialimpactswhenoverseas(Leungetal.2013;Shinn2016).Whatwouldrequirethosestepsisnotclear.

Inadditiontoenvironmentalandsocialpoliciesaddressingforeigndirectinvestment,thereisagrowing

bodyofpoliciesthatarespecifictotheBeltandRoadInitiative.TheinitialvisionforaGreenBeltand

Roadwasreleasedin2015,followedin2017byaguidancedocumentandcooperationplan.Allthree

documentsareambitiousandpaintapictureoftheBRIbeingableto,“…promoteecologicalprogressin

conductinginvestmentandtrade,increasecooperationinconservingecoenvironment,protecting

biodiversity,andtacklingclimatechange,andjoinhandstomaketheSilkRoadanenvironmentfriendly

one.”(ChineseNationalDevelopmentandReformCommission,ChineseMinistryofForeignAffairs,and

ChineseMinistryofCommerce,2015)Allthreedocumentsarestillhigh‐levelandlackcriticaldetailson

implementation,monitoring,andenforcement(Chun2017).Thesepoliciesdodepict,however,howBRI

projectswithappropriateenvironmentalmitigationmechanismsandenforcementcouldserveasa

modelofgooddevelopmentpracticesthroughouttheregion.Ontheotherhand,itisquiteclearthat

positivestatementswithoutenforcedimplementationachievelittle.Again,thequestionisraised

concerninghowcompliancewithrelatedpromisesshouldbeenforced.

D. IndustryandNonprofitStandardsFirmsengagedinBRIprojectsmayvoluntarilyadheretobestpracticeenvironmentalstandards

–typicallystricterthanthoseprescribedbylendersorgovernments–fromindustryassociationsornonprofitorganizations.Firmsmightelecttofollowtheseguidelinesbecausetheycalculatethatitisintheirfinancialinteresttoinvestsustainablyorbecausetheyreceiveexternalpressure–fromtheircustomers,governments,orthegeneralpublic–andreputationalbenefitfromadoptingthesecriteria.

Oneusefulreferenceconcerningsuchvoluntarystandards,codesofconduct,andauditprotocolsrelatingtosustainabilityinsupplychainsistheStandardsMapdatabase(www.standardsmap.org),createdbytheInternationalTradeCentre(Booth2017).Again,theirimpactisanotherquestioninlightofthefactthatmanypositiveandwell‐intentionedstatementsaremadethatdonotshiftoutcomes.

Appendix15listsasampleofsuchvoluntarybestpracticestandardsthatareespeciallyrelevanttotransportationinfrastructureprojects.Thesecanrangefrombroadprinciplesthatfirmspledgetofollow–suchastheEquatorPrinciples–toextremelydetailedrequirementsthatmustbemetinordertoobtaincertificationofsustainability,suchastheCEEQUAL,Envision®,SuRe®,andGreenroads®standards.TheauthorsarenotawareofanyBRItransportationprojectthathasreceivedsustainableinfrastructurecertificationorhasutilizedtheNaturalCapitaltoolsintheirplanningprocess,stepsthatinprinciplecouldshiftoutcomes.ThetwoChinesefirmsthatarecurrentlymembersoftheEquatorPrincipleAssociation,BankofJiangsuandIndustrialBankCo.,Ltd.,arebothprivatecompanies.

WhiletheadoptionofbestpracticesisstilllackingforChineseoutboundinfrastructureinvestments,theChinaChamberofCommerceforMinerals,MetalsandChemicalsImportersandExportersdevelopedstandardsthatcouldserveasaprototype:GuidelinesforSocialResponsibilityinOutboundMiningInvestments.ThesecomprehensivebutvoluntaryguidelinesaimtodirectChineseminingcompaniestowardimprovingcorporatesocialresponsibility(CSR)andsustainabilitystrategies.Theoriginalversionwasproducedin2010;in2015arevisedandmoretransparentversionoftheguidelineswasreleasedandiscurrentlybeingrolledout(CCCMC2015;Chun2017).Again,impactsofadoptionareunclear.AnalystsgenerallybelievethatthelargeChinesestate‐ownedenterprises(SOEs),ascomparedtothesmallerprovincialandmunicipalSOEsandprivatefirms,aremorelikelytohaveboththecapacityandinclinationtofollowvoluntarybestpracticesandotherformsofCSR(MaurinandYeophantong2013;Xu2014;SunandTang2015;WangandHu2017;YuanandLandry2018).ThisisdueinparttothefactthatthelargerSOEshavetheeasiestaccesstoChinesestatefinancingandthemostoversightfromcentralauthorities.Theyareseenasrepresentingthepoliciesofthegovernment.Furthermore,largeSOEsarethetypeofChinesefirmmostlikelytoinvestinBRIinfrastructurethroughjointpartnershipsorpublic‐privatepartnerships(PPP)ratherthanengineer,procurementandconstruction(EPC)servicecontractsorbuild,operateandtransfer(BOT)contracts.Becauseofthelong‐termcommitment,thecompaniesinvestingjointpartnershipsandPPPsaremorelikelytoberesponsivetosomepressuresforlong‐termsustainabilityinvestments(Myxter‐Iinoetal.inprep).LargeSOEsarealsosensitivetotheirinternationalreputation,giventhattheydobusinessacrosstheglobe.BecausemostofthelargeBRIprojectsarecarriedoutbylargeSOEs,theiractions–positiveornegative–dodriveBRIinfrastructureinvesting.

III.RECOMMENDATIONS:BRICORRIDORENVIRONMENTALASSESSMENT

China’scommitmenttoasustainableBeltandRoadcertainlyissuggestedbythegrowingbodyofvision

statementsandguidelinesoutliningvariousconditionsforgreenBRIinvestments.However,todate,

theserecommendationsremainhigh‐levelandvoluntary,thoughpolicyinstrumentsandplanningtools

doexistthatcouldhelpachievethisvision.Whetherandhowtheyareimplementedwillbecritical.

WerecommendthatpoliciesbedevelopedfortheBeltandRoadInitiativethatrequireincorporating

environmentalassessmentproceduresintoearly‐stageplanningofentireBRIEconomicCorridors.ABRI

CorridorEnvironmentalAssessment(BRICEA)wouldfocusonentiretransportationcorridors,thereby

takingadvantageofthescaleandconnectivityofBRItoeffectivelyaddressthecumulativedirectand

indirectenvironmentalrisksfromtransportationinfrastructureprojectsandfollowingmanyofthe

principlesalreadydevelopedforSEAsandregionalenvironmentalassessments.WhileaformalBRI

CorridorEnvironmentalAssessmentprocessdoesnotcurrentlyexist,theUnitedNationsOfficefor

ProjectServices(UNOPS)hasdevelopedaframeworkforsustainableinfrastructureplanningand

developmentwhichcouldprovidemanyofthetoolstosupportsuchaprocess(UNOPS2017).Sucha

processcouldfollowmanyoftheprotocolsalreadyestablishedforChina’sdomesticPlanfor

EnvironmentalImpactAssessments,assetforthunderitsEIALawin2003.Further,sinceweknowthat

theperfectBRICEAswillnotalwaysoccur–inpartbecausesomeprojectsarealreadyunderway–we

stronglyrecommendthatasimilarapproachbeappliedtolatterstagesofrequiredreviewsofprojects

overtime,suchasduediligenceassessments.

ThefollowingattributesofaBRICEAmakeitdistinctlywellsuitedtoreduceenvironmentalrisksfrom

BRI’stransportationinfrastructurenetwork:

SpatialscaleManyenvironmentaleffectsassociatedwithBRItransportationprojects–suchas

impactsonwildlifemigration,carbonemissions,andforestcover–arewidespread,even

transboundary.Forexample,theimpactofarailorroadnetworkasabarriertowildlife

migrationcanonlybeunderstoodinthecontextoftheentiremigratorypathsofdisrupted

speciesandalltheexistingandproposedlinearinfrastructurewithinthespeciesrange.

Similarly,theassessmentofindirectimpactsofroadandrailnetworksondeforestationfrom

landusechangesisonlymeaningfulatthescaleofanentireBRIEconomicCorridor.When

planningthemostcriticalmitigationactivities–avoidanceactions–thescaleofaBRICEAis

indispensable:Plannerscanidentifyacceptableroutesthatavoidsensitiveareaswithina

corridorand,foreachofthealternativeroutes,prioritizetheirviabilitybytheirnetbenefitsthat

includeeconomic,environmental,andsocialinputs(Kieseckeretal.2010).Suchananalysis

couldalsoidentifyareasappropriateforrestorationandcompensatoryoffsets,whichshould

improvetheefficiencyandeffectivenessofmitigationforthecorridorasawhole.Each

individualBRItransportationsegmentwouldbeexpectedtocontributetoprescribedcorridor‐

widemitigationactions,eitherdirectlyorindirectly(throughcompensatoryoffsets).Forwide‐

rangingriskssuchasdeforestationandbiodiversityextinction,mitigationactionscanonlybe

efficientandeffectiveatthisregionalscale.

TemporalscaleBRIEconomicCorridorsstartwithaseriesofpre‐definedendpointsandmajor

hubs,buthavesomeflexibilityconcerninghowtoconnectthehubs.IntegratingSEAsintoearly‐

stageplanningforBRIEconomicCorridors–beforeindividualprojectshavebeenadvanced–

canprovidedirectiononhowtoplacetransportationinfrastructureswheretheyarelikelyto

causetheleastenvironmentalharmwhilemaintainingmostoftheeconomicandconnectivity

benefits.Itiscriticalthatthisprocessbeholisticallyintegratedintotheearly‐stageeconomic

planningforthecorridor,includingallaffectedsectors.Knowingearlyintheprocesswhich

sensitiveareasorinfrastructuredesignsareofflimitsaswellaswhichinvestmentshavepositive

netbenefitsshouldresultinproductiveinvestmentsoftime,money,andpoliticalcapitaland

reducerisksofdisruptionanddelay.

RegionalconnectivityABRICEAwillbemosteffectiveifitincludesinputfromallstakeholders

affectedwithinaneconomiccorridor.FollowingtheSEA‐typemodeldevelopedfortheMekong

BasinRegionalCommission,aregionalcommissioncouldbeestablishedforeachBRICSEAthat

includesgovernmentrepresentativesfromallaffectedcountriesandaprocessforpublic

participation(seeKeskinenandKummu2010).Thiscommissioncouldalsocollect,analyze,and

processinformationfromeachhostcountry’sregulatoryandlegalsystems,allowingall

stakeholderstocoordinatethecomplexgovernanceissuesofatransboundarytransportation

network.AhighlyvisibleandtransparentBRICEAprocessshouldpromotegoodrelations

betweenlendersandhostcountrygovernmentsandreducepotentiallocalcommunity

opposition(Leungetal.2013).

InvestorparticipationABRICEAprocesscouldhelplendersreducetime,money,andobstacles

dealingwithseriousenvironmentalissuesdownstreamduringtheEIAprocess.Whileindividual

EIAswouldstillneedtobecarriedoutforBRItransportationprojectstoaddresslocalimpacts,

thelikelihoodisthatthemajorobstaclesandneedforlate‐stageadjustmentswouldhave

alreadybeaddressedthroughtheBRICEAprocess.Withinputofexistingexpertisefrom

internationalfinancialinstitutionsandChineseagenciessuchasMEE,NDRC,andpolicybanks,

BRICEAsshouldallowlenderstorecognizethehighestinternationalenvironmentalstandards

whileatthesametimeminimizingthebureaucracy,time,andcostofdownstreamEIAs.Awell

conductedBRICEAshouldallowBRItransportationprojectstoachievethehighestGIIP

standardswhileatthesametimeminimizingthebureaucracy,time,andcostofdownstream

EIAs(AIIB2016).Inotherwords,SEAscouldprovidelenderssuchasAIIBtheopportunitytobe

“lean”and“green.”

SUMMARYANDCONCLUSIONEnvironmentalRisksfromBRITransportationInfrastructure

Wedistinguishdirecteffects–abiotic,ecosystems,wildlife–fromindirecteffectsinducedbychangesin

transportcostsandlanduse.Theformertendtobelocalizedandunambiguouslylinkedtoaroadorrail

project.Thelatteroftenhavemorecomplexconnectionsbut,still,potentiallymorepervasiveimpacts.

DirectEffectsDirectenvironmentalimpactsincludeabioticimpactslikeairandwaterpollution

aswellassoilerosion;ecosystemimpactssuchashabitatdestructionandfragmentation;and

wildlifeimpactssuchasvehiclecollisions,barrierstomigration,andillegalwildlifetrafficking.

BRItransportationinvestmentsmaybeespeciallypronetosuchrisksbecausemanyoftheBRI

EconomicCorridorspassthroughsteepterrainthatisvulnerabletoerosion,soildegradation

andsedimentation,andcontainsensitiveecosystemswithhighlevelsofspeciesendemism.

IndirectEffectsHistorically,roadandrailprojectscansetinmotionindirectandunintended

consequences.Changesintransportcostsshiftmarketsandhumanpopulations,whichcanopen

frontierstosettlementswithhabitatloss,deforestation,wildlifeandtimbertrafficking,among

othereffects.Themagnitudesofeffectsdifferbycontext.Deforestation–astraightforward

proxyforenvironmentalrisks–hasheterogenousresponsestotransportationinvestments

dependingontheirecologicalanddevelopmentsettings.BRItransportationprojectsthatare

sitedinhighlydevelopedregionswithextensivepriordeforestationsuchassouthwestChina,

Bangladesh,andpartsofCambodiaandKazakhstanmayhavelessofariseindeforestationand

itisevenpossiblethatincreasedeconomicdevelopmentspurredbyloweredtransportcosts

couldresultinreforestationorafforestation.BRItransportprojectslocatedinareasofmedium

development–thatis,areasatthemarginofdevelopmentfrontiersthatmaybeprofitablewith

furtherinvestments–canexpectsignificantforestlossfromtransportationprojectsnear

existingforestssuchasinMyanmar,LaoPDR,theMalaypeninsula,andwesternRussia.Finally,

inregionswithlowdevelopmentandlittlepriordeforestation,suchasineasternRussiaand

northwestThailand,theshort‐termdeforestationeffectsofBRIroadorrailprojectsmaybe

smallthoughthereremainsahighdegreeofrisk−anduncertainty−aboutlong‐termimpacts.

Roads–especiallyundividedhighwayswithunrestrictedaccesstosurroundingfrontiers–are

especiallyvulnerabletosuchpressures.Dividedhighwaysandrailwaysmayexperienceless

pressureduetotheirrestrictedaccess.

EnvironmentalMitigation

Themitigationhierarchydistinguishesfourtypesofactions–avoidrisks,reducerisks,restore

ecosystems,andoffsetdamages–todiminishimpactswhileallowingeconomicgainsfromtransport

investments.Theyareorderedintermsofdesirability.Inshort,adjustingprojectsupfronttoavoid

lossesiscrucial.

AvoidanceAvoidanceisbyfarthemostimportantcategory.Selectingroadorrailroutesthat

avoidenvironmentallysensitiveareasisusuallythemosteffectiveandleastexpensivewayto

minimizeenvironmentalharm.Earlyininfrastructureplanning,routescanbeselectedwith

minimalcosttoeconomicefficiency.Yetasplanningprogresses,mitigation–evenminorshifts

toroadorrailroutes–becomesincreasinglydifficult.Lateintheprocess,therearenomore

wholesaleavoidanceoptionsbutonlyremediationisavailable.Thatcanbecostlyandoften

ineffectual.

ReductionWhenfullavoidanceisnotpossible,mitigationshouldreduceenvironmentalharms.Engineeringanddesignsolutionsareamongthefrequentstrategiestoaddressdirectimpactsoftransportationinfrastructureonabioticconditions,ecosystems,andwildlife.Reducingindirectimpactsfromtransportprojects–suchasinducedlandusechanges–islessstraightforward.Complementarypoliciesarethepreferredstrategyformodifyinghumandecisionmaking.

RestorationRestorationisconductedatprojectsitestorepaireitherdirectorindirectimpactstobiodiversityorecosystemservices.Ifextensivelandscapetransformationhasoccurredduringconstruction,thenvegetative,wetland,orwaterwayrestorationisoftenrequired.

OffsetsCompensatoryoffsetsareafinalstrategy,aftertheotheractionshavebeenexhausted.Establishingprotectedareasorbuyingcarbonorwatercreditscaninprinciplecounteractthelocallossesofforests,biodiversity,carbon,orotherenvironmentalservicesduetotransportinfrastructure.Becauseitmaybedifficulttofullyanticipateindirecteffectssuchaschangesinlanduseoverthelongterm,andbecausecomparingdamagesatprojectsitestogainsfromoffsiteoffsetscanbedifficult,webelieveoffsetsareinherentlyriskyand,thus,thattakingthisapproachshouldincludeaimingforpositiveenvironmentalnetimpactsfromBRIinvestments.

