CASSAVA PROGRAM DISCUSSION PAPERS

Value Chain Analysis, Household Survey

and Agronomic Trial Results in Son La, Vietnam

Pham Thi Sen, Dominic Smith, Lava Yadav, Cu Thi Le Thuy, Le Viet Dung, Phan Huy Chuong and Jonathan Newby

Discussion Paper Number 1 May, 2018

School of Agriculture and Food Sciences

ISSN: 2209-2684

2

Value Chain Analysis, Household Survey and Agronomic Trial Results in Son La, Vietnam1

Pham Thi Sen2, Dominic Smith3, Lava Yadav3, Cu Thi Le Thuy4, Le Viet Dung2, Phan Huy Chuong2 and Jonathan Newby4

Country Information ProductionStatisticsCassavaproductioninVietnamincreasedfromaround350,000tonsin2001toalmost1.1milliontonsin2016.Duringthesameperiod,cultivatedarearoughlydoubled,from292,000hectaresin2001to570,000hectaresin2016.Thesignificantincreaseinyieldoverthesametimeperiodisduetotheintroductionofhigh-yieldingvarietiesintheearly2000s.

Figure1:CassavaCultivatedAreaandProductioninVietnam2001-2016

1ThisseriesCassavaProgramDiscussionPaperspresentsresultsoftheAustralianCentreforInternationalAgriculturalResearch(ACIAR)supportedprojectsASEM/2014/053DevelopingcassavaproductionandmarketingsystemstoenhancesmallholderlivelihoodsinCambodiaandLaoPDRandAGB/2012/078Developingvalue-chainlinkagestoimprovesmallholdercassavaproductionsystemsinVietnamandIndonesia2NorthernMountainousAgriculturalandForestryScienceInstitute,Vietnam3SchoolofAgricultureandFoodSciences,TheUniversityofQueensland,Brisbane,Qld4072,Australia.4InternationalCentreforTropicalAgriculture(CIAT),VientianeOffice,LaoPDR

0

2000

4000

6000

8000

10000

12000

0

200

400

600

800

1000

1200

2001 2005 2010 20152016

(‘000ha) (‘000tons)

Area Production

3

ThemaincassavaproducingregionsinVietnamareshowninTable1.NorthCentralCoasthasthehighesttotalproductionlevel,butthehighestyieldamongtheregionsisinSouthEast,atalmost26tonsperhectare.Table1:Cassavaplantedareaandproduction,byregion,Vietnam(2016)

Region PlantedArea(ha) Production(t)NorthernMountains

117,000 1,485,500

NorthCentralCoast 174,000 3,027,500

Centralhighlands

149,500 2,542,000

SouthEast

96,000 2,485,000

ProcessingStatisticsIn2014,therewere94cassavastarchprocessingfactories,producingatotalof2.2-2.3milliontons of starch per year. Tay Ninh province alone has 41 starch factories. There 6 ethanolprocessingplantsinthecountry,butonly3(TungLam,DaiViet,Nhiênliệusinhhọcmiềntrung)are currently operating. These 3 factories are operating at 50-60 percent of capacity, using130,000tonsofcassavachipsperyear.TradeStatisticsVietnamexportsbothfreshrootsandstarchwithatotalexportvalueofmorethanUSD1billionperyear.ThemainmarketforbothstarchandchipsisChina,accountingformorethan85percentof exports. The remainder is mostly destined for other markets in Asia, including Taiwan,Philippines,MalaysiaandIndonesia.ProjectActivityLocationsProjectactivitiesinVietnamarebeingundertakenintwoprovinces.AsshowninTable2,SonLaandDakLakbothhavesignificantareasofcassavaproduction,andthecombinedproductionofthe two provinces account for around 10 percent of Vietnam’s total cassava production.Differencesinagroclimaticconditions,ethnicgroupings,valuechainlinkagelevelsandthelevelof commercialization mean that the two provinces have sufficient contrasts to allow veryinterestingcomparisonstobemadebetweenvaluechainsindifferinglocations.

4

Figure2:ResearchLocations,Vietnam

5

Table2:Characteristicsofcassavaproductionbysite,Vietnam(2013)

Province Area ofcassava(ha)

Averagefreshyield(t/ha)

AnnualProduction(t)

Mainindustries

Number offactories

DakLak 25,720 18.4 473,248

Starch,EthanolDry chips(industrial)

5starch1ethanol(DakNong)

SonLa 31,216 11.5 359,485 StarchDry chips(industrial)

2starch

Province Information ProductionStatisticsSon La is one of themajor cassava producing provinces in Vietnam,with increased demandleadingtoproductionincreasesbetween2001and2011.Theincreaseinproductioncanmostlybeaccountedforbysignificantincreasesinplantedareaandsomesmallincreasesinyield(Figure3).Productionhasremainedrelativelystablesince2011,reflectingtheinfluenceofgovernmentpolicyaswellastheincreasingattractivenessofalternativecrops–especiallyArabicacoffee.

Figure3:ProductionandPlantedAreaofCassava2001-2015,SonLa

ProcessingStatisticsTheSonLaStarchProcessingFactory isownedby theFococevconglomerate5. Fococevarearecentlyprivatized(sinceMay2016)formerstate-ownedfoodproductinvestmentcompanywith

5http://fococev.com/vn/trang-chu/

0

5

10

15

20

25

30

35

0

50

100

150

200

250

300

350

400

2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015

Area(tho

usandha)

Prod

uctio

n(th

ousand

tons)

Production(thousandtons) Area(thousandha)

6

atotalof12cassavastarchprocessingfactoriesundertheirownership.Fococevpurchasedthefactoryin2012forVND10billionfromtheSonLaprovincialgovernmentandsincethattimehaveinvested a total ofmore than VND60 billion in upgrading equipment and installing a biogasdigester.PriortotheinvolvementofFococev,thefactoryhadnotbeenoperationalsince2003due to the limited investment capacity of the province, especially to meet environmentalstandards.In 2015-2016 the factory purchased around 40,000 tons of fresh root and produced around11,900tonsofstarch.Productionseasonlastsfrommid-OctobertoearlyAprilandthefactoryemploys90workersonafull-timebasisand25seasonalworkersoncontract.Figure4showsasummaryoftheproductflowsandvaluesforthe2015-2016seasonforthefactory.Thefactorypurchasesfreshrootfromtwomainsources.Atotalof30percentofthefreshrootsupplyispurchasedfromaround80localtradersfromMaiSon.Thesetradersaresmallscaleandgenerallydeliverproductin10tontrucks.Themajority(70percent)ofprocurementisfrom20tradersfromdifferentdistrictswithintheprovince.Thesetradersuse40tontruckstodeliverfreshrootstothefactory.TheaveragepricepaidforfreshrootoverthecourseoftheseasonwasVND1540/kgwithanaveragestarchcontent26.8percent.ThiswasthesameforbothsmallercollectorsfromMaiSonandforlargertradersfromotherdistrictsofSonLa.Thefactorybuysthefreshrootfromtraderswithoutusingwrittencontractsandpayseitherwithcashonthespotorthroughbanktransferifrequired.Evenwithoutformalwrittencontractarrangements,thelinkswiththecollectorsandtradersarerelativelystableandlonglasting.Whentheyfirsttookoverthefactory,FococevundertookacontractfarmingsystemwithgroupsoffarmersinMaiSonandotherdistricts.Underthecontractfarmingarrangement,theyprovidedplantingmaterialandadvancedfundsforfertilizer.However,theynolongerhaveanycontractfarmingsystemastheylostmoneyduetofarmerssideselling.AllofthesalesofstarchproducedbythefactoryarehandledthroughthemarketingdepartmentofFococev.Around90percentofproductionisexportedtoChinathroughMongCaibordergate,withtheremaining10percentdestinedforotherAsianexportdestinationsincludingKoreaandthePhilippinesaswellasdomesticnoodleandpaperproducingenterprises.

7

Figure4:Productflowsandvalues,starchfactory,MaiSon,SonLa

Value Chain Information ThecassavavaluechaininSonLahastwomainendproducts–cassavastarchandcassavachips.Regardlessoftheendproduct,almostalloftheprocessingoccurswithintheprovinceandalmostnofreshrootistransportedoutofSonLaforprocessinginotherprovinces.Thecassavachipvaluechainissignificantlylargerthanthatforstarch,accountingforalmost90percentofthetotalannualproductionoffreshroot.Theonelargescalestarchfactory intheprovince (Mai Son starch factory) consumedaround40,000tof fresh roots in 2015,with thebalanceofproduction(around320,000toffreshroots)beingutilizedtoproducedriedcassavachips.ThereissignificantcassavaprocessinginMaiSon,includingstarchprocessingattheSonLaStarchProcessingCompany,anddrychipprocessingbynumeroussmallandmediumscaleenterprisesatorneartheairport.InadditiontotheconcentratedprocessinginMaiSon,farmersinotherdistrictsalsoproducerelativelysmallamountsofdriedchips,usuallyeitherforlivestockfeed,becausetheywereunabletosellfreshroot,orbecausethepriceofcassavachipswasrelativelyfavourableatthetime.Thissmall-scalefarmerprocessingaccountsforanestimated5000tofthe125,000tonsofchipsproducedannuallyintheprovince.Smallscalecollectorsgenerallypurchasefreshrootsdirectlyfromfarmersat,orclosetotheirfieldsandtransportrootsin1t-3ttruckstocommunecentres.Smalltraderswith10ttrucksarebasedatcommunelevelandpurchasefreshrootsfromthesmall-scalecollectors.InMai Son district and in the south-east of Thuan Chau district, the communes are locatedrelativelyclosetothestarchfactoryandchipprocessors.SmalltradersfromMaiSonandThuan

8

Chautransportfreshrootspurchasedfromcollectorsdirectlytoprocessingintochipsorstarch.Inaddition,smalltradersfromMaiSonalsotraveltonearbydistrictsandpurchasefreshrootsfromcollectors.Inthenorth-westofThuanChaudistrictandinothermoreremotedistrictsofSonLa,thelongdistancesfromcommunestothestarchfactoryorchipprocessorsmakesitimpracticalformanysmalltraderstotransportfreshrootsfromcommunetoprocessor.Intheseareas,asignificantproportionofthefreshrootsisbroughtbysmalltraderstolargertradersbasedalongthemajorroadsoftheprovince.Thesetraderstransportthefreshrootsin40ttruckstochipprocessorsbasedinMaiSondistrictandtotheMaiSonstarchfactory.ThelargetradersalsotransportdrychipsdirectlytochiptraderslocatedinMaiSon.Starch produced by theMai Son starch factory is predominately for export, with around 90percent destined for China and 10 percent for Korea, Philippines, Taiwan and the domesticmarket.DrychipsaresoldtoanimalfeedproductioncompaniesinSonLaandHoaBinhaswellasforexport.