EnvironmentalPolicies

Policieslargelydeterminewhichenvironmentalrisksareaddressedandhow.ForlargeBRItransport

investments,policiesofnationalgovernments,internationalagreements,lendinginstitutions,

corporations,andcivilsocietyorganizationsallcaninfluencethemitigationactionstaken.

NationalEnvironmentalPoliciesInprinciple,host‐countrypoliciesprovideatleasttheminimal

environmentalprotectionsforeverytransportationinfrastructureproject.Chinahaspledgedto

followhost‐countrypoliciesandnormsforallBeltandRoadprojects,forinstance.However,the

BRIcountriesdifferconsiderablyintheirenvironmentalpolicyframeworksand,crucially,intheir

capabilitiesandpoliticalwilltoenforceeventheirownsuchpolicies.Policiesmandatingtheuse

ofenvironmentalimpactassessments(EIA)areamongthemostprevalentrequirementofhost

countrygovernmentsconcerningenvironmentalprotectionfrominfrastructuredevelopment.

However,EIA’sabilitytoaddressenvironmentalrisksishamperedbecausetheyaretypically

conductedrelativelylateintheinfrastructureplanningprocess,whenmostimportantdecisions

havealreadybeenmade.Consequently,somegovernmentshaveadoptedearly‐stageplanning

suchasstrategicenvironmentalassessments(SEAs).SEAsdonewellcanaddresspoliciesand

plansataprogram,landscapeorsector‐level,beforeindividualprojectshavebeenadvanced.

GlobalEnvironmentalPoliciesBestpracticestandardsestablishedbyinternationalentities–

agreements,lendingsafeguards,voluntarystandardsfromnonprofitorganizationsorindustry–

areoftenmorerestrictivethanthoseofnations.AneffectivemeansofinfluencingBRIprojects

hasbeenconditionsestablishedbyfundingsources.IFCPerformanceStandardsortheWorld

BankEnvironmentalandSocialFrameworkareoftenconsideredGoodIndustryInternational

Practicestandardsformultilateral,bilateral,orcommercialloans.Theenvironmentalsafeguards

ofthetwoChinesepolicybanks–thelargestBRIfundingsources–arelesscomprehensiveand

rigorous,however,thanthoseofthemultilateraldevelopmentbanks(MDBs).YetwhenMDBs

haveco‐financedBRItransportprojects,theirmorestringentstandardsareapplied.Domestic

Chineseenvironmentalpolicies,regulations,standards,andinstitutionalreformshavebeen

greatlystrengthenedsince2015.SomeregulationsandstandardsspecifictoChinesepolicy

banksandfirmsinvestingabroadhavealsobeenexpandedinrecentyears.However,much

workstillneedstobedoneforChinatoexpanditsnationalenvironmentalstandardstoforeign

investments,especiallyintheareasofmonitoringandenforcement.

Recommendation:BRICorridorEnvironmentalAssessment

Givenitsscope,theBRI’spotentialenvironmentalimpactsareextensive.Whilemitigationoftheserisks

isdaunting,policyinstrumentsandplanningtoolsdoexisttodramaticallydecreasethem.Westrongly

recommendthattheBRIincorporateregionalversionsofSEA‐styleapproachesintotheearliest‐stage

planningforentireEconomicCorridors.ABRICorridorEnvironmentalAssessment(BRICEA)wouldfocus

upontheentiretransportationcorridor,takingadvantageofthescaleandconnectivityoftheBRIto

addressthecumulativedirectandindirectrisksfromtheseprojects.ManyoftheelementsofaBRICEA

alreadyexistwithinChina’sdomesticnationalenvironmentalpoliciesandcouldbeextendedto

encompassforeigninvestments.Further,sinceweknowthatperfectBRICEAswillnotalwaysoccur–in

partbecausesomeprojectsarealreadyunderway–westronglyrecommendthatasimilarapproachbe

appliedtolaterstagesofrequiredreviewsofprojectsovertime,suchasduediligenceassessments.

ThefollowingattributesmakeanEnvironmentalAssessmentfocusedonBRIEconomicCorridors

distinctlywellsuitedtoreducerisks:

SpatialscaleForwide‐ranging,cumulativeriskssuchasdeforestationandextinction,mitigation

actionsaremosteffectiveataregionalscale.Foravoidance,thescaleofaCorridor

EnvironmentalAssessmentisindispensable,asplannerscanidentifyroutestoavoidsensitive

areasand,acrossalternativeroutes,prioritizebynetbenefits.Onecanalsoidentifyareas

appropriateforrestorationandoffsets.Eachsegmentwouldbeexpectedtocontributeto

compliancedirectlyorindirectly(viacompensatoryoffsets).

TemporalscaleIntegratingenvironmentalassessmentsintotheveryearliestplanningstages,

beforeindividualprojectsadvance,bestpermitsguidanceonhowtoplacetransport

infrastructurewhereitcausesleastenvironmentalharmwhilemaintainingmosteconomic‐

connectivitybenefits.Itiscriticalthatearly‐stagecorridorplanningintegrateallaffectedsectors.

Knowingearlywhichsensitiveareasshouldbeoff‐limits,andwhichinvestmentshavegreatest

netbenefit,shouldyieldproductiveinvestmentsoftime,money,andpoliticalcapitalandreduce

risksofdisruptionanddelay.

RegionalconnectivityABRICorridorEnvironmentalAssessment(BRICEA)willbemosteffective

withinputfromallstakeholdersinaneconomiccorridor.Aregionalcommissioncouldbe

establishedforeachBRICEAwithgovernmentrepresentativesfromallaffectedcountries,plus

broadpublicparticipation.AhighlyvisibleandtransparentBRICEAprocessshouldpromote

goodrelationsbetweenlendersandhostcountrygovernmentsandreducepotentialopposition

fromlocalcommunities.

InvestorparticipationAwellconductedBRICEAshouldallowlenderstoinsistuponthehighest

internationalenvironmentalstandardswhileatthesametimeminimizingthebureaucracy,

time,andcostofdownstreamEIAs.WhileindividualEIAswouldstillneedtobecarriedoutto

addressspecifictechniquechoiceswithlocalimpacts,thelikelihoodisthatthemajorobstacles

andneedforlate‐stageadjustmentswouldalreadybeaddressedthroughsuchanSEAprocess.

APPENDICES

APPENDIX1:ForestCoverChangeacrosstheEntireBeltandRoadChangeinforestcover12canbe

usedasaproxyforenvironmentalrisks

relatedtoawiderangeof

environmentalvariablesofinterest,

includingspecies’habitats,healthy

ecosystems,carbonstorage,water

provision,andotherecosystemservices.

Becausesatellitedataonforestcover

arerelativelyeasytoobtain,thisisone

ofthesimplestandmost

straightforwardtoolstohelpidentify

theimpactoftransportation

infrastructureonecosystems.13

Whenexaminingforestcoverchangeat

thescaleoftheentireBeltandRoad,

somedistinctpatternsemergethat

invitefurtherexploration.Thevariationsinforestcoverandloss,forinstance,highlighttwoimportant

questions:WithinwhichoftheBRIEconomicCorridors(Figure5)areforestsmostvulnerableto

infrastructuredevelopment?Andisthischangeinforestcoveragoodproxyforvulnerabilitytoother

environmentalrisks?

Acursoryevaluationindicatesthatsomecorridorsgeneratemoreriskthanothersbasedonthesimple

factthatnotallaffectedareashavemuchforest.TheChina‐IndochinaPeninsulaEconomicCorridor

(CICPEC)andChina‐Mongolia‐RussiaEconomicCorridor(CMREC)areamongareasatgreatestriskwith

largeareasfacingactivedeforestationinthelast15years(Figure6).Further,acrossandwithinpotential

economiccorridors,wecantakeintoaccountlevelsofpriordevelopment−e.g.,highversusmediumor

low–whenconsideringimpactsoftransportationinvestment.(SeePart1,SectionII.B.i.bfora

theoreticalframeworkconcerningdeforestationinrelationtoeconomicdevelopment.)Withinhighly

developed(anddeforested)regionssuchassouthwestChina,Bangladesh,andpartsofCambodiaand

Kazakhstan,littlenaturalforestremainstobeclearedforagriculture.Hence,littleforestimpactis

expectedfromBRItransportationprojectsinthesehighlydevelopedregions.Itisevenpossiblethat

increasedeconomicdevelopmentspurredbyloweredtransportcostsinsuchregionsleadstorural

12“Forest”forcontinentalSoutheastAsiaandsouthwestChinareferstoarangeofdifferenttropicalforesttypes,includingsomemoistrainforestsbutmostlyevergreenforestsoccupyingthemountainouszones;seasonalordeciduousforestsinthemonsoondominatedsub‐region;verydryforestsandwoodlandsoccupyingplains,plateaus,andotherwater‐limitedsites;andmangroveforestsalongthecoast(H.Stibigetal.2007).13Whilevegetationcoverisrelativelyeasytodistinguishfrombarrenareasinsatelliteimagery,primaryforestscanbedifficulttodistinguishfromtreeplantations,degradedforests,andregeneratingforests,thuscomplicatingtheabilityofsatelliteimagerytoprovidegoodestimatesofchangestohabitats,carbon,andotherecosystemservices(Harrisetal.2012).

Figure5.SixeconomiccorridorsoftheBeltandRoadInitiative

Source:china‐trade‐research.hktdc.com/business‐news/article/The‐Belt‐and‐Road‐

Initiative/The‐Belt‐and‐Road‐Initiative/obor/en/1/1X000000/1X0A36B7.htm

transformationsthatresultinreforestationorafforestationviatreeplantationsornaturalregrowth(see

Kazcan2016).

Ontheotherhand,

withinareasof

mediumdevelopment

–thatis,areasatthe

marginof

developmentfrontiers

thatarecloseto

profitablenowandwill

beprofitablewith

furtherinvestments–

significantforestloss

canbeexpectedfrom

BRItransportation

projectsnearexisting

forests.Suchareascan

befoundwithin

CICPEC(inMyanmar,

theLaoPeople’s

DemocraticRepublic,

theMalaypeninsula)

andCMREC(near

Moscow).

Finally,inregionswithlowdevelopmentandpriordeforestation,suchasineasternRussiaFederation

andnorthwestThailand,theshort‐termeffectsofBRIroadorrailprojectsmaybesmallthoughthere

remainsahighdegreeofrisk−anduncertainty−aboutlong‐termimpacts.ShouldBRIroadandrail

projectstriggerdevelopmentdynamics,withcomplementaryinvestmentsacrossadecadeormore

maketheseareasattractiveforincreasedsettlement,thensuchareasmayexperienceconsiderable

long‐termforestloss.However,ifotherfactors(e.g.,topography,climate,andpolicies)dissuadesuch

dynamics,therecouldbelittlechange.

Otherenvironmentalfeatures–suchaswatershedprotectionwhichishighlyvalued,e.g.,withinbig

cities,orbiodiversityhotspotsthatarecriticalhavensforclustersofuniqueandthreatenedspecies–

arenotnecessarilycloselycorrelatedoverspacewithforestcoverdynamics.AsseeninFigure7,

identifiedbiodiversityhotspotsmayprovideadditionalguidanceatacorridorscale.Atsuchabroad

level,theCMRECcorridor,e.g.,facesfewerspecifichotspotrisksfromBRItransportationprojectsthan

doestheCICPECcorridor.Meanwhile,plannedBRIroadandrailprojectsoftheChina‐Central‐and‐West‐

AsiaCorridorcutthroughbiodiversityhotspotswithinallofKyrzgystan,Tajikistan,andUzbekistan.

EvensuchcoarsepatternsclearlyillustrategreatheterogeneitiesforenvironmentalrisksfacingtheBRI,

duetovariationsinbotheconomicandecologicalfactorsaffectingtheindividualandcumulativerisks

Figure6.Railandroadprojects(existing,underconstruction,andplanned)across

entireBeltandRoadregioninrelationtoforestcover,forestloss,andforestgain

fromtransportationprojectswithindifferentBRIEconomic

Corridors.Consequently,itisimperativethatintegrated

planningprocessesatearlystagesofcorridordevelopment

shouldnotconsidertheaverageorthemodalBRIproject

but,instead,allthecorridordetailsrelevantforassessingthe

cost‐effectivemeansforreducingenvironmentalriskwithin

eachcorridor.Thiscancontributegreatly,atlowcost,to

avoidingsocietallyinefficientinvestmentsinwhich,e.g.,

evenminoreconomicgains(ornogainsornetfinancial

lossesfrominvestments)couldgeneratesignificant

environmentalrisks,i.e.,a‘lose‐lose’.Suchassessmentsare

eminentlypossible,forinstancealongthelinesillustratedfor

roadinvestmentsacrossvariedlandscapesinDasguptaand

Wheeler(2016)andDanyo,DasguptaandWheeler(2018).

TowardsapplicationofsuchthinkingwithintheBRI,inthe

succeedingappendiceswewillexploresomeofthese

potentialrisksandapproachestotheirpotentialmitigationin

greaterdetailforonecorridor,CICPEC.

Figure7.Railandroadprojects(existing,under

construction,andplanned)acrossentireBelt

andRoadregioninrelationtoConservation

International’sBiodiversityHotspots

APPENDIX2:AssessingEnvironmentalRiskfromTransportationInfrastructure

DevelopmentintheChina‐IndochinaPeninsulaEconomicCorridor(CICPEC)Manyindividualtransportationprojectsorsegmentsare

requiredtoconnectanentireBRIEconomicCorridor.While

eachofthesesegmentsofrailorroadmaycrossarelatively

shortdistance,asawholethefulleconomiccorridorcan

traversemanydifferentecosystemsandpoliticalboundaries.

Atwhatscaleshouldenvironmentalrisksbeassessed–at

theleveloftheentirecorridorortheindividualprojects?

Thereisnotoneanswertothisquestion,asitwilldependon

thetypesofrisksbeingevaluated;invariablyassessments

needtobemadeatmultiplescalestotrulycaptureallrisks.

Carryingoutanassessmentattheprojectlevelthroughan

environmentalimpactassessmentmakessensewhen

consideringtheimpactsofconstructiononlocalizedharms

suchasparticulatepollutionorchangesinsoilstabilization.

However,theeffectofinfrastructureoperationsonmanyof

themostpervasiveandlong‐lastingfactors–suchashabitat

fragmentationandlandusechange–canonlybeproperly

evaluatedatthebroaderlandscapeorentire‐corridorscale,

ideallythroughanearly‐stagestrategicenvironmental

assessmentprocess.Overthecourseofaseriesof

AppendiceswetakeacloserlookatoneoftheBRIEconomic

Corridors,theChina‐IndochinaPeninsulaEconomicCorridor

(CICPEC),toillustrateafewofthepotentialenvironmental

risksfromalternativerailandroadroutesatthescaleoftheentirecorridor.Specifically,wewillanalyze

recentlyconstructedandplannedBRIhigh‐speedrailandhighwayprojectsconnectingKunming,China

toSingapore,viaBangkok,Thailand.

ThevisionofimprovingthetransportationnetworkbetweenKunmingandSingaporeisnotnewtoBRI

andCICPEC.EffortstoconnectsouthwestChinaandSoutheastAsiathrougharailnetworkgobackto

theearly1900sduringthecolonialperiod.Threepotentialrouteshavebeenenvisioned:Awestern

routepassingthroughMyanmar;acentralrouterunningthroughtheLaoPeople’sDemocraticRepublic

andThailand;andaneasternroutethatcrossesthroughVietnamandCambodia(Figure8).Allthree

routesarecurrentlypassableatleastbyroadforpassengersbutnoneofthethreerouteshasa

completeahigh‐speedraillinecapableofefficientfreighttransportation.

Inthefollowingappendices,wefocusprimarilyonriskstoforestsandbiodiversityconservation.Thisis

onegroupofrisksamongmany,includingenvironmentalandsocialrisksrelatedtowaterprovisionand

quality,airandnoisepollution,GHGemissions,firevulnerability,earthquakevulnerability,resettlement,

andindigenousrights.Wefocusonbiodiversityandconservationasdistinctiveandvulnerablenatural

assetsofCICPEC.IthasbyfarthehighestlevelsofbiodiversityofanyoftheBRIEconomicCorridors.In

addition,habitatandbiodiversityrisksaresometimesquitespatiallyspecific,offeringtheopportunityto

avoidoratleastgreatlyreducedamagesthroughchoicesacrossroutesoradjustmentsinchosenroutes.

Figure8.ExistingandBRIroadandrail

segmentsalongtheeastern,central,and

westerntransportationroutesconnecting

Kunming,ChinaandBangkok(highlightedin

yellow).

Toassesstheconservationrisksfromrailandroadimprovementsornewconstructionalongthesethree

alternativeroutes,wewillexplorefivepotentialindicatorsacrosstheentireCICPECcorridor:forest

cover,intactfrontierlandscapes,biodiversityhotspots,umbrellaspecies,andillegalwildlifetrafficking.

APPENDIX3:EnvironmentalIndicator:EcosystemEffectsasMeasuredbyForestCover

ChangeChangeinforestcover14canbeusedas

anindicatorofenvironmentalriskthat

relatestoawiderangeofenvironmental

variablesofinterest,includingspecies’

habitats,healthyecosystems,carbon

storage,waterprovision,andother

ecosystemservices.Becausesatellitedata

onforestcoverarerelativelyeasyto

obtain,thisisoneofthesimplestand

moststraightforwardtoolstohelp

identifytheimpactoftransportation

infrastructureonecosystems.15

Inrecentyears,mostofthechangein

forestcoverwithintheCICPECcorridor

hasoccurredinCambodia,theLao

People’sDemocraticRepublic,and

Myanmar,especiallyinmountainous

areas(H.Stibigetal.2007;H.J.Stibiget

al.2014;Zengetal.2018;Figure9).

Interestingly,manyofthemostaffected

areashavebeenalongtheboundaries

betweencountries,suchasbetween

YunnanProvince,ChinaandMyanmar;

LaoPDR,CambodiaandVietnam;and

betweenCambodiaandThailand.Most

afforestationandreforestationhas

occurredinVietnamandnorthern

Thailand.Theprimarydriverofchangeinforestcoveristheconversiontocashcropplantations,

followedbyconversiontotimberplantationsandlogging(H.Stibigetal.2007;H.J.Stibigetal.2014).

Whatrolehaveroadsplayedinforestconversion?Stibigandcolleagues(2007)foundthatroad

constructionwasresponsibleforsomeforestchange,particularlyintheregionfromChinatoThailand

throughthenorthwestofLaoPDR;fromeasternThailandthroughLaoPDRtothecoastofcentral

Vietnam;andinCambodia.Theseauthorsdidnotidentifyanyforestchangethatresultedfromrail

14“Forest”forcontinentalSoutheastAsiaandsouthwestChinareferstoarangeofdifferenttropicalforesttypes,includingsomemoistrainforestsbutmostlyevergreenforestsoccupyingthemountainouszones;seasonalordeciduousforestsinthemonsoondominatedsub‐region;verydryforestsandwoodlandsoccupyingplains,plateaus,andotherwater‐limitedsites;andmangroveforestsalongthecoast(H.Stibigetal.2007).15Whilevegetationcoverisrelativelyeasytodistinguishfrombarrenareasinsatelliteimagery,primaryforestscanbedifficulttodistinguishfromtreeplantations,degradedforests,andregeneratingforests,thuscomplicatingtheabilityofsatelliteimagerytoprovidegoodestimatesofchangestohabitats,carbon,andotherecosystemservices(Harrisetal.2012).

Figure9.RailandroadprojectswithinCICPECandforestin

relationtoforestcover,forestloss,andforestgain

construction(H.Stibigetal.2007).Thismaybebecausetherehavebeenmanyfewerrailwaysthan

roadsconstructedinthelasttwodecadesincontinentalSoutheastAsia;becauseroadsweremorelikely

tobebuiltintopreviouslyinaccessiblefrontierlandscapes;orbecauserailwaysprovidelessaccessto

openforestedareasbecausepassengersandfreightarelimitedtostationaccess.

Lookingforward,wouldforest

coverbeequallyaffectedby

BRItransportationprojects

alongtheeastern,central,and

westernroutes?Manyofthe

recentandproposedBRI

projectsalongthesethree

corridorsaresitedinareasof

highpriordeforestationoron

topoffootprintsofexisting

smallerroadsorrail(Figure9),

whichlessensthedegreeto

whichtheycouldfurther

fragmentthelandscape,

degradeexistingforests,or

providenewaccess.Allthree

oftheseroutes,though,pass

throughsomeforested

habitatsthatmaybe

vulnerabletohabitat

destructionasaresultofanew

orupgradedrailorroad.As

canbeseeninFigure9,BRIrail

linescouldcreateextensive

riskofdeforestationin

northernMyanmar(Western

route)andLaoPDR(Central

route).Onesegmentthatmeritsparticularflaggingforcarefulandearlyinvestigationistheproposed

BRIBurmarailNamTok‐ThanbyuzayatprojectintheThailandsectionofthewesternroutecrossesone

ofthemostforestedandbiologicallyimportantareasofThailand(Figure10inset),cuttingthrougha

nationalparkandskirtingalongsideseveralotherprotectedareas.Bycontrast,theThailandsegmentof

thecentralroutepassesthroughareaswithextensivepriordeforestation(Figure9)andposesnothreat

offurtherforestcoverloss.

Figure10.RailandroadprojectswithinCICPECinrelationtoforestcover,

forestloss,forestgain,andprotectedareas.InsetofproposedBRIBurma

railNamTok‐Thanbyuzayatproject

APPENDIX4:EnvironmentalIndicator:EcosystemEffectsasMeasuredbyIntactFrontier

LandscapesinCICPECIntactforestlandscapes(IFLs)canbe

amongthemostcriticalareasto

conservebecausetheyfrequently

serveasrefugesfornative

biodiversityandproduceextensive

ecosystemservices.Becauseof

extensiveloggingandlanduse

changeovermanydecades

throughoutcontinentalSoutheast

Asia,fewIFLsremainwithinCICPEC

(Figure11).Itisnotsurprisingthat

BRIrailorroadlinesmostlyavoid

thesescarceIFLsbecausethe

eastern,central,andwesternroutes

arelargelybuiltalongthefootprint

ofexisting,thoughlessdeveloped,

transportationcorridors.16Because

thetropicalregionofSoutheastAsia

isamongthemostdiverseand

productiveregionsoftheworldfor

biodiversityandcarbonproductivity

(Lietal.2016),itisespeciallycritical

tosafeguardtheintegrityofthefew

remainingIFLsintheregion.Ofthe

BRIrailandroadprojectsalongthe

CICPECcorridor,twoareofconcern:

Theupgradingandwideningofa

highwayprojectandpotentialHSR

projectscontinuetoexpandnearthe

UNESCOWorldHeritagesiteDong

Phayayen‐KhaoYaiForestComplex

(arareIFLnearBangkok)andtheproposedBurmarailNamTok‐Thanbyuzayatprojectthatbisects

severalThainationalparksneartheMyanmarborder.

16Bydefinition,anIFLisroadless.

Figure11.RailandroadprojectswithinCICPECinrelationtointact

frontierlandscapeandprotectedareas.

APPENDIX5:EnvironmentalIndicator:WildlifeEffectsasMeasuredbyBiodiversity

HotspotsandUmbrellaSpeciesWhileforestcoverchangecanbeagood

indicatorofwildlifehabitat,animalsarenot

equallydistributedacrosstheseareas.

Other,complementaryindicatorsare

neededtoidentifyhigh‐valuebiodiversity

habitats.IFLscapturehigh‐valueareasthat

serveascriticalrefugesformanyanimal

speciesthatrequireextensiveundisturbed

habitat,butthereareotherspeciesthat

mightbethreatenedpreciselybecause

theirhabitathasbeenreducedinsizeand

arenotfoundinIFLs(manyofwhichare

mountainous).Hereweintroducetwo

alternativeindicatorsthatcanhelpidentify

someofthemosthigh‐valuebiodiversity

areasthatshouldbeprioritizedwhen

mitigatingriskscreatedbytransportation

corridors.Thefirstindicator,“biodiversity

endemismhotspots,”delineatesareaswith

highlevelsoffaunalspeciesrichness,

speciesendemism,andspecies

endangerment.Liandcolleagues(2016)

developedsuchametricforcontinental

SoutheastAsiautilizingdatafromtheIUCN

RedListofSpeciesofConcernfor

mammals,birds,andamphibians,

augmentedwithexisting,publiclyavailableremotesensingdata(Figure12).Overlappingdistributionsof

speciesofconcern,Liandcolleaguesprioritizedareasforconservation.FortheCICPECcorridor,the

highestvalue,mostsensitiveareasforbirds,mammals,andamphibiansmostlyoccurinmountainous

areasofSoutheastAsiaduetotheirhighendemismandspeciesrichnessbutcanalsobefoundinlower

elevations,especiallyinnorthernregionsoftheLaoPeople’sDemocraticRepublic,Vietnam,and

Thailand.TheplannedBRIrailprojectsalongthecentralrouteinLaoPDRclearlywouldcutdirectly

throughthisbiodiversityendemismhotspotwhiletheothertworouteslargelyskirtthesehotspots.

Thedistributionofumbrellaspeciescanserveasanalternativeindicatorforidentifyinghigh‐valueareas

forwildlife.“Umbrellaspecies”arewildlifespeciesthathavelargehabitatneedsorotherrequirements

andwhoseconservationresultsinmanyotherspeciesbeingconservedattheecosystemorlandscape

level.FortheCICPECcorridor,largecatssuchastigers(Pantheratigris)andcloudedleopards(Neofelis

nebulosi)canservethisrolenicely(Figure13).Formanydecades,tigershavebeenthreateneddueto

theexpansionofhumandevelopmentincludinganextensiveroadsystemthroughoutmostofthetigers’

rangethatcreatedaccessforpoachersandfragmentedtigerhabitat(Mcmillan2018;WWF2016).Due

tooverhuntingandlossofhabitat,fewerthan3,500tigerslivedinthewild,occupyinglessthan7%of

Figure12.BRIroadsandrailprojectsinCICPECasthey

intersectwithhotspots9ofbiodiversityendemismfor

threatenedspecies(afterLietal.2017)

theirhistoricalrangedespite

extensiveeffortstosetup

protectedareastoconservethese

largecats.Tigersarelistedas

endangeredbytheIUCNRedList

(Mcmillan2018;Walstonetal.

2010).Thesmallercloudedleopard

isalsothreatenedwithextinction,

butlesssothantigers.Theclouded

leopardsarelistedasvulnerable

becausethespecieshaslostan

estimated30%ofitsadult

populationinthelasttwodecades

duetopoaching,habitatloss,

fragmentation,andotherhuman

pressures(Mcmillan2018).Theonly

placeswhereBRIprojectsintersect

withthetigerhabitatsarethe

BurmarailNamTok‐Thanbyuzayat

neartheThailand‐Myanmarborder

andtheBangkok‐KualaLampurHSR

projectinMalaysia.Theformer

projectbisectsthetigerrange,

potentiallyseparatingpopulations

whilethelatterskirtsthesideofthe

tigerrange,thuspotentiallyfurther

restrictingtigerhabitat.Clouded

leopardshavealargerrangeand

morepopulationsthataremore

resilienttohumandevelopment;theirhabitatsareatriskfromBRItransportationprojectsalongall

threecorridors.OfgreatestconcernaretheKunming‐Vientianerailwaythatwouldbisectextensive

cloudedleopardhabitatinLaosandtheKunming‐Mandalaycorridorbisectingthecat’srangein

Myanmar.

Whichofthesetwoindicators–biodiversityendemismhotspotsorumbrellaspecies–isabettertool

foridentifyingsensitiveandhigh‐valueareaswhenassessingtransportationinfrastructurerisks?The

umbrellaspeciesrangesarerelativelysimpletousebecausetheyrequiretherangeofonlysingleor

severalspecies.Theyhavetheaddedvalueofgeneratingpublicawarenessandpotentiallygreater

publicsupportbecauseofthehigh‐profilenatureofthesespecies.However,individualspecieswill

invariablyhaverangesthatarespecifictotheirhabitatrequirementsandthreats.Theabsenceoftigers

fromLaoPDR,ahubofbiodiversityendemism,showsthelimitationofusingonlyumbrellaspecies.Asa

compositeindicator,biodiversityhotspotsaremorelikelytocaptureareasofhighvaluetoawider

communityofspeciesatrisk.Ideally,bothindicatorswouldbeused,utilizingtheircomparative

advantages.

Figure13.TigerandcloudedleopardrangesandBRIroadandrail

infrastructureinCICPEC.Insetshowpotentialhabitatcorridorthat

couldconnecttwodisjuncthabitatranges.

APPENDIX6:IllegalloggingandwildlifetradeImprovingtransportation

corridorscancontribute

significantlytoIllegallogging

andwildlifetrade,especially

alongtheCICPECcorridor.The

SoutheastAsiaregionisbotha

sourceforillegalwildlifeand

timberandahubfor

internationaltrafficking

(Broussard2017;Chouvy2013;

Felbab‐Brown2013a).The

threatposedbytransportation

infrastructuredevelopmentis

twofold:First,newand

improvedroadsinSoutheast

Asiahavebeenopeningupnew

frontiersforillegalloggingand

wildlifepoaching(Clementset

al.2014;Felbab‐Brown2013b,

2013a).Additionally,these

improvedtransportation

networksfacilitatethe

movementoftimberand

wildlife(BenYishayetal.2016;

Felbab‐Brown2013b,2013a).

Giventhatmuchoftheillegal

timberandwildlifetradeare

destinedforEastAsia,

facilitatingthemovementof

goodsalongBRItransportation

corridorsinCICPECwill

undoubtedlyalsofacilitate

wildlifesmuggling.Considerthe

particularcaseofthepangolin,

aspecieshighlyendangeredprimarilybecauseofextensivepoachingfortheinternationalillegalwildlife

markets.AscanbeseeninFigure14,theproposedBRISihanoukvillePortRoadandRailextensionsin

Cambodiaarelocatedveryclosetooneoftheprimesourceareasforillegalpangolinpoaching,the

nearbyBokorNationalParkandsurroundingforests(Chouvy2013;Felbab‐Brown2013a).Anupgraded

portlinkwouldnotonlyfacilitatesmugglingaccesstothecurrentpangolintraffickingroutes,itcould

potentiallyopenanewmaritimehubforsmugglingtraffic.

Figure14.PangolintraffickingroutesandBRIinfrastructureinCICPEC.

InsetshowsproximityofproposedBRISihanoukvillePortConnector

RoadtotheBokorNationalPark,asourceofpangolins(afterChouvy

2013).

APPENDIX7:EnvironmentalIndicator:IndirectEffectsasMeasuredbyForestCover

ChangeSouthworthetal.(2011)findevidenceforascarcity‐basedregionalforesttransitionincontinental

SoutheastAsiabycomparingCambodia–thecountrywiththehighestproportionofintactfrontier

landscapeintheregionandalsothefastestrateofdeforestation–withThailandaswellasNepaland

India–wheredeforestationappearstohaveslowedasforestbecomesscarcer.Cropperetal.(2001)

alsofindsuggestiveevidencethatforesttransitioneffectsmaybepresentinThailandwheretheyfind

thattheroadsbuiltfartherawayfromintactforest

fringeshavelittleeffectonforestclearing(consistent

withAmazonianevidence).AscanbeseeninFigure

15,reforestationinnorthernThailandandthe

contrastingdeforestationacrosstheborderintheLao

People’sDemocraticRepublicprovidesfurther

supportthattheforesttransitionisunderwayin

ThailandbutLaoPDRandCambodiaarestillinthe

earlystagesofdeforestation.MeyfroidtandLambin

(2009)giveanaccountofforesttransitioninVietnam,

whichischaracterizedbyapproximately40%leakage

tonearbyLaoPDRandCambodiawhereillegalforest

harvestingismorecommon.Thisregionalexperience

ofweakerprotectionwithinsomecountries,

particularlythosewithweakerstates,suchas

Cambodia,underscorestheroleoftradeindisplacing

deforestation.Itcouldbeexpectedthatimproved

transportationwouldincreasethesalienceofthis

mechanismastheregionbecomesmoreconnected.

WhatdoesthisimplyfortheproposedBRI

transportationprojects?Threeplannedprojectshave

beenhighlightedinearlierappendicesfortheirhigh

conservationvalue:TheNamTok‐Thanbyuzayatrail

alongthewesternrouteinThailandforitsextensive

forestcover,IFLs,andbigcathabitat;theLaoPDR

section(Boten‐Vientian)ofthecentralrouteforits

endemismbiodiversityhotspotandextensiveforest

cover;andtheCambodianSihanoukvillePortroadand

railextensioninthewesternrouteforitsIFLandrole

insupportingendangeredpangolinhabitat.Whileall

ofthesedeservecarefulattentionandcarefully

plannedmitigationstrategies,theincreasedpressure

forlandusechangeanddeforestationcreatedbyBRItransportationnetworkimprovements,increased

trade,andurbanizationwouldlikelyresultinmoredeforestationintheCambodianandLaoPDRsites.