9

Figure5:SonLaValueChainMap

ThefarmgatepriceforfreshrootspaidbycollectorswasaroundVND1000/kg,collectorsthensoldtosmalltradersatcommunelevelforaroundVND1100/kg.SmalltradersinMaiSonandinthe South East of Thuan Chau who were able to sell directly to the starch factory or chipprocessors inMaiSonwereable togainasellingpriceofVND1400-1550/kg.Small traders inmoreremoteareasofSonLasoldtolargertradersatapriceofaroundVND1200/kg.ThelargetradersthensoldtothestarchfactoryortochipprocessorsatVND1400-1550/kg.

10

Table3:PricesoffreshcassavarootfordifferentvaluechainactorsinSonLa(2015)

ValueChainActor BuyingPrice SellingPriceFarmer VND1000/kgCollector VND1000/kg VND1100/kgSmallTrader VND1100/kg VND1200/kg(saletolargetrader)

VND1400-1550/kg(saletoprocessor)LargeTrader VND1200/kg VND1400-1550/kgStarchFactory VND1400-1550/kg ChipProcessor VND1400-1550/kg Despitethepresenceofnumerouslayersofactors(andinmanycasesthelongdistance)betweenfarmersandprocessors,thefarmgatepriceeveninremoteareasofSonLaisbetween65and70percentofthepricepaidbyprocessors.ThepricefordrychipspaidbycollectorstohouseholdprocessorswasaroundVND3200-3300/kg,collectorsthensoldtosmalltradersatcommunelevelorlargetradersatdistrictlevelforaroundVND3400/kg.Traderssoldtothechipprocessor/tradersinMaiSondistrictataroundVND3600-3700/kg.Sellingpriceforthechipprocessors/tradersinMaiSondistrictisaroundVND4000/kg.Farmgatepriceisaround80percentofthefactorygateprice.Table4:PricesofdrycassavachipsfordifferentvaluechainactorsinSonLa(2015)

ValueChainActor BuyingPrice SellingPriceFarmer/householdchipprocessor

VND3200-3300/kg

Collector VND3200-3300/kg VND3400/kgSmall/LargeTrader VND3400/kg VND3600-3700/kgChipProcessor/Trader VND3600-3700/kg VND4000/kgInformationFlowsDespitethemanyvaluechainactor layersand(inmanycases)significantgeographicdistancebetweenfarmersandprocessors,thetransmissionofpriceinformationthroughthenetworkoftradersandcollectorsworksrelativelyefficiently.Collectorsreportedbasingtheirpurchasepricedecisionson thepricesofferedby small traders at commune level. Small tradersbased theirpurchasingpricedecisionsonthepricesofferedbylargertradersandprocessors.Largetradersbasedtheirpurchasepricedecisionsonthepriceofferedbyprocessors.Valuechainactorsatalllevelsusedmobilephonesfrequentlytocontactbuyersforupdatedpriceinformation. In theabsenceof formal contractswithpriceguarantees,basingpurchasepricedecisionson frequentlyupdated informationon sellingprice canbe seenas aneffective riskminimizationstrategy.Informationontherelativepriceoffreshrootsandchipsandonthelevelsofdemandforeachproductalsoistransmittedrelativelywellthroughthevaluechain,withfarmersandcollectorsreportingthattheywereabletobasetheirdecisionsonwhethertomakechipsorsellfreshrootsoninformationreceivedfromsmalltradersatcommunelevel.

11

RelationshipsBeforebeingboughtbyFOCOCEV,theMaiSonstarchfactorypurchasedaproportionoftheirfresh root inputs from farmersand traders througha contract supply systemwhich includedspecifiedprice,quantityanddeliverytimes.Facedwithdecliningmarketpricesandincreasinginstability in themarket,FOCOCEVdoesnotuseacontractsystemforprocurementandnowbuysonaspotmarketbasis.MaiSonfactorypurchasesthemajorityofinputsfrommediumandlarge-scaletradersfrommanydistrictswithinSonLa.Thefactoryhaslong-termrelationswiththesetraders(andinthepastwasinvolvedincontractsupplyarrangementswithmanyofthem)butnowoperatesonaspotmarketbasiswithnoformalcontracts.Tradingrelationsbetweenlargetradersandsmalltraders,smalltradersandcollectors,andcollectorsandfarmershaveinmanycasesbeeninplaceforasignificantamountoftime.Althoughtherelationshavepersistedoverthelong-term,theydonotinvolveformalcontractsandpurchasesaremadeonaspotpricebasis.

Location of Project Activities within province ProjectactivitiesinSonLaarefocusingontwooftheelevendistrictswithintheprovince.ThuanChauandMaiSondistrictshavebeenselectedastheyrepresentthemaincassavaproducingdistrictsinSonLawithacombinedproductionofmorethan150,000tonsaccountingforalmosthalfoftheprovincialtotal.Table5:Cassavaarea,yieldandproductionbydistrictinSonLa(2015)

District CassavaArea(ha) Yield(t/ha) AnnualProduction(t)SonLaCity 217 18.0 3,900QuynhNhai 3,109 8.8 27,328ThuanChau 7,028 13.7 96,284MuongLa 2,712 10.2 27,662BacYen 2,836 8.7 24,554PhuYen 3,629 9.7 35,201MocChau 527 16.0 8,457YenChau 310 12.1 3,761MaiSơn 3,445 16.3 56,278SongMa 3,647 10.0 36,437SopCop 2,903 9.8 28,449VanHo 853 13.1 11,174SonLaProvince 31,216 11.5 359,485

Source:SonLaDepartmentofAgricultureandRuralDevelopmentThetwodistrictshavealsobeenselectedastheypresentinterestingcontrastsforresearchintermsofagro-climaticconditions,topographyandvaluechainlinks.

12

Figure6:Districtswithprojectactivities,SonLa

MaiSonisamidlanddistrictwithuplandsintheSouth-WestandNorth-Eastandabroadcentralvalley.ThedistrictbordersSonLaCitytotheNorthandhasrelativelygoodtransportationlinks,with theAH13highway running through the central valleyandanairport locatedatHat Lot.CassavaisgrowninboththeuplandsandthemidlandsoftheprovinceandMaiSonisthethirdlargestcassavagrowingdistrictinSonLa,behindThuanChauandSongMadistricts.Themorecommercializednatureofproduction is reflected in thedistricthavingoneof thehighestperhectareyields intheprovinceandthesecondlargestoverallproductionoffreshcassavaroot(seeTable5).ThereissignificantcassavaprocessinginMaiSon,includingstarchprocessingattheSonLaStarchProcessingCompany,anddrychipprocessingbynumeroussmallandmediumscaleenterprisesatorneartheairport.ThuanChauisapredominatelyuplanddistrictinthenorth-westernportionoftheprovince,withcassavabeinggrowninremotelocationsonrelativelysteepslopes.ThuanChauhasthelargestcassavagrowingareaandlargestfreshrootproductionofanydistrictinSonLa.Cassavaisgrownonmorethan7000hectaresinthedistrict,accountingforalmostaquarterofthetotalcassavagrowingareainSonLa.ThecentreofThuanChaudistrictisaround60kilometresfromtheSonLaStarchFactoryandthemajor chip producing area around the Na San airport. The rugged terrain means that thedistancesfromcassavafieldsthroughcommunestothedistrictcentreandthentothestarchfactoryandchippingareascanbeasmuchas200kilometres.ValueChainandHouseholdSurveyLocationsInMaiSon,farmerfocusgroupdiscussionswereheldintwouplandcommunes,ChiengChanintheNorth-EastofthedistrictandNaOtintheSouth-West.AlthoughbothcommunesareasimilardistancetoAH13,thetransportationlinksinNaOtarebetterthaninChiengChanasNationalRoad(QL)40whichlinkstheLaoPDRborderwithAH13passesthroughNaOt.ValueChainactorinterviewswereundertakenwithsmallscalecollectorsinNaOt,collectors/chippersinMuongBon(closetotheairport)andwiththeSonLaStarchFactory.

13

Figure7:ResearchLocations,MaiSonDistrict

InThuanChau, farmer focusgroupdiscussionswereheld in twouplandcommunes.BoMuoicommuneisaround35kilometresfromSonLaCityandPungTraislocatedaround36kilometresfrom Son La City. Pung Tra has significantly better transportation links, being located only 7kilometresfromthemainhighway(AH13).ValueChainactorinterviewswereundertakenwithsmallscaletradersinBoMuoiCommune,smallscalechipprocessorsandtradersinPungTracommuneandalarge-scaletraderinChiengPhaCommune.HouseholdsurveyswereundertakeninChiengChan,NaOt,PungTraandBoMuoicommunes.Ineachcommune,32householdsweresurveyed ineachof2villages. Ineachcommune, thechoiceofvillageswasmadeinordertohaveonemid-landvillageclosetothecommunecentreandonemorehighlandvillagefarfromthecommunecentre.Withineachvillagerespondentswereselectedrandomlyamongsthouseholdsproducingcassava.