TheThaiandCambodiansitesbothhaveexistingprotectedareasinplace,butmuchgreater

enforcementisneededinBokorNationalParktoeffectivelysafeguardthisarea.

Figure15.RailandroadprojectswithinCICPECin

relationtocropland,forestcover,forestloss,and

forestgain

APPENDIX8:MitigationStrategies:ProtectedAreasProtectedareascanprovideanimportant

toolformitigatingenvironmentalrisksfrom

BRIrailandroadprojects.Establishmentor

increasedenforcementofaprotectedarea

nearatransportationcorridorcanhelp

reducethepotentialimpactfromlanduse

changefromhumansettlement,illegaltimber

extraction,andwildlifepoaching.Figure16

showsthat,whiletherearemanyprotected

areasalongexistingroadsandrailinCICPEC

aswellasthoseunderconstructionor

planned,theyarestrikinglyabsentfromthe

LaoPeople’sDemocraticRepublic.17This

clearlyflagstheneedtoconsiderthisaspart

ofamitigationstrategyfortheplannedrail

linethatwillpassthroughareasofhigh

endemismforthreatenedbiodiversity.

Establishingaprotectedareadoesnot

provideautomaticsafeguardagainsthuman

settlementanddestructiveactivitiessuchas

wildlifepoachingandillegallogging.Aroador

railbisectingaprotectedareawillcreatean

edgeeffectthatcanpenetratedeepintothe

zone.Moreover,theprotectedstatusisonly

asstrongasitsenforcement.Considerthe

caseofSnuolWildlifeReserveinCambodia.

Landsatimageryrevealsthatinlessthana

decadeafteraroadwasconstructedpassing

directlythroughthepark,extensiveclearing

occurredwithintheparkalongtheroadinpart

duetopoorenforcement(Clementsetal.2014;

Figure17).

TwooftheplannedBRIprojects–asegmentof

theThaiNamTok‐Thanbyuzayatrail(Figure10

inAppendix3)andtheCambodianSihanoukville

Portroad(Figure11inAppendix6)–bothpass

throughexistingnationalparks.Thedegreeto

whichtheseparkscanprovidesufficient

17WhilethereareafewareasprotectedwithinLaoPDR,noneiswithintheIUCNCategoriesI‐IVwithitshigherlevelsofprotection.

Figure17.LandcoverchangeofSnuolWildlifeReserve,Cambodia.Landsatimageswereobtainedforthreetimepoints:whentheroadwas(1)absent(1990),(2)recentlycompleted(2001),and(3)hadexistedforsometime(2009).FromClements,etal.2014.

Figure16.RailandroadprojectswithinCICPECinrelationto

protectedareas(IUCNCategoriesI‐IV),forestcover,forestloss,

andforestgain

deterrencewilldependlargelyonenforcementcapacity.

APPENDIX9:MitigationStrategies:UsingUmbrellaSpeciesforMitigationPlanningAppendix5describedhowumbrellaspeciesrangescouldbeusedtohelpidentifyhigh‐valueareasfor

wildlife,usingtheexampleoftigerandcloudedleopardranges.Thissameinformationcanbeusedto

helpguidethedevelopmentofmitigationstrategiesusingamitigationhierarchyframework(Figure1).

Giventhecriticallyendangeredstatusoftigers,theirhabitatsshouldbeconsideredno‐goor

“avoidance”areas,withnoBRIroadorrailprojects

permittedwithinaseveralkilometerbufferoftheir

range.Cloudedleopards,withsubstantiallylarger

populations,alowerendangermentstatus,andmore

resiliencetohumandevelopmentmayallowfor

moreflexiblemitigationstrategies.Whenpossible,

BRIrailandroadprojectsshouldalsoavoidthe

cloudedleopardrangesandbuffers,butifnot,

strategiesshouldbeputinplacetoreduceimpacts.

Possiblereductionoptionsincludedesignatingor

strengtheningprotectedareasthatsafeguardhabitat

affectedbytherangeandcreatingtunnelsor

overpassesforpassagearoundtransportation

corridors.Becausethesereductionsinthethreat

maynotcompletelyeliminatetheriskfrom

transportationinfrastructureontheaffected

cloudedleopardpopulations,compensatoryoffset

actioncanalsobeconsidered.Onepotentialoffset

mightbethecreationandprotectionofhabitat

corridorsthatlinkdisjuncthabitatpatcheswithinthe

CICPECregionbutaredistantfromthebusy

transportationcorridors(Figure18).

Figure18.Potentialhabitatcorridorthatcould

connecttwodisjuncthabitatrangesforclouded

leopards.

APPENDIX10:DatabasesofNationalEnvironmentalPolicies

Databasesofnationalenvironmentalregulations,financialincentives,overarchingpoliciesandframeworks,andvoluntaryapproaches

Database Website DataDescription

Faolex faolex.fao.org Nationallaws,regulationsandpoliciesrelatingtofood,agriculture,andrenewablenaturalresources

Ecolex www.ecolex.org Nationalenvironmentallegislation,courtdecisions,andliterature

ClimateChangeLawsoftheWorld

web.law.columbia.edu/climate‐change/resources/climate‐change‐laws‐world

Informationoncountry‐specificclimatechangepolicies,laws,plans,andinitiativesbysector(climate,energy,airpollution,forestandlanduse,environmentalimpactassessments,andadaptationandresiliency)

NewClimatePolicyDatabase

climatepolicydatabase.org InformationonGHGmitigationpoliciesbysector(electricityandheat,industry,buildings,transport,andagricultureandforestry)

LegalAtlas www.legal‐atlas.net/ Dynamicdatabaseofnationalenvironmentallegislationthatcanbecomparedacrosstopicsandcountries

APPENDIX11:InternationalAgreements

Agreement Description DateSigned

Conventiononwetlands(RamsarConvention)

TreatyestablishedbyUNESCOfortheclassification,conservation,andsustainableuseofwetlands.Theconventionidentifieswetlandsofinternationalimportance(especiallythoseimportantforwaterfowl)andplacesrestrictionsondevelopmentatthosesites.

1971*

Conventionconcerningtheprotectionoftheworldculturalandnaturalheritage

TreatytoprotectnaturalandculturalheritagethroughthedesignationofWorldHeritageSites(WHS).WHSareoverseenbyUNESCO.Signatorycountriesmust“taketheappropriatelegal,scientific,technical,administrativeandfinancialmeasuresnecessaryfortheidentification,protection,conservation,presentationandrehabilitationofthisheritage.”

1972*

Conventiononinternationaltradeinendangeredspeciesofwildfaunaandflora(CITES)

Internationalagreementtoensurethatinternationaltradeofplantandanimalspeciesdoesnotthreatentheirsurvival.SpeciescoveredbyCITESarelistedinthreeappendices,witheachappendixrequiringadifferentlevelofprotection.TradeofspeciesinAppendixIisprohibitedcompletely,whiletradeofAppendixIIandIIIspeciesiscloselyregulatedbutallowedwithincertainlimitsandwithproperpermitting.

1973*

Europeanagreementonmaininternationaltrafficarteries(AGR)

AGRdefinestheE‐roadnetworkofroutesofstrategicimportanceforinternationaltrafficflowswithinEurope.Theagreementsetsthestandards,includingconsiderationofthedirectandindirecteffectsofroadsandtrafficonpeople,faunaandflora;soils,sub‐soils,water,air,microclimate;landscape,physicalpropertyandculturalheritage.EIAsarementionedasmeasurestoaddressadverseimpacts.

1975

Baselconventiononthecontroloftransboundarymovementsofhazardouswastesandtheirdisposal(BaselConvention)

AimsoftheBaselconventionarethreefold,1)toreducethegenerationofhazardouswasteandpromotesoundmanagementofanyhazardouswastethatisgenerated,2)torestricttransboundarymovementofhazardouswaste,unlessmovingthewasteacrossbordersisforenvironmentallysoundmanagement,and3)todeveloparegulatorysystemforthetransboundarymovementofhazardouswaste.

1989*

Theconventionofenvironmentalimpactassessmentinatransboundarycontext(EspooConvention)

ThetreatywascreatedbytheUnitedNationsEconomicCommissionforEurope.ItobligatesPartiestoconductEnvironmentalImpactAssessmentsatanearlystageofplanningforcertainactivitiesthatareexpectedtocauseenvironmentalharms.ItalsorequiresthatStatesmustnotifyeachotherwhenactivitiesareexpectedtoresultintransboundaryenvironmentalharms.

1991*

UnitedNationsframeworkconventiononclimatechange(UNFCCC)

Theobjectiveofthetreatyistostabilizethegreenhousegasconcentrationoftheatmospheretopreventanthropogenicinterferencewiththeclimatesystembylimitingcountry‐levelemissionsofthesegasses.UNFCCCConferencesofthePartiesareheldannuallytoaddresscurrentissuesofclimatechangeandhaveresultedintheKyotoProtocolandtheParisClimateAgreement.

1992(KyotoProtocol)*

2015(ParisAgreement)*

Conventiononbiologicaldiversity(CBD)

Thetreatyhasthegoalsofconservingbiodiversity,sustainablyusingthecomponentsofbiodiversity,andequitablysharingthebenefitsprovidedbygeneticresourcesofbiologicaldiversity.Signatoriesmustdevelopnationalstrategiesdetailinghowthecountrywillachievethesegoals.The2011‐2020strategicplanforbiodiversity(createdbytheCBD)includestheAichibiodiversitytargets,whicharemeasurabletargetsfor2020thatwillhelpachievetheconvention’sgoals.

1992*

UnitedNationsconventiontocombatdesertification

Theconventionaimstocombatdesertificationandmitigatetheeffectsofdroughtthroughsustainablelandmanagement.Theconventionencouragescooperationbetweendevelopedanddevelopingcountriestoenhanceknowledgesharingandtechnologytransferforsustainablelandmanagement.Thetreatycallsfornationalandregionalactionprogramstoimplementtheconvention;theseprogramscallforspecificmeasuresthatcanbetakentocombatdesertification.

1994*

UNECE conventiononaccesstoinformation,public

Theconventionestablishesrightsofthepublic(individualsandtheirassociations)toreceiveenvironmentalinformationthatisheldbypublicauthorities;toparticipateinenvironmentaldecision‐makingbycommentingonandreceivingfeedbackon,for

1998

participationindecision‐makingandaccesstojusticeinenvironmentalmatters(AarhusConvention)

example,proposalsforprojects,plans,andprogramsaffectingtheenvironment;andtochallengepublicdecisionsthathavebeenmadewithoutrespectingthetwoaforementionedrightsorenvironmentallawingeneral.

*Chinaisasignatorytoagreement.

APPENDIX12:MultilateralDevelopmentBankEnvironmentalPoliciesMDB KeyAspectsofRelevantEnvironmentalPolicy Year

Introduced

InternationalFinanceCorporation(IFC)PerformanceStandards

TheIFCPerformanceStandardsonEnvironmentalandSocialSustainability(PS)havegainedrecognitionastheglobalbestpracticestandardforassessingandmitigatingnegativeenvironmentalandsocialoutcomesrelatedtolargeinfrastructureprojects.Thestandardsadherestrictlytothemitigationhierarchy,placinghighimportanceonavoidanceofimpactifpossible.EspeciallyrelevantfortheenvironmentrisksrelatedtotransportationinfrastructureareResourceEfficiencyandPollutionPrevention(PS3)andBiodiversityConservationandSustainableManagementofLivingNaturalResources(PS6).EachperformancestandardhasanaccompanyingguidancenotethatprovidesmoretechnicaldetailsabouthowborrowersshouldadheretothePS.

Revised2012

WorldBankEnvironmentalandSocialFramework

Inadditiontoprotectingthepoorandtheenvironmentandensuringsustainabledevelopment,WB’sEnvironmentalandSocialFramework(ESF)addresses,amongotherthings,transparency,non‐discrimination,socialinclusion,publicparticipation,andaccountability.TheEnvironmentalandSocialStandardsmirrortheIFC’sPerformanceStandardsveryclosely.

Revised2018

AsianDevelopmentBank(ADB)

ADB'sSafeguardPolicyStatement(SPS)governingtheenvironmentalandsocialsafeguardsofADB'soperationsareacornerstoneofitssupporttoinclusiveeconomicgrowthandenvironmentalsustainabilityinAsiaandthePacific.TheobjectivesoftheSPSaretoavoid,orwhenavoidanceisnotpossible,tominimizeandmitigateadverseprojectimpactsontheenvironmentandaffectedpeople,andtohelpborrowersstrengthentheirsafeguardsystemsanddevelopthecapacitytomanageenvironmentalandsocialrisks.

Revised2009

AfricanDevelopmentBank(AfDB)

AfDB’sIntegratedSafeguardsSystemconsistsoffourinterrelatedcomponents:IntegratedSafeguardsPolicyStatement(PS),OperationalSafeguards(OS),EnvironmentalandSocialAssessmentProcedures.ThePSdescribescommonobjectivesoftheBank’ssafeguardsandlaysoutpolicyprinciples.TheOSareasetoffivesafeguardrequirementsthatBankclientsareexpectedtomeetwhenaddressingsocialandenvironmentalimpactsandrisks.TheImpactAssessmentGuidanceNotesprovidetechnicalguidancetotheBank’sborrowersorclientsonstandardsofsectorissues,suchasroadsorfisheries,oronmethodologicalapproachesclientsorborrowersareexpectedtoadopttomeetOSstandards.OS3,Biodiversity,RenewableResources,andEcosystemServices,isespeciallyrelevanttoaddressingenvironmentalrisksfromBRItransportinfrastructure.

2013

EuropeanBankforReconstructionandDevelopment(EBRD)

EBRD’sEnvironmentalandSocialPolicy(ESP)putssafeguardsinplacetopreventorminimizeanyadverseenvironmentalorsocialimpacts,toimprovetheproject’sefficiency,andmaximizebenefitsforthewidercommunityandfuturegenerations.ESP outlineshowtheEBRDwilladdresstheenvironmentalandsocialimpactsofitsprojectsbydefiningtherespectiverolesandresponsibilitiesofboththeBankanditsclientsindesigning,implementingandoperatingprojects;settingastrategicgoaltopromoteprojectswithhighenvironmentalandsocialbenefits;andmainstreamingenvironmentalandsocialsustainabilityconsiderationsintoallitsactivities.

Revised2014

AsianInfrastructureInvestmentBank(AIIB)

TherecentlyestablishedAIIBreleasedthefirstdraftofitsEnvironmentalandSocialFrameworkinFebruaryof2016,whichincludesanEnvironmentalandSocialExclusionList—alistofprojecttypesoractivitiesthatthebankrefusestofinanceonenvironmentalorsocialgrounds.Inmanyways,theAIIBEnvironmentalandSocialFrameworkalignswithsimilarstandardsreleasedbyotherbanks,butitalsoreliesheavilyonitspartners’standards(Weiss2017).

2016

NewDevelopmentBank(NDB)

TheNDB’sEnvironmentalandSocialFramework(ESF)includesanenvironmentalandsocialpolicyaswellasenvironmentalandsocialstandards(ESS).ESS1,theEnvironmentalandSocialAssessment,isparticularlyrelevant.

2016

APPENDIX13:ChinesePolicyBankPoliciesChinaPolicyBanks KeyAspectsofRelevantEnvironmentalPolicy Year

Introduced

Export‐ImportBankofChina(ChinaEximBank)

The2007GuidelinesforEnvironmentalandSocialImpactAssessmentofChinaExportandImportBank’sLoanProjectsrequiresenvironmentalimpactassessments,monitoring,andreviewofprojectimpactsforallprojectsbeforeaprojectgainsapproval.Whendeemednecessary,environmentalandsocialresponsibilitiesmaybeincludedintheloancontract.TheEximBankalsohastherighttomonitortheclient’simplementationofthemitigationactivities(FOE2016;Leungetal.2013).

2007;2015

ChinaDevelopmentBank(CDB)

CDBhastransparentsustainabledevelopmentobjectives–includinganobjectiveonenvironmentalprotectionforclimate,ecology,cleanenergy,andlow‐carbonliving–butspecificenvironmentalpoliciesandtheircontentarenotavailabletothepublic(FOE2016).In2006,CDBpledgedtoabidebytheUnitedNationsGlobalCompact10principlesinhumanrights,labourstandards,environment,andanti‐corruption.CBDproducedaseriesofnon‐bindingframeworkstopromoteenvironmentally‐friendlybusinesses,includinganannualWorkPlanforLoanstoReducePollutionandEmissions,GuidelinesonEnvironmentalProtectionProjectDevelopmentReview,andGuidelinesonSpecialLoansforEnergyConservationandEmissionReduction(FriendsoftheEarth(FOE)2016;Renetal.2017).

2004

APPENDIX14:ChinesePoliciesDomain Specific

PoliciesKeyAspectsofRelevantEnvironmentalPolicy Year

Introduced

GeneralEnvironmentalLawsandGuidelines

EnvironmentalProtectionLawofthePeople’sRepublicofChina

Theoriginal1989lawprovidedtheimpetusforenvironmentalimpactassessmentsforconstructionprojectswithinChina.TheEIAprocesswasadministrative,notstatutorily.In2014,Chinaupdatedthelawtoinclude,amongotherprovisions,stricterpenaltiesandgreateropportunitiesforpublicenvironmentallitigation.

1989/amended2014

LawofthePeople'sRepublicofChinaonEnvironmentalImpactAssessment(EIALaw)

TheEIALawsetsforthrequirementsforseveraltypesofstrategicenvironmentalassessments(SEAs)withinChina.Underthislaw,planenvironmentalImpactassessments(PEIA)arelegallyrequiredformajoreconomicdevelopmentactivities,integratedplans(suchaslanduse,regionaldevelopment,andwatersheddevelopment).ThesePEIAsisusedtointegrateenvironmentalconsiderationsintoallphasesofthepreparationofspatialandlanduseplans.Asecondtypeofstrategicenvironmentalassessmentarerequiredforsectoralplans(forexample,industry,agriculture,husbandry,forestry,energy,waterconservancy,transportation,andnaturalresourcesdevelopment)withinChina(DusiandXi2009;Wuetal.2010).

2003

GuidelinesonEnvironmentalProtectionforOverseasInvestmentandCooperation

Thesevoluntaryguidelinesrepresentakeypolicyreleasedin2013byMOFCOMandtheformerMEP(nowMEP).TheseguidelinesrecommendthatcompaniescompleteEIAs,developenvironmentalmitigationmeasures,andworkwithlocalcommunitiestoidentifypotentialnegativeenvironmentalandsocialimpacts(Leungetal.2013).

2013

GuidelinesforSocialResponsibilityinOutboundMiningInvestments

TheChinaChamberofCommerceofMetals,MineralsandChemicalsImportersandExporters(CCCMC)developedtheseguidelinestoassistChinesecompaniesimprovetheirenvironmentalandsocialperformanceinoverseasmininginvestments.Thoughtheseguidelinesarenotstrictlyfocusedontransportationinfrastructure,theyrepresentthefirstdetailedvoluntarystandardsforforeigninfrastructureinvestmentbyChinesecompanies.

2014

GuidelinesofSustainableInfrastructureforChineseInternationalContractors

TheChinaInternationalContractorsAssociation(CHINCA)developedguidelinestoassistChineseinternationalcontractorsinbuildinginfrastructurethatmeetsleadingsustainabilitystandards.Theguidelinescoverfivemajorphasesofinfrastructuredevelopment:funding,planninganddesign,building,operationandmaintenance,andclosure.

2017

BRI‐SpecificGuidelines

VisionandActionsonJointlyBuildingSilkRoadEconomicBeltand21stCenturyMaritimeSilkRoad

ThisvisiondocumentoutlinesChina’svisionandmissionoftheBRI,highlightingsustainabledevelopment,cooperationbetweennations,andtheimportanceoffreetrade.Issuesrelatedtotheenvironmentarementionedonlybrieflyindocument,signalingprioritiesforcooperationinnaturalresourceinvestmentsandindustries;cooperationinecologicalconservation,biodiversityprotection,andstrategiesforclimatechange;andorganizingpublicinterestinecologicalprotectionforthebenefitofthepublic.

2015

GuidanceonPromotingGreenBeltandRoad

Thehigh‐levelandnon‐bindingdocumentoutlinesabroadrangeofsignificantenvironmentalissuesthatwillbeaddressedunderBRI.ThedocumentdirectsChineseenterprisesengagedinBRIprojectsto“voluntarilyobeylocalenvironmentalprotectionlaws,regulations,standards,codes,honorenvironmentalandsocialresponsibilitiesandreleaseannualenvironmentalreports.”Thedocumentspromotearangeoftasksincludinggreensupplychains;partnershipswithNGOsandresearch

2017

organizations;environmentalprotectionplatformcooperation;capacitybuilding;greentechnologytransfers;sharingofenvironmentalprotectioninformationandbigdata;andenhancinggreenguidanceforcorporatebehavior.

VisionforMaritimeCooperationundertheBeltandRoadInitiative

Thisvisiondocumentacknowledgesthatoceansasanecosystemcontributesvaluablenaturalassets.Thedocumentprovidesspecific,non‐bindingrecommendationsfortheprotectionofthemarineenvironment.Thegreendevelopmentsection(4.1)highlightsensuringthehealthoftheoceanforpresentandfuturegenerations;cooperationamongnationstoundertakeconservationandpreservationofmarineecosystemservices,ecosystems,andspecies;establishingeffortstomonitor,evaluate,preserve,andrestoremarineandcoastalsystems;jointlytacklingmarinepollutionissues;demonstratinglow‐carbondevelopmentinmaritimesectors;supportingsmall‐islandnationsinadaptingtoclimatechange;andstrengtheningcooperationforinternationalbluecarbonprograms.

2017

BeltandRoadEcologicalandEnvironmentalCooperationPlan

ThiscooperationplanwaswrittentofitinwiththepreviousVisionandActionsonJointlyBuildingSilkRoadEconomicBeltand21stCenturyMaritimeSilkRoad,GuidanceonPromotingGreenBeltandRoad,andChina’s13thFive‐YearPlan.TheEnvironmentalcooperationplanpromotesactivitiessuchascooperationbetweencountriesforenvironmentalprotection,applyingandestablishinggreenfinancialinstruments,increasingNGOandthink‐tankinvolvementinenvironmentalplanningandpartnerships,strengtheninggreencorporatebehavior,andenhancinggreensupplychains.

2017

APPENDIX15:IndustryandNGOEnvironmentalStandardsStandard Description

EquatorPrinciples TheEquatorPrinciples(EPs)isariskmanagementframeworkforprivatefinancialinstitutionstodetermine,assess,andmanageenvironmentalandsocialriskinprojects.TheEPframeworkismodelledcloselyontheIFC’sperformancestandards.Currently94globalfinancialinstitutionsfrom37countrieshaveadoptedtheframeworkformanagingenvironmentalandsocialrisk.

SuRe®Standard TheStandardforSustainableandResilientInfrastructure(SuRe®),developedbyGlobalInfrastructureBasel,isaglobalvoluntarystandardwhichintegrateskeycriteriaofsustainabilityandresilienceintoinfrastructuredevelopmentandupgrade,through14themescovering61criteriaacrossgovernance,socialandenvironmentalfactors.

CEEQUALassessmentrating

CEEQUALisanevidence‐basedsustainabilityassessment,rating,andawardsschemeforcivilengineering,infrastructure,landscaping,andpublicrealmprojects.

Envision®ratingsystem

TheEnvision®systemprovidesasetofguidelinestoaidandquantifythesustainabilityofaninfrastructureprojectduringtheplanningandpreliminarydesignphases.

Greenroads®ratingsystem

TheGreenroadsRatingSystemismeasuresandmanagesustainabilityontransportationprojectsusingenvironmental,social,andeconomicperformancemeasures.Projectsareevaluatedbyanindependent,expert,third‐partyreview.

NaturalCapitalProtocol

FrameworkcreatedbytheNaturalCapitalCoalitiontoprovideguidanceonhowtoincorporatehumaninteractionwithnatureandnaturalcapitalintoactionableinformationforbusinessmanagerstoinformdecisions.

DATALINKS/CITATIONSFORMAPSAll maps were created by Dr. Andrew Jacobson, Duke University, unless otherwise noted.

Cities:Reed,T.andS.Trubetskoy(2018)TheBeltandRoadInitiativeandtheValueofUrbanLand.World

Bankmimeo.Ports:Reed,T.andS.Trubetskoy(2018)TheBeltandRoadInitiativeandtheValueof

UrbanLand.WorldBankmimeo.

Roads&rails:Reed,T.andS.Trubetskoy(2018)TheBeltandRoadInitiativeandtheValueofUrbanLand.

WorldBankmimeo.Administrativeboundaries:GADMv2.8‐https://gadm.org/index.html.

Forestchange:GlobalForestWatch2015v1.3‐https://www.globalforestwatch.org/.

Hansen, Matthew C., et al. "High-resolution global maps of 21st-century forest cover change." science 342.6160 (2013): 850-853.

Cropland2015:ESACCIlandcoverV2.0‐https://www.esa‐landcover‐cci.org/?q=node/175

Cite–ESAClimateChangeInitiative–LandCoverProject2017

scientificpublicationontheresultsofresearchactivitiesthatuseoneorseveralCCI‐LCproductsas

input,youshallacknowledgetheESACCILandCoverprojectinthetextofthepublicationandprovide

theprojectwithanelectroniccopyofthepublication(contact@esalandcover‐cci.org).

Protectedareas:UNEP&IUCN.2018.TheWorldDatabaseonProtectedAreas(WDPA).Cambridge:UNEP‐

WCMC.https://protectedplanet.net/

Pangolinhub&traffickingroutes:Chouvy,P‐A.2013.AnAtlasofTraffickinginSoutheastAsia.London:

I.B.Tauris.

IntactForestLandscape2013:Potapov, Peter, et al. "The last frontiers of wilderness: Tracking loss of intact forest landscapes from 2000 to 2013." Science Advances 3.1 (2017): e1600821.

http://www.intactforests.org/

CentersofEndemism:Li BV, Hughes AC, Jenkins CN, Ocampo-Peñuela N, Pimm SL (2016) Remotely Sensed Data Informs Red List Evaluations and Conservation Priorities in Southeast Asia. PLoS ONE 11(8): e0160566. https://doi.org/10.1371/journal.pone.0160566

Speciesdistributions(cloudedleopard&tiger):Goodrich, J., Lynam, A., Miquelle, D., Wibisono, H., Kawanishi, K., Pattanavibool, A., Htun, S., Tempa, T., Karki, J., Jhala, Y. & Karanth, U. 2015. Panthera tigris. The IUCN Red List of Threatened Species 2015: e.T15955A50659951.

http://dx.doi.org/10.2305/IUCN.UK.2015-2.RLTS.T15955A50659951.en. Downloaded on 10 September 2018.

Grassman, L., Lynam, A., Mohamad, S., Duckworth, J.W., Bora, J., Wilcox, D., Ghimirey, Y., Reza, A. & Rahman, H. 2016. Neofelis nebulosa. The IUCN Red List of Threatened Species 2016: e.T14519A97215090.

http://dx.doi.org/10.2305/IUCN.UK.2016-1.RLTS.T14519A97215090.en. Downloaded on 10 September 2018.

LITERATURECITEDAfDB.2015.PaymentforEnviromentalServices:APromisingToolForNaturalResourcesManagementIn

Africa.Abidjan,Coted’Ivoire.

Ahmed,S.E.etal.2014.“RoadNetworksPredictHumanInfluenceonAmazonianBirdCommunities.”ProceedingsoftheRoyalSocietyB:BiologicalSciences281(1795):20141742–20141742.http://rspb.royalsocietypublishing.org/cgi/doi/10.1098/rspb.2014.1742.

AIIB.2016.AIIB:EnvironmentalandSocialFramework.

Albers,H.J.2010.“SpatialModelingofExtractionandEnforcementinDevelopingCountryProtectedAreas.”ResourceandEnergyEconomics32(2):165–79.

Alix‐Garcia,Jennifer,KatharineRESims,PatriciaYanez‐Pagans,andElizabethNShapiro.2013.“OnlyOneTreefromEachSeed?EnvironmentalEffectivenessandPovertyAlleviationinProgramsofPaymentsFroEcosystemServices.”7(4):56.

Alpízar,Francisco,AnnaNordén,AlexanderPfaff,andJuanRobalino.2017a.“SpilloversfromTargetingofIncentives:ExploringResponsestoBeingExcluded.”JournalofEconomicPsychology59:87–98.

———.2017b.“UnintendedEffectsofTargetinganEnvironmentalRebate.”EnvironmentalandResourceEconomics67(1):181–202.

Ambani,MildredM.2017.“GISAssessmentofEnvironmentalFootprintsofStandardGaugeRailway(SGR)onNairobiNationalPark,Kenya.”UniversityofNairobi,SchoolofEngineering.

Andam,KwawSetal.2008.“MeasuringtheEffectivenessofProtectedAreaNetworksinReducingDeforestation.”ProceedingsoftheNationalAcademyofSciencesoftheUnitedStatesofAmerica105(42):16089–94.

Andersen,LykkeE.1996.“CausesofDeforestationintheBrazilianAmazon.”JournalofEnvironment&Development5(3):309–28.

Anderson,ChristaM.,ChristopherB.Field,andKatharineJ.Mach.2017.“ForestOffsetsPartnerClimate‐ChangeMitigationwithConservation.”FrontiersinEcologyandtheEnvironment15(7):359–65.

Andrews,KimberlyM.,TomA.Langen,andRichardP.J.H.Struijk.2015.“Reptiles:OverlookedbutOftenatRiskfromRoads.”InHandbookofRoadEcology,eds.RodneyvanderRee,ClaraGrilo,andDanielJ.Smith.WestSussex,U.K.:JohnWiley&Sons,Incorporated.

Angelo,GinoJDetal.2006.“EvaluationofWildlifeWarningReflectorsforAlteringWhite‐TailedDeerBehaviorAlongRoadways.”WildlifeSociety…34(4):1175–83.http://onlinelibrary.wiley.com/doi/10.2193/0091‐7648(2006)34[1175:EOWWRF]2.0.CO;2/abstract%5Cn%3CGotoISI%3E://WOS:000243173600036.

Angelsen,Arild.2007.WorldBankPolicyResearchWorkingPaperForestCoverChangeinSpaceandTime :CombiningthevonThünenandForestTransitionTheories.

Armsworth,P.R.,G.C.Daily,P.Kareiva,andJ.N.Sanchirico.2006.“LandMarketFeedbacksCanUndermineBiodiversityConservation.”ProceedingsoftheNationalAcademyofSciences103(14):5403–8.

Asher,Sam,TeevratGarg,andPaulNovosad.2017.TheEcologicalFootprintofTransportation

Infrastructure.

Asher,Sam,andPaulNovosad.2015.“MarketAccessandStructuralTransformation:EvidencefromRuralRoadsinIndia.”

Babińska‐Werka,Joanna,DagnyKrauze‐Gryz,MichałWasilewski,andKarolinaJasińska.2015.“EffectivenessofanAcousticWildlifeWarningDeviceUsingNaturalCallstoReducetheRiskofTrainCollisionswithAnimals.”TransportationResearchPartD:TransportandEnvironment38:6–14.

Barber,ChristopherP.,MarkA.Cochrane,CarlosM.Souza,andWilliamF.Laurance.2014.“Roads,Deforestation,andtheMitigatingEffectofProtectedAreasintheAmazon.”BiologicalConservation177:203–9.http://dx.doi.org/10.1016/j.biocon.2014.07.004.

Baylis,Kathy,DonFullerton,andPayalShah.2016.WhatDrivesForestLeakage?

Bechtel,Peter,andSeanNazerali.2016.ANationalBiodiversityOffsetSystem:ARoadMapforMozambique.https://www.profor.info/sites/profor.info/files/MozambiqueOffsetRoadmap_1.pdf.

Benítez‐López,Ana,RobAlkemade,andPitaA.Verweij.2010.“TheImpactsofRoadsandOtherInfrastructureonMammalandBirdPopulations:AMeta‐Analysis.”BiologicalConservation143(6):1307–16.

Bennett,Genevieve,MelissaGallant,ContributingAuthor,andKerryTenKate.2017.“StateofBiodiversityMitigation2017:MarketsandCompensationforGlobalInfrastructureDevelopment.”:92.http://www.forest‐trends.org/dir/em_newsletter%0Ahttp://www.forest‐trends.org/dir/em_newsletter%0Ahttp://www.forest‐trends.org/dir/em_newsletter.

Bennett,Genevieve,andFranziskaRuef.2016.AlliancesforGreenInfrastructure:StateofWatershedInvestment2016.http://www.forest‐trends.org/documents/files/doc_5463.pdf.

Bennett,MTetal.2014.“China’sConversionofCroplandtoForestProgramforHouseholdDeliveryofEcosystemServices:HowImportantIsaLocalImplementationRegimetoSurvivalRateOutcomes?”Forests5(9):2345–76.