14

Figure8:ResearchLocations,ThuanChauDistrict

Livelihood Information TimeoffirstcultivatingcassavaInadditiontothesignificantproportionofthefarmersthatstartedcultivatingcassavapriorto1990,threedistinct“peak”yearsforcommencingcassavaproductioncanbeseen.Thefirstpeak,in1997,sawfarmersinallcommunesstartingtoplantcassava.Anotherpeakoccurredin2007,withfarmersinallcommunescommencingcultivation.Thelastpeak,around2012sawfarmersinBoMuoi,ChiengChanandNaOtstartingtogrowcassava.Numbersoffarmersstartingtoplantcassavahasdeclinedeachyearfrom2012onwards.

Figure9:YearofFirstCassavaProduction,bycommune

15

ImportanceofCassavainoveralllivelihoodandincashincomeAlmostallsurveyedhouseholdshaveeitherlowlandoruplandricefields.TheproductionvalueofpaddyanduplandriceisanimportantcontributortolivelihoodsinPungTra,ChiengChanandBoMuoi(Figure10).MaizeisasignificantuplandcropinBoMuoiandChiengChan,whilecoffeeis cultivated by amajority of households in Na Ot and PungTra. Livestock – especially largelivestockisanimportantcontributortolivelihoodsinallcommunes.Off-farmincomesareimportantcontributortolivelihoods,particularlyinPungTraandChiengChan.MoredetailedinformationaboutannualincomesfromvarioussourcesisgiveninTable47.

Figure10:Sourceofincome,bycommune

Cassavacontributesanaverageof30percentofoverallhouseholdlivelihoodtohouseholdsinthelowestincomequartileandaprogressivelysmallerproportionoflivelihoodofhouseholdsinhigherincomequartiles,toalowoflessthan5percentoflivelihoodofhouseholdsinthehighestincomequartile(Table6andFigure11).

0.00

20.00

40.00

60.00

80.00

100.00

120.00

140.00

BoMuoi ChiengChan NaOt PungTra Total

MillionVN

D

totalcassavaincome paddyriceproductionvalue uplandriceproductionvalue

maizeincome Incomefromotherannualcrops coffee

incomeothertreecrops LivestockIncome Off-farmincome

16

Table6:AnnualIncomefromdifferentsources,byincomequartile(VND)

IncomeQuartile Q1 Q2 Q3 Q4 TotalTotal CassavaIncome 5,954,688 10,212,136 7,404,688 7,735,000 7,835,910

Non-CassavaCroppingIncome 7,444,531 19,074,769 38,762,375 74,344,063 34,844,833

Total LivestockIncome 404,688 5,936,923 17,683,344 36,604,688 15,121,533

Off-farmIncome 4,129,063 7,921,538 20,919,219 50,161,719 20,732,840Giventhesmalleraveragelandholdingsofhouseholdsinthelowest incomequartile, it isnotsurprisingthatoff-farmincomealsocontributesanaverageofmorethan15percentofoveralllivelihood to this quartile. From quartile 2 up to quartile 4, crops contribute a diminishingproportionoflivelihoodandlivestockandoff-farmincomeprovideaprogressivelyhighershare.

Figure11:SourceofLivelihood,byincomequartile

Figure12showsthesourcesofcashincomebyincomequartile.Thisisderivedbynotincludingthevalueofthestaplecrop(paddyoruplandrice)inthecalculationofgrossincome.Cassava’srelative importance to lower income households is shown clearly,with cassava providing onaverage nearly half of the cash incomeof households in the lowest incomequartile, and anaverageof30percentofincomeofhouseholdsinthe2ndincomequartile.Cassavaprovideslessthan5percentofcashincomeofhouseholdsinthehighestquartile,comparedwiththealmost60percentgainedfromlivestockandoff-farmincome.MoredetailedinformationispresentedinTable6.

0%10%20%30%40%50%60%70%80%90%

100%

Q1 Q2 Q3 Q4

IncomeQuartiles

TotalCassavaIncome Non-CassavaCroppingIncome TotalLivestockIncome Off-farmIncome

17

Figure12:CashIncomeSource,byincomequartile

LabourForceAverage household size was 5.5, with an average of 3.1 members having at least someinvolvementinagriculture,ofwhich2.4onaveragewereemployedfull-time.Table7:numberoffamilymembersbyemploymentstatus

AverageNumberofFamilyMembersEmploymentstatusinAgriculture Males Females TotalFulltime 1.2 1.2 2.4Never 1.2 1.2 2.3Parttime 0.2 0.2 0.3Rarely 0.2 0.2 0.4Total 2.7 2.7 5.5Useoflabourbygenderandhousehold/non-householdThereseemstobenospecificgenderrolesincassavaproduction,withmaleandfemaleperson-daysperyearforeachcassavaproductionrelatedtaskbeingrelativelyeven(Figure13).Thisisdifferenttothecaseofpaddyrice,wherethereissignificantgenderdisparitybetweendifferentproductiontasks6.

6Seeforexample,TruongThiNgocChi,NguyenThiKhoa,BuiThiThanhTam,andT.R.Paris(2004),GenderrolesinricefarmingsystemsintheMekongRiverDelta:anexploratorystudy,inG.L.DenningandVoTongXuan(eds).VietnamandIRRI:APartnershipinRiceResearch.ProceedingsofaconferenceheldinHanoi,Vietnam,4-7May1994.

0%10%20%30%40%50%60%70%80%90%

100%

Q1 Q2 Q3 Q4

IncomeQuartiles

TotalCassavaIncome Non-CassavaCroppingIncome TotalLivestockIncome Off-farmIncome

18

Figure13:HouseholdLabourPerson-Daysperhectare,byGender

The extremely steep slopes thatmuch of the cassava in Son La is grown onmean that landpreparationispredominatelycarriedoutmanuallyandthisisreflectedinthelargenumberofpersondaysdedicatedtofieldestablishmentandlandpreparation.Therelativelysmallquantitiesofchemicalfertilizerused(partlyafunctionofthesteepnessoftheslopes)isreflectedinverylownumbersofpersondaysoflabourforfertilizerapplication.Thechallengesoftransportingheavyloadsofherbicideupsteepslopesmeanthatalargenumberofpersondaysoflabourisutilisedfortwoseparateroundsofweeding.Thelargestsingleactivityabsorbinghouseholdlabourisharvesting,accountingformorethan50persondaysofhouseholdlabourperhectareperyear.DetailedlabourutilisationandcostfiguresareshowninTable48.Given the relatively low income levels of cassava farmers in Son La, it is not surprising thathouseholdsarethemainsourceoflabour,withhiredlabourandexchangelabouronlyaccountingfor a very small proportion of total labour used for production. Only harvesting andtransportation activities included any outside labour of note, and even this only totalled anaverageoflessthan20persondaysperhectareperyear(Figure14).

30 20 10 0 10 20 30 40

FieldEstablishment

LandPreparation

PlantingMaterialPreparation

PlantingStakes

FertiliserApplication1

PestandDiseaseControl

FirstWeeding

SecondWeeding

ThirdWeeding

Harvesting

Transporting

ChippingandDrying

Otherpost-harvest

Male Female

19

Figure14:LabourPerson-Daysperhectare,bySource

AccesstocreditAlmost58percentofhouseholdshadtakenat leastoneloaninthepast12months,withthemajorityofthoseonlyhavingoneloanandnohouseholdreportinghavingmorethan3loans.Quartile2reportedthehighestproportionofhouseholdswithloans(almost71percent),whileonly45.31percentofhouseholdsinthehighestincomequartiletookaloaninthelast12months(Table8).Theaverageamountofdebttakenoninthepast12monthswasaroundVND19.6million.Table8:Proportionofhouseholdshavingtakenloans

AccesstoCredit Q1 Q2 Q3 Q4 TotalPercent of households thatreceived a loan in the past 12months

57.81% 70.77% 57.81% 45.31% 57.98%

%householdswith1loan 48.44% 64.62% 54.69% 40.63% 52.14%

%householdswith2loans 7.81% 3.08% 3.13% 4.69% 4.67%

%householdswith3loans 1.56% 1.54% 0.00% 0.00% 0.78%

Average value of total loansreceived(VND)

13,828,125 19,030,769 24,343,750 21,359,375 19,638,132

Therewereproblemsaroundmanageabilityofdebt,withmorethan70percentofhouseholdsreporting at least some concern with their debt level and of those, more than 37 percentreportingthattheirdebtwas“unmanageable”or“veryunmanageable”(Table9).

60 50 40 30 20 10 0 10 20

FieldEstablishment

LandPreparation

PlantingMaterialPreparation

PlantingStakes

FertiliserApplication1

FirstWeeding

SecondWeeding

ThirdWeeding

Harvesting

Transporting

ChippingandDrying

Otherpost-harvest

Household Non-Household

20

Table9:Manageabilityofdebt

Howmanageableisthecurrentlevelofdebt Frequency PercentVeryunmanageable 67 31.31%Unmanageable 13 6.07%Someconcern 72 33.64%Manageable 56 26.17%Verymanageable 6 2.80%Total 214 100%The most common source of loans was the Bank for Social Policies, with the second mostfrequentloansourcebeingfromfamily/friends/relatives.Only7ofthemorethan140reportedloanswerefromshopkeepersortradersandnonewerereportedtohavecomefromthestarchfactoryorfromchipprocessors.Table10:LoanSources

SourceofLoan FrequencyBankforSocialPolicies 68Family/Friend/Relative 43Agribank 27Creditfund 6fertiliser,seedseller 6OtherBank 1farmersunion 1trader 1veteranscreditfund 1AccesstoinformationFarmers accessed information on agricultural production most frequently from friends andneighbours,fromtheirfamilyorfromTV.Tradersandcassavaprocessorswereonlynotedasasourceofinformationatotalof10times(Table11).