BenYishay,Ariel,BradleyParks,DanielRunfola,andRachelTrichler.2016.“ForestCoverImpactsofChineseDevelopmentProjectsinEcologicallySensitiveAreas.”AidDataWorkingPaper(December2016).

Bhattarai,K,DConway,andMYousef.2009.“DeterminantsofDeforestationinNepal’sCentralDevelopmentRegion.”JournalofEnvironmentalManagement91(2):471–88.

Blackman,Allen.2015.“StrictversusMixed‐UseProtectedAreas:Guatemala’sMayaBiosphereReserve.”EcologicalEconomics112:14–24.

Blackwell,BradleyF.,TravisL.DeVault,andThomasW.Seamens.2015.“UnderstandingandMitigatingtheNegativeEffectsofRoadLightingonEcosystems.”InHandbookofRoadEcology,eds.AndrewvanderRee,DanielJ.Smith,andClaraGrilo.WestSussex,U.K.:JohnWiley&Sons,Incorporated,143–50.

Booth,Sarah.2017.AnalysisofExistingEnvironmentalPolicyDatabases.

Bremer,LeahL.etal.2016.“OneSizeDoesNotFitAll:NaturalInfrastructureInvestmentswithintheLatinAmericanWaterFundsPartnership.”EcosystemServices17:217–36.

http://dx.doi.org/10.1016/j.ecoser.2015.12.006.

Broussard,Giovanni.2017.“BuildinganEffectiveCriminalJusticeResponsetoWildlifeTrafficking:ExperiencesfromtheASEANRegion.”ReviewofEuropean,ComparativeandInternationalEnvironmentalLaw26(2):118–27.

Brown,PhillipH.,andChristopherL.Lant.1999.“TheEffectofWetlandMitigationBankingontheAchievementofNo‐Net‐Loss.”EnvironmentalManagement23(3):333–45.

Bruschi,Daniele,DavideAstiasoGarcia,FrancoGugliermetti,andFabrizioCumo.2015.“CharacterizingtheFragmentationLevelofItalian’sNationalParksDuetoTransportationInfrastructures.”TransportationResearchPartD:TransportandEnvironment36:18–28.http://dx.doi.org/10.1016/j.trd.2015.02.006.

Bull,JosephW.etal.2013.“BiodiversityOffsetsinTheoryandPractice.”Oryx47(3):369–80.

Bulte,ErwinH.,LeslieLipper,RandyStringer,andDavidZilberman.2008.“PaymentsforEcosystemServicesandPovertyReduction:Concepts,Issues,andEmpiricalPerspectives.”EnvironmentandDevelopmentEconomics13(3):245–54.

BusinessandBiodiversityOffsetsProgramme(BBOP).2009.Business,BiodiversityOffsetsandBBOP:AnOverview,.Washington,DC.

———.2012.BiodiversityOffsetDesignHandbook‐Updated.Washington,DC.

Cader,Masud,AlexanderCantor,ShujieShao,andMingqingLiu.EconomicsoftheBRI:Co‐InvestmentandCreatingMarketsalongtheBeltandRoad.Washington,DC.

Caro,Tim,AndrewDobson,AndrewJ.Marshall,andCarlosA.Peres.2014.“CompromiseSolutionsbetweenConservationandRoadBuildingintheTropics.”CurrentBiology24(16):R722–25.http://dx.doi.org/10.1016/j.cub.2014.07.007.

Carter,MasonC.,R.C.Kellison,andR.ScottWallinger.2015.ForestryintheU.S.South.BatonRouge:LouisianaStateUniversityPress.

CCCMC.2015.GuidelinesforSocialResponsibilityinOutboundMiningInvestments.https://www.emm‐network.org/wp‐content/uploads/2015/03/CSR‐Guidelines‐2nd‐revision.pdf.

Chai,Jian,QuanYingLu,ShouYangWang,andKinKeungLai.2016.“AnalysisofRoadTransportationEnergyConsumptionDemandinChina.”TransportationResearchPartD:TransportandEnvironment48:112–24.http://dx.doi.org/10.1016/j.trd.2016.08.009.

Chaker,A.,K.El‐Fadl,L.Chamas,andB.Hatjian.2006.“AReviewofStrategicEnvironmentalAssessmentin12SelectedCountries.”EnvironmentalImpactAssessmentReview26(1):15–56.

Chen,Zhenhua,JunboXue,AdamZ.Rose,andKingsleyE.Haynes.2016.“TheImpactofHigh‐SpeedRailInvestmentonEconomicandEnvironmentalChangeinChina:ADynamicCGEAnalysis.”TransportationResearchPartA:PolicyandPractice92:232–45.http://dx.doi.org/10.1016/j.tra.2016.08.006.

Chomitz,KennethM,andDavidAGray.1996.“Roads,LandUse,andDeforestation:ASpatialModelAppliedtoBelize.”TheWorldBankEconomicReview10(3):487–512.

Chomitz,KennethM,andTimothySThomas.2003.“DeterminantsofLandUseinAmazônia :AFine‐

ScaleSpatialAnalysis.”AmericanJournalofa*griculturalEconomics85(4):1016–28.

Chouvy,Pierre‐Arnaud,ed.2013.AnAtlasofTraffickinginSoutheastAsia:TheIllegalTradeinArms,Drugs,People,CounterfeitGoodsandNaturalResourcesinMainlandSoutheastAsia.NewYork:I.B.Tauris&Co.Ltd.

Chun,Zhang.2017.“ChinaNeedsUrgentOversightofInvestments.”Chinadialogue.https://www.chinadialogue.net/article/show/single/en/10258‐China‐needs‐urgent‐oversight‐of‐investments.

Clements,GopalasamyReubenetal.2014.“WhereandHowAreRoadsEndangeringMammalsinSoutheastAsia’sForests?”PLoSONE9(12):1–26.

Cochrane,MarkA.,andWilliamF.Laurance.2002.“FireasaLarge‐ScaleEdgeEffectinAmazonianForests.”JournalofTropicalEcology18(3):311–25.

Coffin,AlisaW.2007.“FromRoadkilltoRoadEcology:AReviewoftheEcologicalEffectsofRoads.”JournalofTransportGeography15(5):396–406.

Correa,Juanetal.2018.EvaluatingInternationalClimateFinanceattheLocalLevel:ImpactsoftheOlhosD’AguaProject,anAmazonFund/REDDInitiative.

Cropper,Maureen,andCharlesGriffiths.1994.“TheInteractionofPopulationGrowthandEnvironmentalQuality.”TheAmericanEconomicReview84(2).

Cropper,Maureen,CharlesGriffiths,andMuthukumaraMani.1999.“Roads,PopulationPressures,andDeforestationinThailand,1976‐1989.”LandEconomics75(1):58.

Cropper,Maureen,JyotsnaPuri,andCharlesGriffiths.2001.“PredictingtheLocationofDeforestation:TheRoleofRoadsandProtectedAreasinNorthThailand.”LandEconomics77(2):172.

D’Alfonso,Tiziana,ChangminJiang,andValentinaBracaglia.2016.“AirTransportandHigh‐SpeedRailCompetition:EnvironmentalImplicationsandMitigationStrategies.”TransportationResearchPartA:PolicyandPractice92(635867):261–76.http://dx.doi.org/10.1016/j.tra.2016.06.009.

D’Angelo,Gino,andAndrewvanderRee.2015.“UseofReflectorsandAuditoryDeterrentstoPreventWildlife‐VehicleCollisions.”InHandbookofRoadEcology,eds.AndrewvanderRee,DanielJ.Smith,andClaraGrilo.WestSussex,U.K.:JohnWiley&Sons,Incorporated,213–18.

Dalgaard,Carl‐Johan,NicolaiKaarsen,OlaOlsson,andPabloSelaya.2018.“RomanRoadstoProsperity:PersistenceandNon‐PersistenceofPublicGoodsProvision.”UniversityofCopenhagen,mimeo:1–50.http://www.econ.ku.dk/pabloselaya/papers/RomanRoads_v9‐6_Feb2018.pdf.

Danyo,Stephen,SusmitaDasgupta,andDavidWheeler.2018.PotentialForestLossandBiodiversityRisksfromRoadImprovementinLaoPDR.

Dasgupta,Susmita,andDavidWheeler.2016.MinimizingEcologicalDamagefromRoadImprovementinTropicalForests.

Defries,RuthS.,ThomasRudel,MariaUriarte,andMatthewHansen.2010.“DeforestationDrivenbyUrbanPopulationGrowthandAgriculturalTradeintheTwenty‐FirstCentury.”NatureGeoscience3(3):178–81.

Deininger,Klaus,andBartMinten.2002.“DeterminantsofDeforestationandtheEconomicsof

Protection:AnApplicationtoMexico.”AmericanJournalofa*griculturalEconomics84(4):943–60.

Deng,Xiangzhengetal.2011.“PressureCookersorPressureValves:DoRoadsLeadtoDeforestationinChina?”JournalofEnvironmentalEconomicsandManagement61(1):79–94.

Dobson,Andrewetal.2006.“HabitatLoss,TrophicCollapse,andtheDeclineofEcosystemServices.”Ecology87(8):1915–24.https://esajournals.onlinelibrary.wiley.com/doi/abs/10.1890/0012‐9658%282006%2987%5B1915%3AHLTCAT%5D2.0.CO%3B2.

Dobson,AndrewP.etal.2010.“RoadWillRuinSerengeti.”Nature467(7313):272.

Dollar,David.2018.13BrookingsInstitutionIsChina’sDevelopmentFinanceaChallengetotheInternationalOrder?Washington,DC.

Donaldson,Dave.2010.NBERWorkingPaperSeriesRailroadsoftheRaj:EstimatingtheImpactofTransportationInfrastructure.

Dong,Suochengetal.2018.“AnEvaluationoftheEconomic,Social,andEcologicalRisksofChina‐Mongolia‐RussiaHigh‐SpeedRailwayConstructionandPolicySuggestions.”JournalofGeographicalSciences28(7):900–918.

Dorsey,Benjamin,MattiasOlsson,andLisaJ.Rew.2015.“EcologicalEffectsofRailwaysonWildlife.”InHandbookofRoadEcology,eds.RodneyvanderRee,DanielJ.Smith,andClaraGrilo.WestSussex,U.K.:JohnWiley&Sons,Incorporated,219–27.https://ebookcentral.proquest.com/lib/duke/detail.action?docID=1895495.

Dusi,Jiri,andJianXi.2009.StrategicEnvironmentalAssessmentandSoutheastAsia:AProgressReviewandComparisonofCountrySystemsandCases.Washington,DC.

Ekstrom,Jon,LeonBennun,andRobinMitchell.2015.ACross‐SectorGuideforImplementingtheMitigationHierarchy.Cambridge:TheBiodiversityConsultancyLtd.

Engel,Stefanie,StefanoPagiola,andSvenWunder.2008.“DesigningPaymentsforEnvironmentalServicesinTheoryandPractice:AnOverviewoftheIssues.”EcologicalEconomics65(4):663–74.

Esty,DanielC.E.,andMichaelPorter.2002.“RegulationandPerformance:ALeadingIndicatorofFutureCompetitiveness?”InTheGlobalCompetitivenessReport2001–2002,eds.MichaelPorteretal.NewYork:OxfordUniversityPress,78–101.

FederalTransitAdministration.2009.FTARailModernizationStudy:ReporttoCongress.

Felbab‐Brown,Vanda.2013a.“TheIllegalTradeinWildlfieinSoutheastAsiaandItsLinkstoEastAsianMarkets.”InAnAtlasofTraffickinginSoutheastAsia,ed.Pierre‐ArnaudChouvy.NewYork:I.B.Tauris&Co.Ltd.,137–54.

———.2013b.“TheJaggedEdge:IllegalLogginginSoutheastAsia.”InAnAtlasofTraffickinginSoutheastAsia,NewYork:I.B.Tauris&Co.Ltd.,113–36.

Ferraro,PaulJ.2008.“AsymmetricInformationandContractDesignforPaymentsforEnvironmentalServices.”EcologicalEconomics65(4):810–21.

———.2009.“RegionalReviewofPaymentsforWatershedServices:Sub‐SaharanAfrica.”JournalofSustainableForestry28(3–5):525–50.

———.2013.“MoreStrictlyProtectedAreasAreNotNecessarilyMoreProtective:Evidencefrom

Bolivia,CostaRica,Indonesia,andThailand.”EnvironmentalResearchLetters8(2).

Ferraro,PaulJ.,andR.DavidSimpson.2002.“TheCost‐EffectivenessofConservationPayments.”LandEconomics78(3):339–53.

Ferretti‐Gallon,Kalifi,andJonahBusch.2014.CenterforGlobalDevelopmentWorkingPaperWhatDrivesDeforestationandWhatStopsIt?AMeta‐AnalysisofSpatiallyExplicitEconometricStudies.

Forman,R.TT.1998.“RoadEcology:ASolutionfortheGiantEmbracingUs.”LandscapeEcology13(4).

Forman,RichardT.T.,DSperling,JABissonette,andAPClevenger.2003.RoadEcology :ScienceandSolutions.Washington,DC:IslandPress.

Foster,AndrewD,andMarkRRosenzweig.2003.“EconomicGrowthandtheRiseofForests.”TheQuarterlyJournalofEconomics118(2):601–37.

Freitas,SimoneR.,ToddJ.Hawbaker,andJeanPaulMetzger.2010.“EffectsofRoads,Topography,andLandUseonForestCoverDynamicsintheBrazilianAtlanticForest.”ForestEcologyandManagement259(3):410–17.

FriendsoftheEarth(FOE).2016.EmergingSustainabilityFrameworks:ChinaDevelopmentBankandChinaExport‐ImportBank.Washington,DC.https://1bps6437gg8c169i0y1drtgz‐wpengine.netdna‐ssl.com/wp‐content/uploads/wpallimport/files/archive/emerging‐sustainability‐frameworks‐CDB‐ChinaExim.pdf.

———.2017.InvestinginaGreenBeltandRoad?AssessingtheImplementationofChina’sGreenCreditGuidelinesAbroad.Washington,DC.https://1bps6437gg8c169i0y1drtgz‐wpengine.netdna‐ssl.com/wp‐content/uploads/2017/12/ChinaGreenBelt_Dec‐2017.pdf.

Fyumagwa,R.etal.2013.“RoadsasaThreattotheSerengetiEcosystem.”ConservationBiology27(5):1122–25.

Gallagher,KellySims,andQiQi.2018.PoliciesGoverningChina’sOverseasDevelopmentFinanceImplicationsforClimateChange.Boston,MA.https://sites.tufts.edu/cierp/files/2018/03/CPL_ChinaOverseasDev.pdf.

Geisler,Charles,andRagendradeSousa.2001.“FromRefugetoRefugee:TheAfricanCase.”PublicAdministrationandDevelopment21(2):159–170.

Geist,HelmutJ,andEricFLambin.2001.WhatDrivesTropicalDeforestation?

———.2002.“ProximateCausesandUnderlyingDrivingForcesofTropicalDeforestation.”Bioscience52(2).

Givoni,Moshe.2006.“DevelopmentandImpactoftheModernHigh‐SpeedTrain:AReview.”TransportReviews26(5):593–611.

GlobalEnvironmentalInstitution(GEI).2018.“HowPolicyBanksCanGoforGreen:CaseStudyonGreenFinance.”http://www.geichina.org/en/how‐policy‐banks‐can‐go‐for‐green‐case‐study‐on‐green‐finance/(October3,2018).

Goosem,Miriam.2015.“VulnerabilityandClimaticConditions:ParticularChallengesforRoadPlanning,ConstructionandMaintenance.”InHandbookofRoadEcology,eds.RodneyvanderRee,DanielJ.Smith,andClaraGrilo.WestSussex,U.K.:JohnWiley&Sons,Incorporated,397–406.

Gray,ThomasN.E.etal.2018.“TheWildlifeSnaringCrisis:AnInsidiousandPervasiveThreattoBiodiversityinSoutheastAsia.”BiodiversityandConservation27(4):1031–37.

Gunn,Jill,andBramF.Noble.2011.“ConceptualandMethodologicalChallengestoIntegratingSEAandCumulativeEffectsAssessment.”EnvironmentalImpactAssessmentReview31(2):154–60.http://dx.doi.org/10.1016/j.eiar.2009.12.003.

Gunson,KariE.,GiorgosMountrakis,andLindiJ.Quackenbush.2011.“SpatialWildlife‐VehicleCollisionModels:AReviewofCurrentWorkandItsApplicationtoTransportationMitigationProjects.”JournalofEnvironmentalManagement92(4):1074–82.http://dx.doi.org/10.1016/j.jenvman.2010.11.027.

Haddad,NickMetal.2015.“HabitatFragmentationandItsLastingImpactonEarth’sEcosystems.”ScienceAdvances1(2):1–9.

Hamer,AndrewJ.,ThomasE.S.Langton,andDavidLesbarreres.2015.“MakingaSafeLeapForward:MitigatingRoadImpactsonAmphibeans.”InHandbookofRoadEcology,eds.RodneyvanderRee,ClaraGrilo,andDanielJ.Smith.WestSussex,U.K.:JohnWiley&Sons,Incorporated.

Harangozo,Gabor,andCeciliaSzigeti.2017.“CorporateCarbonFootprintAnalysisinPractice–withaSpecialFocusonValidityandReliabilityIssues.”JournalofCleanerProduction167:1177–83.http://dx.doi.org/10.1016/j.jclepro.2017.07.237.

Hargrave,Jorge,andKrisztinaKis‐Katos.2013.“EconomicCausesofDeforestationintheBrazilianAmazon:APanelDataAnalysisforthe2000s.”EnvironmentalandResourceEconomics54(4):471–94.

Harris,NancyLetal.2012.“BaselineMapofCarbonEmissionsfromDeforestationinTropicalRegions.”Science336(June):1573–76.

Haya,BarbaraKresch.2010.“CarbonOffsetting:AnEfficientWaytoReduceEmissionsortoAvoidReducingEmissions?”UniversityofCalifornia,Berkeley.http://escholarship.org/uc/item/7jk7v95t.

Helsingen,Hanna(WWF‐Myanmar),Ben(UniversityCollegeLondon)Milligan,Michele(WWFGlobalScience)Dailey,andNirmal(WWF‐US)Bhagabati.2018.GreeningChina’sBelt&RoadInitiativeinMyanmar.

Herrera,Diego.2015.“SpilloverConservationaroundProtectedAreasintheBrazilianAmazon:FrontierDynamicsandVariationinNearbyDeforestation‘Blockage’(ReverseLeakage).”DukeUniversity.

Hillman,Jonathan.2018.““China’sBeltandRoadInitiative :FiveYearsLater”ATestimonyBy :”:1–12.

Holmgren,Peter,EdwardJumaMasakha,andHåkanSjöholm.1994.“NotAllAfricanLandIsBeingDegraded:ARecentSurveyofTreesonFarmsinKenyaRevealsRapidlyIncreasingForestResources.”Ambio23(7):390–95.

Honey‐Rosés,Jordi,KathyBaylis,andM.IsabelRamírez.2011.“ASpatiallyExplicitEstimateofAvoidedForestLoss.”ConservationBiology25(5):1032–43.

Hoyos,LauraE,MarceloRCabido,andAnaMCingolani.2018.“AMultivariateApproachtoStudyDriversofLand‐CoverChangesthroughRemoteSensingintheDryChacoofArgentina.”InternationaljournalofGeo‐Information7(170):1–15.

Humphrey,Chris.2016.“TimeforaNewApproachtoEnvironmentalandSocialProtectionatMultilateralDevelopmentBanks.”(April).

Ibisch,PierreL.etal.2016.“AGlobalMapofRoadlessAreasandTheirConservationStatus.”Science354(6318):1423–27.

ICMM,andIUCN.2013.IndependentReportonBiodiversityOffsets.https://www.icmm.com/website/publications/pdfs/biodiversity/biodiversity‐offsets%0A.

IFC.2012.“GuidanceNote6BiodiversityConservationandSustainableManagementofLivingNaturalResources.”(2006):1–70.http://www.ifc.org/wps/wcm/connect/a359a380498007e9a1b7f3336b93d75f/Updated_GN6‐2012.pdf?MOD=AJPERES.

IUCN.2016.IUCNPolicyonBiodiversityOffsets.http://cmsdata.iucn.org/downloads/iucn_biodiversity_offsets_policy_jan_29_2016.pdf.

Iuell,Bjørnetal.2003.COST341‐HabitatFragmentationDuetoTransportationInfrastructure:TheEuropeanReview.https://handbookwildlifetraffic.info/handbook‐wildlife‐traffic/.

Jackson,NathanD.,andLenoreFahrig.2011.“RelativeEffectsofRoadMortalityandDecreasedConnectivityonPopulationGeneticDiversity.”BiologicalConservation144(12):3143–48.http://dx.doi.org/10.1016/j.biocon.2011.09.010.

Janic,Milan,andTomislavJovanović.2012.“EstimatingSomeSocialandEnvironmentalEffectsfromRail/RoadSubstitutionintheTrans‐EuropeanTransportCorridors.”PROMET‐Traffic&Transportation24(4):335–41.http://www.fpz.unizg.hr/traffic/index.php/PROMTT/article/view/438.

Jayachandran,Seemaetal.2017.“CashforCarbon:ARandomizedTrialofPaymentsforEcosystemServicestoReduceDeforestation.”Science357(6348):267–73.

Johnson,Michael,RufusEdwards,andOmarMasera.2010.“ImprovedStoveProgramsNeedRobustMethodstoEstimateCarbonOffsets.”ClimaticChange102(3):641–49.

Joppa,LucasN.,andAlexanderPfaff.2009.“HighandFar:BiasesintheLocationofProtectedAreas.”PLoSONE4(12):1–6.

———.2011.“GlobalProtectedAreaImpacts.”ProceedingsoftheRoyalSocietyB:BiologicalSciences278(1712):1633–38.

Joppa,Lucas,andAlexanderPfaff.2010.“ReassessingtheForestImpactsofProtection:TheChallengeofNonrandomLocationandaCorrectiveMethod.”AnnalsoftheNewYorkAcademyofSciences1185:135–49.

Joshi,Prathibha,andKrisBeck.2017.“EnvironmentalKuznetsCurveforDeforestation:EvidenceUsingGMMEstimationforOECDandNon‐OECDRegions.”iForest10(1):196–203.

Kaczan,DavidJames.2016.“CanRoadsContributetoForestTransitions?”2016.https://davidkaczan.com/job‐market‐materials/.

Kaczan,David,AlexanderPfaff,LuzRodriguez,andElizabethShapiro‐Garza.2017.“IncreasingtheImpactofCollectiveIncentivesinPaymentsforEcosystemServices.”JournalofEnvironmentalEconomicsandManagement.

Kaul,I.,I.Grunberg,andM.Stern.1999.“DefiningGlobalPublicGoods.”InGlobalPublicGoods:InternationalCooperationinthe21stCentury,OxfordUniversityPress.

Kelly,Peter,andXuexiHuo.2013.“LandRetirementandNonfarmLaborMarketParticipation:AnAnalysisofChina’sSlopingLandConversionProgram.”WorldDevelopment48:156–69.

Keskinen,Marko,andMattiKummu.2010.ResearchGateImpactAssessmentintheMekong:ReviewofStrategicEnvironmentalAssessment(SEA)&CumulativeImpactAssessment(CIA).http://www.researchgate.net/publication/266415643.

Kiesecker,JosephM.,HollyCopeland,AmyPocewicz,andBruceMcKenney.2010.“DevelopmentbyDesign:BlendingLandscapelevelPlanningwiththeMitigationHierarchy.”FrontiersinEcologyandtheEnvironment8(5):261–66.

Koop,Gary,andLiseTole.1999.“IsThereanEnvironmentalKuznetsCurveforDeforestation?”JournalofDevelopmentEconomics58(1):231–44.

Laurance,SusanG.W.,PhilipC.Stouffer,andWilliamF.Laurance.2004.“EffectsofRoadClearingsonMovementPatternsofUnderstoryRainforestBirdsinCentralAmazonia.”ConservationBiology18(4):1099–1109.

Laurance,William.2017.“TheDarkLegacyofChina’sDriveforGlobalResources.”YalEnvironment360,YaleSchoolofForestry&EnvironmentalStudies.

Laurance,WilliamF.etal.2014.“AGlobalStrategyforRoadBuilding.”Nature513:229–33.

———.2015.“BadRoads,GoodRoads.”InHandbookofRoadEcology,eds.AndrewvanderRee,DanielJ.Smith,andClaraGrilo.WestSussex,U.K.:JohnWiley&Sons,Incorporated,10–15.

———.2015.“ReducingtheGlobalEnvironmentalImpactsofRapidInfrastructureExpansion.”CurrentBiology25(7):R259–62.http://dx.doi.org/10.1016/j.cub.2015.02.050.

Laurance,WilliamF.,andIreneBurguésArrea.2017.“RoadstoRichesorRuin?”Science358(6362):442–44.

Leblois,Antoine,OlivierDamette,andJulienWolfersberger.2017.“WhatHasDrivenDeforestationinDevelopingCountriesSincethe2000s?EvidencefromNewRemote‐SensingData.”WorldDevelopment92:82–102.

Ledec,George,SallyJohnsonCampos,andDianaReayReay.2016.BiodiversityOffsets:AUserGuide.Washington,DC.

Leung,Denise,YingzhenZhao,TaoHu,andAthenaBallesteros.2013.WorldResourcesInstituteIssueBriefEnvironmentalandSocialPoliciesinOverseasInvestments:ProgressandChallengesforChina.papers2://publication/livfe/id/144506.

Li,BinbinV.etal.2016.“RemotelySensedDataInformsRedListEvaluationsandConservationPrioritiesinSoutheastAsia.”PLoSONE11(8):1–18.

Li,M.,J.Wu,andX.Deng.2013.“IdentifyingDriversofLandUseChangeinChina:ASpatialMultinomialLogitModelAnalysis.”LandEconomics89(4):632–54.

Li,Manetal.2014.“ImpactsofRoadExpansiononDeforestationandBiologicalCarbonLossintheDemocraticRepublicofCongo.”EnvironmentalandResourceEconomics60(3):433–69.

Li,Tianweietal.2016.“StrategicEnvironmentalAssessmentPerformanceFactorsandTheirInteraction:AnEmpiricalStudyinChina.”EnvironmentalImpactAssessmentReview59:55–60.

Li,Xin,YinglingFan,andLanWu.2017.“CO2emissionsandExpansionofRailway,Road,Airlineandin‐LandWaterwayNetworksoverthe1985–2013PeriodinChina:ATimeSeriesAnalysis.”TransportationResearchPartD:TransportandEnvironment57(September):130–40.http://dx.doi.org/10.1016/j.trd.2017.09.008.

Litvaitis,JohnA.,andJeffreyP.Tash.2008.“AnApproachtowardUnderstandingWildlife‐VehicleCollisions.”EnvironmentalManagement42(4):688–97.

Liu,Hong,andGuanieLim.2018.“ThePoliticalEconomyofaRisingChinainSoutheastAsia:Malaysia’sResponsetotheBeltandRoadInitiative.”JournalofContemoraryChina:1–16.

Liu,Shiliangetal.2011.“EffectsofRoadNetworkonVegetationPatterninXishuangbanna,YunnanProvince,SouthwestChina.”TransportationResearchPartD:TransportandEnvironment16(8):591–94.http://dx.doi.org/10.1016/j.trd.2011.08.004.

Locatelli,Bruno,andRaffaeleVignola.2009.“ManagingWatershedServicesofTropicalForestsandPlantations:CanMeta‐AnalysesHelp?”ForestEcologyandManagement258(9):1864–70.

Mahmoud,MahmoudI.etal.2017.“AlternativeRoutesforaProposedNigerianSuperhighwaytoLimitDamagetoRareEcosystemsandWildlife.”TropicalConservationScience10.

Maliszewska,Maryla,andDominiquevanderMensbrugghe.Inpreparation.TheBeltandRoadInitiative:MacroandSectoralImpacts.Washington,DC.

Maurin,Cristelle,andPichamonYeophantong.2013.“GoingGlobalResponsibly?China’sStrategiesTowards‘Sustainable’OverseasInvestments.”PacificAffairs86(2):281–303.

McKenney,Bruce,andJessicaWilkinson.2015.AchievingConservationandDevelopment:10PrinciplesforApplyingtheMitigationHierarchy.https://www.nature.org/ourinitiatives/urgentissues/land‐conservation/smart‐development/applying‐the‐mitigation‐hierarchy.pdf.

Mcmillan,Madeleine.2018.“ThreatstoBigCatsinSoutheastAsia.”DukeUniversity.

Mena,CarlosF.,RichardE.Bilsborrow,andMichaelE.McClain.2006.“SocioeconomicDriversofDeforestationintheNorthernEcuadorianAmazon.”EnvironmentalManagement37(6):802–15.

Meyfroidt,Patrick,andEricFLambin.2009.“ForestTransitioninVietnamandDisplacementofDeforestationAbroad.”106(38).

Meyfroidt,Patrick,ThomasKRudel,andEricFLambin.2010.“ForestTransitions,Trade,andtheGlobalDisplacementofLandUse.”

Miteva,DanielaA.,BrianC.Murray,andSubhrenduK.Pattanayak.2015.“DoProtectedAreasReduceBlueCarbonEmissions?AQuasi‐ExperimentalEvaluationofMangrovesinIndonesia.”EcologicalEconomics119:127–35.

Monteiro,J.A.,andJ.P.Trachtman.InPress.“Environment‐RelatedProvisionsinPreferentialTradeAgreements.”InHandbookofDeepTradeAgreements,eds.A.Mattoo,NadiaRocha,andMicheleRuta.

Moran,Daniel,AliHasanbeigi,CeciliaSpringer,andGlobalEIntelligence.2018.TheCarbonLoopholein

ClimatePolicy:QuantifyingtheEmbodiedCarboninTradedProducts.

Morelli,Federicoetal.2014.“CanRoads,RailwaysandRelatedStructuresHavePositiveEffectsonBirds?‐AReview.”TransportationResearchPartD:TransportandEnvironment30:21–31.http://dx.doi.org/10.1016/j.trd.2014.05.006.

Myczko,Łukaszetal.2017.“EffectsofLocalRoadsandCarTrafficontheOccurrencePatternandForagingBehaviourofBats.”TransportationResearchPartD:TransportandEnvironment56(August):222–28.

NationalRail.2009.“ComparingEnvironmentalImpactofConventionalandHighSpeedRail.”NetworkRail,UK,August,Availableat:http://www.….http://scholar.google.com/scholar?hl=en&btnG=Search&q=intitle:Comparing+environmental+impact+of+conventional+and+high+speed+rail#0.

Nelson,Andrew,andKennethMChomitz.2011.“EffectivenessofStrictvs.MultipleUseProtectedAreasinReducingTropicalForestFires:AGlobalAnalysisUsingMatchingMethods.”PLoSONE6(8).

Nelson,G.C.,andD.Hellerstein.1997.“DoRoadsCauseDeforestation?UsingSatelliteImagesinEconometricAnalysisofLandUse.”AmericanJournalofa*griculturalEconomics79(1):80–88.

NewDevelopmentBank.2016.EnvironmentandSocialFrameworkV4.Shanghai,China.https://www.ndb.int/wp‐content/uploads/2017/02/ndb‐environment‐social‐framework‐20160330.pdf.

Nolte,C.,A.Agrawal,K.M.Silvius,andB.S.Soares‐Filho.2013.“GovernanceRegimeandLocationInfluenceAvoidedDeforestationSuccessofProtectedAreasintheBrazilianAmazon.”ProceedingsoftheNationalAcademyofSciences110(13):4956–61.

OECD.2006.AssessmentApplyingStrategicEnvironmentalAssessment:GoodPracticeGuidanceforDevelopmentCooperation.http://www.oecd‐ilibrary.org/development/applying‐strategic‐environmental‐assessment_9789264026582‐en.

Pagiola,Stefano,AgustinArcenas,andGunarsPlatais.2005.“CanPaymentsforEnvironmentalServicesHelpReducePoverty?AnExplorationoftheIssuesandtheEvidencetoDatefromLatinAmerica.”WorldDevelopment33(2SPEC.ISS.):237–53.

Panlasigui,Stephanieetal.2017.ImpactsofCertification,UncertifiedConcessionsandProtectedAreasonForestLossinCameroon,2000to2013.

Parris,Kirsten.2015.“EcologicalImpactsofRoadNoiseandOptionsforMitigation.”InHandbookofRoadEcology,eds.AndrewvanderRee,DanielJ.Smith,andClaraGrilo.WestSussex,U.K.:JohnWiley&Sons,Incorporated,151–58.

Paton,DaleG.,SimoneCiuti,MichaelQuinn,andMarkS.Boyce.2017.“HuntingExacerbatestheResponsetoHumanDisturbanceinLargeHerbivoresWhileMigratingthroughaRoadNetwork.”Ecosphere8(6).

Pfaff,Alexanderetal.2007.“RoadInvestments,SpatialSpillovers,andDeforestationintheBrazilianAmazon.”JournalofRegionalScience47(1):109–23.