21

Table11:SourcesofInformationonagriculturalproduction

SourceofInformation FrequencyFriendsandneighboursinthevillage 224Family 198TV 119Friendsandneighboursoutsidethevillage 76Districtgovernmentextension 64FarmerGroup 45Other 20CassavaTraders 19Radio 8Internet 7Provincegovernmentextensionstaff 6CassavaProcessors 3Researchers 2Farmersaccessedinformationonagriculturalmarketsmostfrequentlyfromtraders,friendsandneighboursandfromtheirfamily.Cassavaprocessorswereonlynotedasasourceofinformationatotalof22times(Table11).Table12:SourcesofInformationonagriculturalmarkets

SourceofInformation FrequencyCassavaTraders 208FriendsandNeighboursinvillage 190Family 136FriendsandNeighboursoutsidethevillage 61TV 23CassavaProcessors 22FarmerGroup 14Other 8Districtgovernmentextension 3Internet 3Radio 2GroupmembershipAtotalof186households(72%ofallhouseholds)indicatedthattheyhadahouseholdmemberparticipating in a group or a mass organization. The most common organizations were theWomen’sUnionandtheFarmer’sUnion.

22

Table13:HouseholdMembershipofGroupsandMassOrganizations

NameofOrganization FrequencyWomen’sUnion 119Farmersunion 83HoChiMinhCommunistYouthUnion 56Veteran’sunion 26SeniorCitizensAssociation 18FatherlandFront 3communistparty 3cooperative 1publicsecurity 1OwnershipofassetsOverall,around90percentoffarmersownedmotorbikes.However,onlyabout67percentoffarmersinthelowestincomequartileownedmotorbikes.Around20percentoffarmersowned2or4wheeltractors,butasshowninTable29,thesewerenotgenerallyutilisedforcassavacultivation. More than 80 of farmers had a mobile phone and almost 35 percent had asmartphone.Table14:AssetOwnershipbyIncomeQuartile

Assets Q1 Q2 Q3 Q4 TotalTruck 0.00% 0.00% 0.00% 3.13% 0.78%car 0.00% 0.00% 0.00% 0.00% 0.00%motorbike 67.19% 96.92% 98.44% 98.44% 90.27%two wheeltractor

1.56% 6.15% 14.06% 28.13% 12.45%

four wheeltractor

3.13% 3.08% 9.38% 14.06% 7.39%

Waterpump 1.56% 16.92% 15.63% 32.81% 16.73%generator 4.69% 3.08% 0.00% 1.56% 2.33%mobilephone 71.88% 84.62% 89.06% 85.94% 82.88%smartphone 23.44% 27.69% 37.50% 50.00% 34.63%tv 85.94% 90.77% 98.44% 96.88% 93.00%dvdplayer 54.69% 63.08% 79.69% 67.19% 66.15%radio 12.50% 4.62% 7.81% 14.06% 9.73%refrigerator 3.13% 20.00% 46.88% 71.88% 35.41%

23

Agronomic Information Area,production,CurrentyieldsandtrendsAverage cassava production area per household was 0.57 hectares, varying between 0.31hectaresinPungTraand0.96hectaresinNaOt.Averageproductionwas7.9tons,givingayieldof15.5tonsperhectare(Table15).Theyieldperhectarewasrelativelyconstantacrossallfourcommuneswithalowof14.8tonsperhectareinPungTraandahighof16.6tonsperhectareinBoMuoi.Table15:HouseholdCassavaProductionCharacteristics,byCommune

BoMuoi

ChiengChan NaOt PungTra Total

Cassavaproduction2016(tons) 7.7 6.7 13.0 4.0 7.9CassavaHarvestArea2016(ha) 0.48 0.52 0.96 0.31 0.57CassavaYield2016(tons/ha) 16.6 15.3 15.3 14.8 15.5Theaveragehighestcassavayieldinthepast5yearswas19.3tonsperhectare,whiletheaveragelowestyieldwas13.9tonsperhectare.Table16:HighestandLowestProductioninlast5years,byCommune

BoMuoi

ChiengChan NaOt PungTra Total

HighestCassavaProductioninthelastfiveyears(tons) 13.6 8.6 16.8 8.5 11.9AreaUtilized forHighestCassavaYieldinthelastfiveyears(ha) 0.67 0.52 1.03 0.53 0.69

Highest Cassava Yield in the lastfiveyears(tons/ha) 21.3 19.4 18.9 17.7 19.3LowestCassavaProductioninthelastfiveyears(tons) 7.0 5.7 10.0 3.7 6.6Area Utilized for Lowest CassavaYieldinthelastfiveyears(ha) 0.46 0.51 0.95 0.31 0.56

Lowest Cassava Yield in the lastfiveyears(tons/ha) 16.0 14.1 12.4 12.9 13.9Cassava yields were declining either moderately or rapidly for a majority of farmers in allcommunes.Overall,only4.3percentoffarmersreportedthatyieldwasincreasing,whilealmost74percentreporteddecliningyields(Table17).

24

Table17:CassavaYieldTrends,byCommune

YieldTrend BoMuoi ChiengChan NaOt PungTra TotalIncreasing 3.1% 7.8% 4.7% 1.6% 4.3%Relativelyconstant 10.8% 18.8% 14.1% 7.8% 12.8%Decliningmoderately 40.0% 37.5% 39.1% 51.6% 42.0%Decliningrapidly 43.1% 26.6% 28.1% 29.7% 31.9%fluctuating,butnocleartrend 3.1% 9.4% 14.1% 9.4% 8.9%PlansforgrowingcassavainthefutureMorethan76percentoffarmersindicatedthattheyintendedtoplantcassavaintothefuture,withonly8.2percentnot intending togrowcassavaafter thecurrentseason.Theremainderwere unsure about their future plans for cassava production (Table 18). The proportion offarmersnotintendingtogrowcassavainthefuturewashighestinNaOtandlowestinChiengChan.Table18:FutureProductionIntention,byCommune

WillyougrowCassavaintheFuture?

BoMuoi ChiengChan NaOt PungTra Total

Yes 80.0% 71.9% 70.3% 82.8% 76.3%No 7.7% 3.1% 17.2% 4.7% 8.2%Unsure 12.3% 25.0% 12.5% 12.5% 15.6%Table19:FutureProductionIntention,byIncomeQuartile

WillyougrowCassavaintheFuture? Q1 Q2 Q3 Q4 TotalYes 76.6% 81.5% 73.4% 73.4% 76.3%No 12.5% 7.7% 7.8% 4.7% 8.2%Unsure 10.9% 10.8% 18.8% 21.9% 15.6%VarietiesFarmersreportedalargenumberofnamesofvarietiesthattheyplanted(Table20).Themajorityof theseare localnamesanddonotgiveany informationabout theactualvariety.Themostcommon“variety”reportedisCaoSan–thismeansHighYieldingCassava.TheonlyactualvarietynamereportedwasKM94,whichwasonlyreportedby1farmer.

25

Table20:VarietiesofCassavausedbyfarmers

VarietyName ProportionoftotalvarietiesCaoSan 55.3%LaTre 27.5%SanDen 12.1%SanXanh 1.9%SanTau 0.6%GiongNgheAn 0.6%KM94 0.3%GiongCaoBang 0.3%Sanlaunam 0.3%sanMocChau 0.3%SanMotThan 0.3%Sannguoikinh 0.3%SoilErosionProblemsandControlTechniquesAlmost 90 percent of farmers viewed soil erosion as a problem, with almost 60 percentconsideringitasseriousorveryserious.Around45percentoffarmerswereawareoferosioncontrolmeasuresbutonly7percenthadreceivedanytrainingonsoilconservationmeasuresinthepast.Encouragingly,almost90percentoffarmerswereinterestedinparticipatinginerosioncontrolmeasuretrialsontheirland.Table21:SoilErosionPerception,byCommune

Nameofcommune BoMuoi ChiengChan NaOt PungTra TotalSoil Erosion perceived as aproblem

92.3% 81.3% 93.8% 92.2% 89.9%

VerySeriousProblem 24.6% 18.8% 26.6% 17.2% 21.8%SeriousProblem 38.5% 29.7% 46.9% 32.8% 37.0%MediumProblem 21.5% 25.0% 17.2% 35.9% 24.9%SmallProblem 7.7% 7.8% 3.1% 6.3% 6.2%Areyouawareofanymeasuretoreducesoilerosion?

53.8% 23.4% 51.6% 51.6% 45.1%

Haveyouhadanytrainingonanysoilconservationmeasures?

4.6% 9.4% 10.9% 3.1% 7.0%

Are you interested in triallingconservation practices on yourland?

96.9% 84.4% 92.2% 82.8% 89.1%

Adoption of intercropping is very low, with only 7.4 percent of farmers ever having grownintercropswithcassavaandonly2.7percentoffarmerscurrentlygrowingintercrops.Morethan36precentoffarmersareinterestedintriallingintercrops,rangingfromon14percentoffarmersinChiengChantoalmost66percentoffarmersinNaOt.

26

Table22:AwarenessofIntercropping,byCommune

Nameofcommune BoMuoi ChiengChan NaOt PungTra TotalIntercropping Have you ever grown intercropswithyourcassava?

4.6% 4.7% 17.2% 3.1% 7.4%

Do you currently grow anyintercropswithyourcassava?

3.1% 1.6% 6.3% 0.0% 2.7%

Areyouinterestedintriallingnewintercrops?

29.2% 14.1% 65.6% 35.9% 36.2%

Fertiliseradoption,awarenessandcorrectapplicationOnly1.2percentoffarmersapplyorganicfertilizertotheircassava.Incontrast,theadoptionrateof inorganic fertilizer is relativelyhigh, at almost74percent.While adoption ishigh,only11percentoffarmersunderstandwhattheNPKvaluesontheirfertilizermean(Table23).Table23:FertiliserPractice,byCommune

BoMuoi

ChiengChan

NaOt PungTra Total

Do you apply organic fertiliser to yourcassava?