———.2009.“ParkLocationAffectsForestProtection :LandCharacteristicsCauseDifferencesinPark

ImpactsacrossCostaRicaParkLocationAffectsForestProtection :LandCharacteristicsCauseDifferencesinParkImpactsacrossCostaRica∗.”TheB.E.JournalofEconomicAnalysis&Policy9(2).

———.2014.“Governance,LocationandAvoidedDeforestationfromProtectedAreas:GreaterRestrictionsCanHaveLowerImpact,DuetoDifferencesinLocation.”WorldDevelopment55:7–20.

———.2018.“Roads&SDGs,TradeoffsandSynergies:LearningfromBrazil’sAmazoninDistinguishingFrontiers.”Economics12:1–25.

Pfaff,Alexander,andJuanRobalino.2012.“ProtectingForests,Biodiversity,andtheClimate:PredictingPolicyImpacttoImprovePolicyChoice.”OxfordReviewofEconomicPolicy28(1):164–79.

———.2017.“SpilloversfromConservationPrograms.”AnnualReviewofResourceEconomics9(1):299–315.

Pfaff,Alexander,JuanRobalino,DiegoHerrera,andCatalinaSandoval.2015.“ProtectedAreas’ImpactsonBrazilianAmazonDeforestation:ExaminingConservation‐DevelopmentInteractionstoInformPlanning.”PLoSONE10(7):1–17.

Pfaff,Alexander,JuanRobalino,CatalinaSandoval,andDiegoHerrera.2015.“ProtectedAreaTypes,StrategiesandImpactsinBrazil’sAmazon:PublicProtectedAreaStrategiesDoNotYieldaConsistentRankingofProtectedAreaTypesbyImpact.”PhilisophicalTransactionsoftheRoyalSocietyB(370).

Pfaff,AlexanderSP.1999.“WhatDrivesDeforestationintheBrazilianAmazon?”43:26–43.

Pfaff,Alexander,FranciscoSantiago‐Ávila,andLucasJoppa.2017.“EvolvingProtected‐AreaImpactsinMexico:PoliticalShiftsasSuggestedbyImpactEvaluations.”Forests8(1):1–14.

Pfaff,Alexander,andRobertWalker.2010.“RegionalInterdependenceandForest‘Transitions’:SubstituteDeforestationLimitstheRelevanceofLocalReversals.”LandUsePolicy27(2):119–29.

Pilgrim,J.D.,andJonEkstrom.2014.TechnicalConditionsforPositiveOutcomesfromBiodiversityOffsets.”AnInputPaperfortheIUCNTechnicalStudyGrouponBiodiversityOffsets.:IUCN.Gland,Switzerland.

Popp,J.N.,andS.P.Boyle.2017.“RailwayEcology:UnderrepresentedinScience?”BasicandAppliedEcology19:84–93.http://dx.doi.org/10.1016/j.baae.2016.11.006.

Porras,I,MGreig‐Gran,NNeves,andInternationalInstituteforEnvironmentandDevelopment.2008.NaturalResourcesIssuesAllThatGlitters:AReviewofPaymentsforWatershedServicesinDevelopingCountries.http://www.iied.org/pubs/pdfs/13542IIED.pdf.

Potapov,Peteretal.2017.“TheLastFrontiersofWilderness:TrackingLossofIntactForestLandscapesfrom2000to2013.”ScienceAdvances3(1).

Rainey,HugoJ.etal.2015.“AReviewofCorporateGoalsofNoNetLossandNetPositiveImpactonBiodiversity.”Oryx49(2):232–38.

vanderRee,Rodney,JeffreyW.Gagnoy,andDanielJ.Smith.2015.“Fencing:AValuableToolforReducingWildlife‐VehicleCollisionsandFunnellingFaunatoCrossingStructures.”InHandbookofRoadEcology,eds.AndrewvanderRee,ClaraGrilo,andDanielJ.Smith.WestSussex,U.K.:JohnWiley&Sons,Incorporated.

VanDerRee,Rodney,DanielJ.Smith,andClaraGrilo.2015.HandbookofRoadEcologyHandbookofRoadEcology.

Reed,Tristan,andAlexandrTrubetskoy.2018.TheBeltandRoadInitiativeandtheValueofUrbanLand.Washington,DC.

Ren,Peng,ChangLiu,andLiwenZhang.2017.“China’sInvolvementinCoal‐FiredPowerProjectsalongtheBeltandRoad.GlobalEnvironmentalInstitute.”(May).

Ren,Peng,LiwenZhang,RognZhu,andJingweiZhang.2017.HowChina’sPolicyBanksCanSupportSustainableForeignInvestment.

Renaud,FabriceG.,Karen\Sudmeier‐Rieux,andMarisolEstrella,eds.2013.TheRoleofEcosystemsinDisasterRiskReduction.Tokyo:UnitedNationsUniversityPress.http://dx.doi.org/10.1016/j.ijdrr.2015.08.001.

Rico,Jimenaetal.2018.DeforestationImpactsofConservation&DevelopmentinthePeruvianAmazon:EvaluatingLand‐UseRestrictionsinCombinationswithRightstoExtraction*.

Robalino,JuanA.2007.“LandConservationPoliciesandIncomeDistribution:WhoBearstheBurdenofOurEnvironmentalEfforts?”EnvironmentandDevelopmentEconomics12:521–33.

Robalino,Juan,andAlexanderPfaff.2013.“EcopaymentsandDeforestationinCostaRica:ANationwideAnalysisofPSA’sInitialYears.”LandEconomics89(July2015):432–48.

Robalino,Juan,AlexanderPfaff,andLauraVillalobos.2015.“DeforestationSpilloversfromCostaRicanProtectedAreas.”DocumentosdetrabajoU.CostaRica.

———.2017.“HeterogeneousLocalSpilloversfromProtectedAreasinCostaRica.”JournaloftheAssociationofEnvironmentalandResourceEconomists4(3):795–820.

Ruan,Xiaofeng,FengQiu,andMilesDyck.2016.“TheEffectsofEnvironmentalandSocioeconomicFactorsonLand‐UseChanges:AStudyofAlberta,Canada.”EnvironmentalMonitoringandAssessment188(8).

Rudel,ThomasK.,LauraSchneider,andMariaUriarte.2010.“ForestTransitions:AnIntroduction.”LandUsePolicy27(2):95–97.

Sanchez‐Azofeifa,G.Arturo,AlexanderPfaff,JuanAndresRobalino,andJudsonP.Boomhower.2007.“CostaRica’sPaymentforEnvironmentalServicesProgram:Intention,Implementation,andImpact.”ConservationBiology21(5):1165–73.

Sausmikat,Noraetal.2017.SilkRoadBottom‐up:RegionalPerspectivesonthe‘BeltandRoadInitiative.’Cologne,Germany.

Sawyer,Halletal.2013.“AFrameworkforUnderstandingSemi‐PermeableBarrierEffectsonMigratoryUngulates.”JournalofAppliedEcology50(1):68–78.

Seiler,Andreas,andJ‐OHelldin.2006.“MortalityinWildlifeDuetoTransportation.”InTheEcologyofTransportation:ManagingMobilityfortheEnvironment,eds.J.DavenportandJ.L.Davenport.Netherlands:Springer.http://link.springer.com/10.1007/1‐4020‐4504‐2.

Selva,Nuria,AdamSwitalski,StefanKreft,andPierreL.Ibisch.2015.“WhyKeepAreasRoadfree?TheImportanceofRoadlessAreas.”InHandbookofRoadEcology,eds.AndrewvanderRee,DanielJ.

Smith,andClaraGrilo.WestSussex,U.K.:JohnWiley&Sons,Incorporated,16–26.www.wiley.com%5Cgo%5Cvanderree%5Croadecology.

Serebrisky,Thomasetal.2018.TechnicalNoteIDBGFrameworkforPlanning,Preparing,andFinancingSustainableInfrastructureProjects:IDBSustainableInfrastructurePlatform.

Shapiro‐Garza,Elizabeth.2013.“ContestingtheMarket‐BasedNatureofMexico’sNationalPaymentsforEcosystemServicesPrograms:FourSitesofArticulationandHybridization.”Geoforum46:5–15.http://dx.doi.org/10.1016/j.geoforum.2012.11.018.

Shinn,DavidH.2016.“TheEnvironmentalImpactofChina’sInvestmentinAfrica.”CornellInternationalLawJournal49:25–67.

Sidle,R.C.,M.Ghestem,andA.Stokes.2014.“EpicLandslideErosionfromMountainRoadsinYunnan,China‐ChallengesforSustainableDevelopment.”NaturalHazardsandEarthSystemSciences14(11):3093–3104.

Sills,ErinOetal.2015.“EstimatingtheImpactsofLocalPolicyInnovation:TheSyntheticControlMethodAppliedtoTropicalDeforestation.”PLoSONE10(7):1–15.

Sims,KatharineRE.2010.“ConservationandDevelopment:EvidencefromThaiProtectedAreas.”JournalofEnvironmentalEconomicsandManagement60(2):94–114.

Sims,KatharineRE,andJenniferM.Alix‐Garcia.2015.“ParksversusPES:EvaluatingDirectandIncentive‐BasedLandConservationinMexico.”JournalofEnvironmentalEconomicsandManagement.

Smith,DanielJ.,RodneyVanDerRee,andCarmeRosell.2015.“WildlifeCrossingStructures:AnEffectiveStrategytoRestoreorMaintainWildlifeConnectivityAcrossRoads.”InHandbookofRoadEcology,.

Smith,M.,D.deGroot,andG.Bergkamp.2006.Pay.EstablishingPaymentsforWatershedServices.Gland,Switzerland.https://www.iucn.org/sites/dev/files/import/downloads/pay.pdf.

Soanes,Kylie,andAndrewvanderRee.2015.“ReducingRoadImpactsonTreeDwellingAnimals.”InHandbookofRoadEcology,eds.AndrewvanderRee,DanielJ.Smith,andClaraGrilo.JohnWiley&Sons,Incorporated,334–40.

Song,X.P.,C.Huang,andJ.R.Townshend.2014.“AnIntegratedFrameworkforEvaluatingtheEffectsofDeforestationonEcosystemServices.”IOPConferenceSeries:EarthandEnvironmentalScience17(1).

Southworth,Jane,HariniNagendra,andLinCassidy.2011.“ForestTransitionPathwaysinAsia–studiesfromNepal,India,Thailand,andCambodia.”JournalofLandUseSciences4:1–15.

deSoyres,Françoisetal.2018.HowMuchWilltheBeltandRoadInitiativeReduceShipmentTimesandTradeCosts?Washington,DC.

Stibig,H.J.etal.2014.“ChangeinTropicalForestCoverofSoutheastAsiafrom1990to2010.”Biogeosciences11(2):247–58.

Stibig,H,FStolle,RDennis,andCFeldkötter.2007.JRCScientificandTechnicalReports,ForestCoverChangeinSoutheastAsia–theRegionalPattern.http://scholar.google.com/scholar?hl=en&btnG=Search&q=intitle:Forest+Cover+Change+in+Southeast+Asia+‐+The+Regional+Pattern+‐#0.

Symes,WilliamS.,MadhuRao,MichaelB.Mascia,andL.RomanCarrasco.2015.“WhyDoWeLoseProtectedAreas?FactorsInfluencingProtectedAreaDowngrading,DownsizingandDegazettementintheTropicsandSubtropics.”GlobalChangeBiology22(2).

TheWorldBank.2017.“EnvironmentalandSocialFramework.SecondDraft.”:141.

TheWorldBankGroup.2015.ANationalBiodiversityOffsetScheme:ARoadMapforLiberia’sMiningSector.Washington,DC.

TNC.2016.CapacityBuildingforMongolianMinistryofEnvironmentandGreenDevelopment(MEGDT)inRelationtoBiodiversityandConservationintheSouthernGobiDesert:FinalReport.

UNEnvironment.2018a.AssessingEnvironmentalImpacts‐AGlobalReviewofLegislation.Nairobi,Kenya.https://www.unep‐wcmc.org/system/dataset_file_fields/files/000/000/494/original/Environmental_Impacts_Legislation.pdf?1518007077%0Ahttp://web.unep.org/publications/.

———.2018b.14MainstreamingBiodiversityintheInfrastructureSector :FosteringSystem‐LevelApproaches.

UNOPS.2017.Evidence‐BasedInfrastructure:NISMOD‐International.

Vessali,KavehV.1996.“LandUseImpactsofRapidTransit:AReviewoftheEmpiricalLiterature.”BerkeleyPlanningJournal11(1).

Viana,Vm,andMcCenamo.2008.“RailroadsintheAmazon:AKeyStrategyforReducingDeforestation.”Carbon&ClimateL.Rev…1:290–97.http://heinonlinebackup.com/hol‐cgi‐bin/get_pdf.cgi?handle=hein.journals/cclr2&section=52.

Walston,Joeetal.2010.“BringingtheTigerBackfromtheBrink‐theSixPercentSolution.”PLoSBiology8(9):6–9.

Wang,Peishen,NingYang,andJuanDQuintero.2012.China :TheEnvironmentalChallengeofRailwayDevelopment.

Watkins,Graham.2014.No.IDB‐TNInter‐AmericanDevelopmentBankTECHNICALNOTEApproachestotheAssessmentandImplementationofSustainableInfrastructureProjectsinLatinAmericanandtheCaribbean.

Watson,JamesE.M.etal.2018.“TheExceptionalValueofIntactForestEcosystems.”NatureEcologyandEvolution2(4).

Weiss,MartinA.2017.CongressionalResearchServiceAsianInfrastructureInvestmentBank(AIIB).https://fas.org/sgp/crs/row/R44754.pdf.

Weller,Cameron.2015.“ConstructionofRoadsandWildlifeMitigationMeasures:PitfallsandOpportunities.”InHandbookofRoadEcology,eds.AndrewvanderRee,DanielJ.Smith,andClaraGrilo.WestSussex,U.K.:JohnWiley&Sons,Incorporated,60–64.

Westin,Jonas,andPerKågeson.2012.“CanHighSpeedRailOffsetit*EmbeddedEmissions?”TransportationResearchPartD:TransportandEnvironment17(1):1–7.http://dx.doi.org/10.1016/j.trd.2011.09.006.

Wingard,J.etal.2014.GuidelinesforAddressingtheImpactofLinearInfrastructureonLargeMigratory

MammalsinCentralAsia.ConventionontheConservationofMigratorySpeciesofWildAnimals(CMS)TechnicalReport.

WorldBank.2012.“StrategicEnvironmentalAssessmentintheWorldBank:LearningfromRecentExperiencesandChallenges.”:103.http://documents.worldbank.org/curated/en/729811468331017746/pdf/728950ESW0whit0200ENV0SEA0pub0final.pdf.

WorldBankTechnicalAssistanceProgram.2009.StrengtheningChina’sEnvironmentalProtectionAdministrativeSystem:AnalysisandRecommendations.Washington,DC.

Wu,Jingetal.2010.“StrategicEnvironmentalAssessmentImplementationinChina‐Five‐YearReviewandProspects.”EnvironmentalImpactAssessmentReview31(1):77–84.

Wunder,Sven.2007.“TheEfficiencyofPaymentsforEnvironmentalServicesinTropicalConservation.”ConservationBiology21(1):48–58.

WWF.2016.TheRoadAhead:ProtectingTigersfromAsia’sInfrastructureDevelopmentBoom.Gland,Switzerland.

Xu,Jintao,RanTao,ZhigangXu,andMichaelTBennett.2010.“China’sSlopingLandConversionProgram:DoesExpansionEqualSuccess?”Landeconomics86(2):219–44.

Xu,Zhigang,MTBennett,RanTao,andJintaoXu.2004.“China’sSlopingLandConversionProgrammeFourYearson:CurrentSituationandPendingIssues.”InternationalForestryReview6(3–4):317–26.

Yu,Heetal.2017.“EffectsoftheQinghai‐TibetRailwayontheLandscapeGeneticsoftheEndangeredPrzewalski’sGazelle(ProcapraPrzewalskii).”ScientificReports7(1):1–14.

Zeng,Zhenzhongetal.2018.“HighlandCroplandExpansionandForestLossinSoutheastAsiaintheTwenty‐FirstCentury.”NatureGeoscience11(8):556–62.http://dx.doi.org/10.1038/s41561‐018‐0166‐9.

Zhou,Lihuanetal.2018.MovingtheGreenBeltandRoadInitiative:FromWordstoActions.

Reducing Environmental Risks from Belt and Road Initiative …documents.worldbank.org/curated/en/700631548446492003/... · 2019-01-25 · Policy Research Working Paper 8718 Reducing - [PDF Document] (2024)

References