1.5% 0.0% 1.6% 1.6% 1.2%

Doyouapply inorganic fertiliser toyourcassava?

95.4% 64.1% 59.4% 76.6% 73.9%

DoyouunderstandwhattheNPKvaluesmeanonthefertiliseryouapply?

15.4% 7.8% 10.9% 10.9% 11.3%

Have you ever seen a fertiliser trial oncassava?

12.3% 14.1% 12.5% 7.8% 11.7%

Are you interested in visitinga fertiliserdemonstration trial to see the resultonproductionandreturns?

87.7% 93.8% 95.3% 87.5% 91.1%

Are you interested in conducting a trialonyourownland?

90.8% 79.7% 95.3% 84.4% 87.5%

Themostcommonfertilizerformulationusedbyfarmerswas5:10:3,aformulationwhichisnotoptimalforcassavaproduction.Almost30percentoffarmersdidnotknowwhatthefertilizerformulation that theyutilisedwas. Clearly there is anopportunity for fertilizer companies todevelopmoreappropriateformulationssuitableforcassavaproduction.

27

Table24:NPKFormulasUsedbyFarmers

Formula Proportionoffertilizerusers5:10:3 61%don’tknow 29%Inadditiontoapplicationofnon-optimalfertilizerformulations,theaveragequantityoffertilizerapplied per hectare is relatively low, at around560kgper hectare. The low level of fertilizerapplicationisnotsurprising,giventheextremelysteepslopesonwhichcassavaisplantedinSonLa.Table25:AverageNPKApplication(kgperhectare)duringplanting,byCommune

BoMuoi ChiengChan NaOt PungTra TotalNPK 839 246 543 460 560WeedingandHerbicidesMorethan95percentoffarmersindicatedthatweedswereaproblemandthatweedslimitedtheproductivityoftheircassavacrop.Thispatternwasrelativelyconstantacrossallcommunes.Table26:WeedImpactPerception,byCommune

BoMuoi ChiengChan NaOt PungTra TotalDo you think thatweeds limit theproductivity of yourcassavacrop?

largeproblem 55.4% 54.7% 45.3% 25.0% 45.1%mediumproblem 33.8% 31.3% 32.8% 45.3% 35.8%Smallproblem 9.2% 12.5% 17.2% 18.8% 14.4%No 1.5% 1.6% 4.7% 10.9% 4.7%Despitealmostallfarmersindicatingthatweedswereasignificantproblemimpactingoncassavaproduction,onlyaround26percentoffarmersusedherbicideontheircassavafields,rangingfrom17.2percentoffarmersinPungTra,to30.8percentoffarmersinBoMuoi(Table27).Thecost of herbicide and the steep slopes may explain the low proportion of farmers applyingherbicides.

28

Table27:HerbicidePractice,byCommune

BoMuoi ChiengChan NaOt PungTra TotalDo you apply anyherbicides?

30.8% 26.6% 31.3% 17.2% 26.5%

Have you receivedany training onherbicideuse?

4.6% 7.8% 9.4% 3.1% 6.2%

Do you useprotective clothingwhen applyingherbicide?

27.7% 21.9% 29.7% 14.1% 23.3%

Given the seriousness of the weed problem and the low level of herbicide use, it is hardlysurprising thatalmost100percentof farmerspracticemanualweedingofcassava fields.Themostcommonnumberoftimesofweedingoveraseasonistwo(Table28).Table28:ManualWeedingPractice,byCommune

BoMuoi ChiengChan NaOt PungTra Total

Do you conductmanualweeding?

100.0% 98.4% 100.0% 96.9% 98.8%

1weeding 16.9% 43.8% 54.7% 32.8% 37.0%2weedings 55.4% 37.5% 39.1% 53.1% 46.3%3weedings 26.2% 15.6% 6.3% 10.9% 14.8%4weedings 1.5% 1.6% 0.0% 0.0% 0.8%LandPreparationGiventhesteepnessofcassavafields,itisnotsurprisingthatonly2percentoffarmerscultivatecassavafieldsusing2or4wheeltractors.Onlyaround22percentoffarmersusebuffaloorcattleforploughing.Thedominantformoflandcultivationisusingmanualtools.ThisisthecaseinallcommunesexceptforChiengChan,wheretherelativelyflattertopographymeansthatamajorityoffarmerscanusecattleofbuffaloforlandcultivation(Table29).Table29:LandCultivationPractice,byCommune

BoMuoi ChiengChan NaOt PungTra TotalTractor 1.5% 3.1% 0.0% 0.0% 1.2%4wheeltractor 1.5% 0.0% 1.6% 0.0% 0.8%Buffaloorcattle 10.8% 71.9% 4.7% 1.6% 22.2%ManualTools 87.7% 31.3% 85.9% 96.9% 75.5%MakeRidges 1.5% 0.0% 3.1% 1.6% 1.6%

29

CassavaUtilisationMostfarmerssoldfreshcassava,accountingforatleast80percentoffarmersinallcommunesexceptPungTra.Atotalof37percentoffarmersalsousedcassavaforlivestockproduction.ThiswasparticularlycommoninPungTra,wheremorethan87percentoffarmersusedcassavaforfeedingtheirownlivestock.DriedchipproductionandsalesalsooccurredinBoMuoiandNaOt(Table30).Table30:CassavaUtilisation,byCommune

BoMuoi ChiengChan

NaOt PungTra Total

Eat 0.0% 0.0% 0.0% 3.1% 0.8%Useforownlivestock 15.4% 42.2% 3.1% 87.5% 37.0%CassavaLeaf 1.5% 1.6% 0.0% 0.0% 0.8%Sellfreshcassava 92.3% 84.4% 85.9% 57.8% 80.2%SellDriedcassava 10.8% 0.0% 21.9% 1.6% 8.6%RelationshipwithTradersOffarmersthatsoldcassavatofreshroottraders,around48percentdescribedtherelationshipasstrongorverystrong.Onlyabout20percentsaidthattherelationshipwasweakorveryweak.Althoughthenumberoffarmerssellingtodriedchiptraderswasmuchsmallerthanthenumbersellingtofreshroottraders,thepatternofrelationshipswasrelativelysimilar(Table31andTable32).Thiscouldbeexplainedbythefactthatatvillageandcommunelevelthesametradersareinvolvedinbothfreshrootanddrychiptrading.Table31:RelationshipwithFreshRootTraders,byIncomeQuartile

Freshroottraders Q1 Q2 Q3 Q4 Totalverystrong 13.7% 12.0% 17.4% 28.8% 18.0%Strong 29.4% 32.0% 40.4% 21.2% 30.7%moderate 35.3% 38.0% 19.2% 32.7% 31.2%weak 3.9% 8.1% 13.4% 11.6% 9.3%veryweak 17.7% 10.0% 9.6% 5.8% 10.8%Table32:RelationshipwithDryChipTraders,byIncomeQuartile

Drychiptraders Q1 Q2 Q3 Q4 Totalverystrong 14.4% 0.0% 0.0% 0.0% 4.7%Strong 14.4% 29.9% 100.0% 33.3% 31.4%moderate 56.8% 50.0% 0.0% 33.3% 45.3%weak 14.4% 0.0% 0.0% 0.0% 4.7%veryweak 0.0% 20.1% 0.0% 33.3% 14.0%

30

Trials 2016-2017 TrialLocationsThetrialswereconductedin4sites(4communes):

InThuậnChâudistrict:BóMườiandBúngTracommunesInMaiSơndistrict:NàỚtandChiềngChăncommunes

Table33:Briefinformationonthe4communes

MaiSơnDistrict ThuậnChâudistrict

ChiềngChăncommune

NàỚtcommune

BóMườicommune

PúngTracommune

Area(km2) 60.03 106.50 62.21 25.64

Population 6449 2976 8163 3138

Ethnicity Thái,H’Mông,Kinh

Thái,KhơMú,SinhMun,H’Mông

100%Thái 97,2%Thái 2,8%LaHa

No.ofvillage 19 17 18 14

Cassavaarea(ha) 244 541 100 750

Ineachofthesecommunes,asmentionedabove,over70%HHsgrowcassava,eachin0.3–0.9haoflandonaverage,mostlyonsteepslopes(upto650),andsoilerosionisperceivedasaseriousproblem.AccordingtotheHHsurvey,mostfarmersherehavenotbeenaccessedtoanytraininginsoilconservation.Regardingfertilizersuse,althoughhighrate(74%)ofHHsapplyNPK,onlyaround10%ofthemunderstandthevaluesofN,PandK.ThemostcommonfertilizerformulationusedisNPK5:10:3,althoughalargepartofHHs(30%)donotknowwhatthefertilizerformulationthattheyutilisedwas.Thefertilizers’levelusedisratherlow,onlyaround50kg/ha,ofwhichallisappliedonce,attheplantingtime.Regardingvarieties,almostallfarmersdonotknowwhatvarietiesactuallyareunderthelocallycallednames,exceptforKM94whichiscorrectlymentionedbyfewfarmers.Actually,themostcommon“variety”reported-CaoSan(meaningHighYieldingCassava)isalsoKM94whichisanewhighyieldingvarietypopularlygrownintheregion.Inthe3communesofBoMuoi,ChiengChanandNaOt,cassavaismainlyforprocessing(HHssellfreshrootsorasdriedtripstotradersorprocessingfactories),onlyasmallportionisusedforlivestockfeeding.However,inPungTracommune,mostHHs(87%)usecassavaasfeedsand3%HHsusealsoforfoods.Forbothfeedsandfoodslocallydevelopedvarieties(LaTre)withlessbitterrootsarepreferred.Over80%ofcassavafarmersarewillingtotakepartintrials,andalmost95%areinterestedinvisitingthetrials.

31

KeyActivitiesOnthebasisoftheabovementionedsituationtrialswereplannedtoaimat(i)introductionofsomenewhighyieldingvarietiessuitabletothelocalneedsandconditions,(ii)optimisationoffertilizers levels and application method and (iii) validation and dissemination of some soilconservationpractices.Selectionoffieldsanddesigningoftrials:Withparticipationof thedistricts’DARDandcommunes’extensionofficersand leaders, fieldtripswere conducted to the4 communes to find suitable fields to conduct the trials.Due tocomplicatedtopographyandsmall-scaleofplots,itwasnotpossibletofindsuitablefieldsinallthe 4 communes meeting the previously set-up requirements (rather flat and uniform landconditions for fertiliserandvariety trials,anduniformlysloping landsofabout20-25o forsoilconservation trial, each trial with 5 replicates, each trial plot area is at least 10 m x 5 m).Therefore,finally,thetrialsweredesignedandestablishedinthefieldsasbelow.

32

Table34:Locationsanddesignsoftrials

Trial Location Area (ha)

Slope (degree)

Design Farmer Ethnicity

1 Variety Quỳnh Lương village, Chiềng Chăn 0.15 5-10

CRB, 5 replicates

Lường Văn Yêu Thai

2 Variety Púng Mé Village, Púng Tra 0.15 45-50

CRB, 5 replicates

Lường Văn Ánh Thái

3 Fertilizer Quỳnh Lương village, Chiềng Chăn 0.15 5-10 Big PLots Lường Văn

Yêu Thai

4 Fertilizer Há Xét village, Nà Ớt 0.15 55-60 Big Plots Vì Văn Hom Thái

5 Fertilizer Long Sàn Village, Bó Mười 0.1 5-10 Big Plots Lò Văn

Phỏng Thái

6 Fertilizer Púng Mé Village, Púng Tra 0.15 5-10 Big Plots Lường Văn

Tưởng Thái

7 Soil management

Sài Lương village, Chiềng Chăn 0.25 35-40

CRB, 4 replicates

Lường Văn Nón Thái

8 Soil management

Há Xét village, Nà Ớt 0.2 55-60

CRB, 3 replicates

Vì Văn Hom Sinh mun

9 Soil management

Long Sàn Village, Bó Mười 0.15 40-45

CRB, 3 replicates

Lò Văn Yêu Thái

10

Soil management

Púng Mé Village, Púng Tra 0.25 45-50

CRB, 5 replicates

Quàng Văn Kiên Thái

Notestothehistoryofsoilmanagementandfertilisertrials(orderasinthe1stcolumnintheabovetable):(3):Rotationbetweenmaize,cassavaandsugarcane,notnecessaryfollowinganycycle.In2016maizewasplanted,fertilizersincluded1,6t/hamanure+600kg/haNPKforthebasal,200kg/haurea+200kg/haureafor2timestopdressing,yieldwas10t/hafreshcobs(4&8):plantedtocassavaformanyyears,withtheyieldofrootsreducedfromyeartoyear;forKM94theyieldwas20t/hain2013,15t/hain2014,andalmost7t/hain2016.Thefertilizerlevelin2016wasabout70kg/haNPK,onlyforthebasalapplication.(5):Duringthepast5yearsthelandwasplantedtomaizeorcassava,dependingonthechangeofmarketpriceofthese2crops,thelandownerchoosesoneofthemtogrow.In2016maizewascultivatedandafterharvestofmaizecowpeawascultivated.Fertiliserlevelappliedtomaizewas600kgNPK(5-10-3)/haforthebasal,150kgurea/hafortopdressing;themaizeyieldwas14t/hafreshcobs.(6):Thelandwasplantedtocassavaformanyyears.In2016theleveloffertilizersappliedwas120kg/haNPKforthebasaland20kg/haureaforthetopdressing.TheHHusedrootsforanimals,andthusharvestedsmallnumberofplantseverytimeandneverknewhowmuchtheyieldwas.(7):Before2014plantedtosugarcane,2014-2016cassava.In2016fertilizerswere200kgNPK,100kgureafortopdressing,yieldwas16t/ha.(9)Thelandwasplantedtomaizeorcassavas,dependingonthechangeofmarketpriceofthese2crops,thelandownerchooseoneofthemtogrow.In2016maizewascultivatedandafterharvestofmaizecowpeawascultivated.Thefertilizerslevelappliedtomaizewas300kg/haNPKforbasaland50kg/haureafortopdressing.Theyieldofmaizewas11t/hafreshcobs.(10):Thelandwasplantedtocassavaformanyyears.In2016theleveloffertilizersappliedwas200kg/haNPKforthebasaland30kg/haureaforthetopdressing.TheHHusedrootsforanimals,andthusharvestedsmallnumberofplantseverytimeandneverknewhowmuchtheyieldwas.

33

Varietytrialsinvolvedatotalof6varieties,includingKM94,Sa21-12,Rayong9,BK,13sa05andLaTre(apopularlocalvariety).Ofthose,KM94andLaTreareusedascontrols.Thetrialwasconductedinonecommuneineachdistrictonly(ChiềngChăncommune,MaiSơnDistrictandinPúngTracommune,ThuậnChâuDistrict).ThetrialdesignwasCRBwith5replications;areaofeachplotwas30m2.Participatoryevaluationwasconductedattheharvestwithinvolvementoffarmers, local officers, traders, cassava processing factory, extension staff and projectresearchers.

Fertilizertrialinvolved5treatments,usingthepopularhighyieldingvarietyintheregion-KM94and,wasconductedinallthe4communes.Theobjectiveistostudytheresponseofthecassavavarietytotheapplicationofvariouscombinationsoffertilizers(N,PandK)inordertofindthebest and most economic fertilizer rate to obtain and maintain high enough cassava yield.Participatory evaluation was conducted with involvement of farmers, local officers, traders,cassava processing factory, extension staff and project researchers. The trial was designedfollowingbigplots,withnoreplication:

InChiềngChănCommune:areaofeachplotwas175m2InNàỚtcommune:areaofeachplotwas120m2InPúngTracommune:areaofeachplotwas150m2InBóMườicommune:areaofeachplotwas120m2

T0(control):NonefertilizerT1:basalfertilizingwith300kg/haNPK(5:10:3),nonetopdressingT2:basalfertilizingwith600kg/haNPK(5:10:3),nonetopdressingT3:separateN,P,andKfertilizers;thetotalvolumewas(40N+10P+40K,equaling87kgUrea+142kgSuperphosphate+80kgKaliClorua)

- Basalfertilizing:AllofP,½ofK,½N- Topdressing:½N,½K(2monthsafterplanting)

T4:FDP(fertilizerdeepplacement,totalvolumewaslikethatofT3(40N+10P+40K)SoilmanagementtrialaimedtostudydifferentintercropsandsoilmanagementtechniquesinordertofindeffectiveoptionsintermofbotheconomyandsoilerosionmanagementforSonLa.Thetrialwasconductedinall4communesandinvolved6treatments:

T0(control):CassavaonlyT1:cassava+cowpeaT2:cassava+mungbeanT3:cassava+peanutT4:cassava+grasstripbyGuinea(Panicum.maximum)T5:cassava+contourlinesbyresiduesofcassavafromthelast

ThedesignwasCRB,with4replicatesinChiengChan(areaofeachplotwas11x6=66m2),3replicatesinNaOt(areaofeachplotwas11x5=55m2),5replicatesinPungTra(areaofeachplotwas11x5=55m2),and3replicatesinBoMuoi(areaofeachplotwas11x6=66m2).

Trialsestablishmentandmanagement:

34

Allthetrialswereestablishedduring3-7April,2017.Rightafterplantingofcassavaintercropsandgrassweresown.Farmersmanagedthetrialswiththeproject’sstafftechnicalsupport.

Organisationofharvestfielddays:

InDecember,oneharvestfielddaywasorganisedineachcommuneforparticipatoryevaluationofthetrials.Participantsincludedlocalleaders(communesandvillages),provincialanddistrictDARD, commune extension officers and representatives of communes ‘women unions andfarmers associations, as well as farmers, both directly involved in the trials and not directlyinvolvedinthetrials,localtradersandSonLacassavafactory(FOCOSEV).

Farmers and local officers expressed their interest in high yielding varieties. Three varieties,including13SA05,SA21-12andRayon9seemedtohavehigheryieldcomparedtothecontrol,KM94.Farmerswerealsointerestedinsoilconservationpracticesandbalancedfertiliserlevels.Likely,cowpeaasanintercropbroughtsignificantadditionalincomeandthereforemanyfarmerswishedtobesupportedtotestthislegumenextyearintheircassavafield.

Harvestoftrials:

Of the trials in Thuận Châu and the variety trials in both districts the harvest process werecompletedrightafterthefielddays.InMaiSonfarmershavenotyetharvestedtheircassava,andsothetrialswillbeharvestedlatter,perhapsbytheendofthismonth.

Parametersrecorded

Germination:germinationratesofallcrops

Cassavagrowthandbiomass:Randomlysampled10plantsineachplottomeasureandcalculatethemeanoftheirheightofstem,freshweightofnon-commercialaerialbiomass(stem+leaves),numberoftubers(onlyrootshavelengthequalorabove12cmanddiameterequalorabove2cm)andfreshweightofalltubers.

Intercropsgrowthandbiomass:Totalfreshbiomassofmungbeanandcowpeawereweightedafter2ndtimeofharvest,andthatofpeanutattheharvest.

Grassyield:Thesumofalltheharvestsintheyearmakestheyield

Yieldofcassava:Totalweightofalltubersharvestedofeachplot,andcalculatetheyieldperhectare.

Harvestindexforcassava:

HI= !"#$%'(")#"!"#$%'()#"*+"#$%($'#-*.#/0#$)

Starchcontent(%):CalculatedaccordingtothemethodofCassavaFactoryinSonLa,usingweightoffreshtubersintheair,andweightoffreshtubersinwater.Cost:allkindofworks(preparingland,planting,weeding,harvesting...);numberofworkingdaysforeachkindofjobs;kind,quantityandcostofanyinputs(seeds/chemicals,fertilisers...)

35

Soilcharacteristics:Beforestartingthetrials,soilsamplesweretaken(beforetheestablishmentof all the trials in 2017). For each trial block sampleswere taken in 5 locations along the 2diagonals,ineachlocation3samplesat3layers(0-10cm,10-20cm,20-30cm).Allsampleswereanalysedseparatelyinthefollowingparameters:pHH2O,pHKCl,OC,OlsenP,N%,CEC,EC,K+,Ca++,Na+andNitrat.Inthelastyear,afterharvestingofthetrialssampleswillbetakenandanalysedagain,butseparatelyforeachtrialplot.WhowasInvolvedNorthernMountainousAgricultureandForestryScienceInstitute(NOMAFSI)wasresponsibleforallthetrials.NOMAFSIalsoanalysethesoilsamples.TheRootCropResearchandDevelopmentCentre (RCRDC)participated in thevariety trial; inchargeofprovidingthestakesandrecordingdata,analysingdataandwritingthetechnicalreportregardingthistrial.Farmers (Table 34), the land owners, participated in all the activities (trials establishment,managementandharvest).DARD of Son La, and DARD of both Thuan Chau andMai Son districts participated as localsupervisors,providinginputsforthetrialsiteselection,implementingandmonitoring.PeoplecommitteesofChiengChan,NaOt,PungTraandBoMuoicommunesparticipatedasthecommunes’ focalpoints,playing the liaison role in connecting farmerswith theproject, andprovidinglogisticalsupportstotheimplementationofalltheactivities.SonLaCassavaFactory(FOCOCEV)andsomelocaltradersparticipatedinparticipatoryevaluationofthetrials.FOCOCEVespeciallywasalsoinvolvedinvarietiesevaluationandindeterminationofstarchcontentofcassavaroots.ResultsVarietytrialsAsseen(Table35andTable36),allthe4trialedvarietiesexpressedgoodgrowthandgavegoodyields,andthuscouldbeintroducedforusingintheproductionintheregion.Theyallhadstarchcontentandrootyieldequaltoorhigherthanthecurrentlocallypopularvarieties–KM94andLaTre.However,accordingtothevarietiesprofiles,Rayong9ishighlysusceptibletoredspiders.Formorecorrectconclusions,thetrialshouldberepeatedforonemoreyear.

36

Table35:YieldandyieldfactorsoftrailedvarietiesinPúngTra

Numberofroots/plant

Weightoffreshroots(kg/plant)

Freshrootyield(t/ha)

Starchcontent(%)

Starchyield(t/ha)

Rayong9 12.10 0.16 19.00b 29,2 5.5513Sa05 13.42 0.18 23.67c 30 7.10BK 10.77 0.18 18.98b 29 5.51Sa21-12 8.77 0.15 13.05a 30 3.92KM94(control1) 8.37 0.16 13.42a 30 4.03Látre(control2) 7.98 0.23 15.13a 30 4.54CV% 20.9 33.7 17.1 Table36:factorsofproductivity,yieldofvarietytrialinPúngTra–ThuậnChâu,2017

VarietiesNumberofroots/plant

Weightoffreshroots(kg/plant)

Freshrootyield(t/ha)

Starchcontent(%)

Starchyield(t/ha)

Rayong9 8.92 0.19 17.17bc 29.6 5.08

13Sa05 8.43 0.24 19.49d 28.1 5.48

BK 9.03 0.21 18.83cd 28.5 5.37

Sa21-12 6.47 0.24 15.14ab 30 4.54

KM94(control1) 7.39 0.22 16.47b 30 4.94

LaTre(control2) 7.08 0.19 13.72a 27.7 3.80

CV% 11.4 14.3 10.2

Soilmanagementtrials:OnlyinThuanChaudistrict(PungTraandBoMuoicommunes)cassavainthetrialwasharvested;inMaiSon(NaOtandChiengChancommunes)notyet.Thus,theresultsregardingcassavayieldandincomepresentedhereareonlyfromThuanChausites.Nevertheless,atthefielddayssomemain points were commonly reached by the participant, and this together with parameters‘recordsallowustomakethefollowingpoints(forcorrectdiscussionsandconclusionshowevertorepeattrialsforsomemoreyearsisrequired):Germinationandplantgrowth:Seedsofallcropsgerminatedwell.However,shortlyaftertheirgermination,during20April–10May,therewasalongandseriousdroughtperiodwhichcausedhighmortalityrateofmungbean(upto70%inMaiSonand46%inThuanChau)andGuineagrass(upto61%inMaiSonand37%inThuanChau).Cassavawasnotsignificantlyimpactedbythis drought spell while cowpea and peanut also were influences. The survival rates arepresented in Table 37. This consequently could significantly impact the yield and biomass ofintercropsandgrass.

37

Table37:Survivalrateoflegumeintercropsandgrass,1.5monthsaftersowing(15May,2017)

Cowpea Mungbean Peanut Guinea

ChiengChan 84.2 30.0 74.6 38.8

NaOt 72.2 40.0 77.8 53.3

PungTra 94.7 70.0 76.7 69.7

BoMuoi 80.0 53.8 68.3 42.5

Biomassandyieldoflegumeintercrops(Table40):Legumes,asintercrops,cowpeainparticular,broughtsignificantadditionalincomeforfarmers.TheyieldandincomeofintercroppedlegumesinChiengChanandBoMuoiwashigherthaninPungTraandNaOt,andthatofcowpeawashigherthanmungbeanandpeanut(Table38). Incaseofmungbeanthereasonforlowyieldcould be the high mortality rate, while regarding peanut there were evidences of insectsdamagingseedsandofpoorlydevelopedseed(verytinyseedsformed).Theyieldoflegumeswasnotproportionaltothesurvivalrate,andthiscouldbeduetothesoilconditions.InChiengChanandBoMuoithelandswerequiteflats,plantedtomaizeandlegumeslastyear,andwithgoodconditions.InNaOtandPungTra,landsweresteeplyslopingandplantedtocassavaformanyyearswithlowfertilizersrates.Asignificantvolumeofbiomasswasformedineachsite,andthiscouldbeagoodsourceoforganicfertilisers.Attheharvesthoweverallbiomassoflegumeshadalready been decayed, and thus intercropping with legumes does not help to build mulchmaterials.Table38:Yieldandbiomassofintercropsandgrassinthesoilmanagementtrials

Cowpea Mungbean Peanut GuineaDriedyield(kg/ha)

FreshBiomass(t/ha)

Driedseeds(kg/ha)

FreshBiomass(t/ha)

Driedseeds(kg/ha)

FreshBiomass(t/ha)

Totalharvest(t/ha)

Chiềngchăn 318.0b 1,37 93b 0,47 426c 1,41 0,90c

Nàớt 227.3a 0,83 52a 0,39 311b 1,59 0,82c

PúngTra 240.0a 1,39 80b 0,43 254a 0,87 0,25a

BóMười 325.0b 1,69 200c 1,04 432c 1,31 0,51b

Grassstripsprovidedfreshfeedsforcattleandalsopreventedsoilfrombeingwashedoffaway;asignificantamountofsoilwasdetainedabovethestrips.Neverthelessitrequiredadditionallabourforplantingandmanagingthegrass,andalsoadditionalcapitalinputforgrassseedsandfertilizers. On the other hand, fields are often steeply sloping and located far from farmers’houses,andthus,itwasdifficultforfarmerstotakegrasshometofeedtheircattle.Whennot

38

usinggrassforfeedingcattlefarmersneithersellgrass,andthereforetheywouldnotreallywanttospendinputsforgrassplantingandmanagement.Lastyear,inChiengChanthelandownerusedallthegrassharvestedforcattlefeedingandalsomanagedthegrasswell;inthissitethegrassyieldwashighest.InNaOt,thefieldwassteeplyslopingandfarfromthefarmer’shouse,andhenceonlysmallpartofgrassharvestedwastakenoutofthefieldforcattle.InBoMuoi,thelandownerhadsomegrassareaneartohishouseforcattle,andthusallgrassinthetrialwasleftinthefieldformulch.InPungTra,thelandownerusedallgrassforcattle,butinthissitepartofthegrasswasstolen.AlsoinPungTra,after15May2017grasscontinuedtodieddueandneededtobere-sownfewmoretimes.Thus,inPungTratheyieldofgrasswaslowest. Contoursofcassavaresidueshadno impactsonthecassavagrowthandyield,butcouldalsopreventasignificantamountofsoilfrombeingwashedoffaway.The yield of cassava (Table 39 and Table 40) in Pung Tra was not impacted by treatments.However,inBoMuoi,asignificantreductionincassavayieldwasobservedinallthetreatmentswithintercroppedlegumesandgrass;thereasoncouldbecompetitionfornutrition?Thestarchcontentwas30%forallthetreatmentsinallsites.Table39:YieldofcassavainsoilmanagementtrialinPúngTra

Numberofroots/plant

Freshroot(kg/plant)

Abovegroundbiomass(kg/plant)

Freshrootyield(t/ha)

Harvestindex(HI)

Control 7.36 1.48 0.97 14.37a 0.60C+cowpea 7.06 1.55 1.0 14.22a 0.61C+mungbean 7.18 1.45 1.01 15.30a 0.59C+peanut 7.74 1.51 1.00 15.02a 0.60C+grass 7.90 1.36 1.04 14.77a 0.56C+contourbyresidues 6.64 1.58 1.17 14.38a 0.57CV% 11.60 17.40 16.50 12.20

Table40:YieldofcassavainsoilmanagementtrialinBóMươi

Numberofroots/plant

Freshroot(kg/plant)

Abovegroundbiomass(kg/plant)

Freshrootyield(t/ha)

Harvestindex(HI)

Control 7.57 2.14 2.15 15.47c 0.50C+cowpea 6.97 1.90 2.07 12.93a 0.48C+mungbean 7.10 2.05 2.15 13.01ab 0.49C+peanut 6.73 1.83 2.03 13.76ab 0.47C+grass 7,20 2.16 1.99 14.06ab 0.52C+contourbyresidues 7.47 2.15 2.28 15.26bc 0.49CV% 9.40 4.70 10.10 6.10

39

Costand income (Table41andTable42): Treatmentswith leguminous intercropshadhighergrossandnetreturnduetoadditionalincomefromthelegumes;thehighestreturn,bothgrossandnet,hadthetreatmentwithcowpeaasanintercrop.Theothertwotreatments(withgrassstripsandcontourbyplantresidues)hadthesamegrossreturnasthecontrol,butthetreatmentwithgrassstriphadreducednetincomeduetoincreasedmaterialcostsrequiredforgrasstripplantingandmanagement(seedsandfertilisers)whilegrassdidnotbringanyadditionalincome.Asignificantreductioninthenetreturnperworkingdayandper1000vndspentwasobservedforallthetreatments,expecttheonewithcontourbyplantresidueshadthesameandtheonewithcowpeahadincreasednetreturnperworkingday.Thehighestreductioninnetreturnperworkingdayandper1000vndspentwasobservedforthetreatmentwithgrassstrip,andthereason could be high labour and inputs required for grass while no additional income wasobtained.Atthefielddays,farmersandlocalofficers,allexpressedtheirinterestincowpeaasintercropforitshighereconomicreturn.Table41:Incomeandinputcost,soilmanagementtrialinPungTra

Control C+cowpea

C+mungbean

C+peanut

C+grass

C+contour

Grossreturn(000vnd) 20,118 27,108 23,820 24,076 20,678 20,137Totalmaterialcost(000vnd) 3,342 4,678 4,778 5,078 7,510 3,342Totallabour(workingdays) 200 232 227 234 215 202Netreturn(000vnd) 16,776 22,430 19,042 18,998 13,168 16,796Netreturnperworkingday(000vnd) 83.88 96.68 83.89 81.19 61.25 83.15Netreturnper1000vndspent(000vnd) 5.02 4.80 3.99 3.74 1.75 5.03

Table42:Incomeandinputcost,soilmanagementtrialinBoMuoi

Control C+cowpea

C+mungbean

C+peanut

C+grass

C+contour

Grossreturn(000vnd) 21,658 25302 20614 22312 19,684 21,364

Totalmaterialcost(000vnd) 3,342 4,678 4,778 5,078 7,510 3,342

Totallabour(workingdays) 200 231 228 242 215 202

Netreturn(000vnd) 18,316 20,624 15,836 17,234 12,174 18,022Netreturnperworkingday(000vnd) 91.58 89.28 69.46 74.29 56.62 89.22

Netreturnper1000vndspent(000vnd) 5.48 4.41 3.31 3.39 1.62 5.39

Noteforbothtables41and42:priceofcowpeawas30.000vnd/kg,mungbean:30.000vnd/kg,peanut:12.000vnd/kg,cassava:1.400vnđ/kg,grass:0(grasswasnotsoldatall)

40

Fertilizertrials:CassavainMaiSonwasnotharvested.InBoMuoicommuneofThuanChau(Table43)theFDPtreatmentandthetreatmentwithseparatefertilizers(40N,10Pand40K)gavethelowestyield;one of the explanations for this could be that the top dressing aswell as gradually releasedfertilizerscaused‘over’vegetativegrowthandconsequentlyreducedyieldandreducedharvestindex(HI).Theother2treatments,onewith300kg/haNPKandonewith600kg/haNPK(appliedallasthebasal),hadthesameyieldand(HI)asthecontrol;fertilizersinthiscasehadnoimpact,andoneofthereasonscouldbethatthelandofthistrialinthissitewasrichinnutrientelements(flat land, in2016maize followedby cowpeawere cultivated, andahigh levelof fertilizers,including600kgNPKkg/haforthebasal,150kgurea/hafortopdressing,wasapplied). Inthiscase,fertilizerapplicationdidnotincreasebutreduceincomeandnetreturn(Table44).Table43:Cassavafreshrootyield,fertilisertrialinBoMuoi

Numberofrootsperplant

Freshrootsperplant(kg/plant)

Abovegroundbiomass(kg/plant)

Yield(t/ha)

Harvestindex(HI)

Starchcontent(%)

Nofertilizer 7.73 2.74 2.47 23.43b 0.53 30.0300kgNPK,onlybasal 7.23 2.27 2.47 23.22b 0.48 30.0

600kgNPK,onlybasal 7.70 2.48 2.42 22.2b 0.51 30.0

40N,10P,40K),basal&topdress 7.60 2.35 3.32 18.28a 0.41 30.0

FDP(40N,10P,40K) 8.13 2.28 3.11 17.1a 0.40 30.0

CV% 5.6 8.4 8.4 5.7 Table44:Costandreturn,fertilizertrialinBóMười

Nofertilizer 300kgNPK,

onlybasal

600kgNPK,onlybasal

40N,10P,40K),basal&topdressing

FDP(40N,10P,40K)

Grossreturn(000vnd) 32,802 32,508 31,080 25,592 23,940

Totalmaterialcost(000vnd) 1,455 2,715 3,975 3,406.8 4,545

Totallabour(workingdays) 210 210 208 207 205

Netreturn(000vnd) 31,347 29,793 27,105 22,185 19,395Netreturnperworkingday(000vnd) 149,3 141.8 130.3 107.2 94.6

Netreturnper1000vndspent(000vnd) 21.54 10.97 6.82 6.51 4.26

Note:thepriceofcassavafreshrootswas1,400vnd/kg

41

In Pung Tra (Table 45), all the treatments had increased yield compared to the control. ThehighestyieldhadthetreatmentwithseparateN,PandKfertilizersapplicationwith1topdressingtime,followedbythetreatmentswith300kg/haor600kg/haNPKappliedallasthebasal,andthen theDFP treatment. The benefits and income increased accordingly in these treatments(Table46).Here,inPungTra,thesoilconditionscouldberatherdifferenttothatinBoMuoi(thelandwasmoresteepandplantedtocassavaformanyyears,in2016theleveloffertilizerswasmuchlower,only120kg/haNPKforthebasal,and20kg/haureaforthetopdressing),andthustheimpactsoffertilizerswasalsodifferentcomparedtothatinBoMuoi.Inthissite,allthetestedfertilizerratesincreasedtheyieldofcassavaaswellasbothgrossandnetincomeandnetincomeperworkingday.However,thenetincomeper1000vndspentwasreduced,especiallyinFDPandhighNPKrate.Weneverthelesswillneedtorepeatthetrialforinsomemoreyears,andalsotolookatthesoilanalysisresultstohavebetterdiscussionsandconclusions.Table45:Cassavafreshrootyield,fertilisertrialinPungTra

Numberofrootsperplant

Freshrootsperplant(kg/plant)

Abovegroundbiomass(kg/plant)

Yield(t/ha)

Harvestindex(HI)

Starchcontent(%)

Nofertilizer 7.03 1.24 0.8 12.20a 0.61 29.1300kgNPK,onlybasal 8.47 1.69 0.95 16.28b 0.64 30.0600kgNPK,onlybasal 8.90 1.67 1.0 16.67bc 0.62 30.040N,10P,40K),basal&topdress 9.50 2.34 2.42 22.37d 0.49 30.0FDP(40N,10P,40K) 8.70 1.99 1.35 18.7c 0.60 30.0CV% 8.50 8.30 27.30 6.70

Table46:Incomeandinputcost,fertilizertrialinPungTra

Nofertilizer 300kgNPK,

onlybasal

600kgNPK,onlybasal

40N,10P,40K),basal&topdressing

FDP(40N,10P,40K)

Grossreturn(000vnd) 17,080 22,792 23,338 31,318 26,180

Totalmaterialcost(000vnd) 1,455 2,715 3,975 3,407 4,545

Totallabour(workingdays) 190 193 198 200 198Netreturn(000vnd) 15,625 20,077 19,363 27,911 21,635Netreturnperworkingday(000vnd) 82.2 104.0 97.8 139.5 109.3Netreturnper1000vndspent(000vnd) 10.73 7.39 4.87 8.19 4.76

42

ChallengesandconstraintsClimate: All the crops were sown/grown during 3 April – 7 April, and shortly aftergeminating/spoutingtherewasalongspellofdrought(during20April–10May)whichcausedhighrateofmortalityofplantlets.Viceversa,duringtheharvestingperiodoflegumesitwasrainyalot,almosteveryday,andthismaycausesomereductionintheyieldandqualityoflegumes.Grasswastore-sownfewtimesbecauseofthehighmortalityratecausedbydroughtandlatter-re-sownonediedduetothelackofsun(cassavaplantscoveredwellthegroundandlittlegrassseedlingswerecompletelyshaded).Lands:Mostofthecasesoffertilisersandsoilmanagementtrials,blockaretoosteeplysloping(45–60degrees)andlocatedfarfromfarmers’houses.Thiscauseddifficultiesforfarmerstovisitthefields,especiallyforharvestandtakegrasshometofeedtheircattle.Fertiliserapplication:Soilconditionsaredifferentbetweensites,butthesameratesweretestedinallsites.Cassavadensity:Thedensityof10000plants/ha(1mx1mdistance)seemstobelowandmightcauselowyieldofcassava.Increased labour requirement forplanting andmanaging intercrops andgrass strips, and thisseemstobeoneofthefactorshinderingtheadoptionofpractices.

Future plans and partnerships Opportunitiesandnewideasfor2018Thevaluechainsurveyandhouseholdsurveyresultspointtoanumberofclearconclusionsforfutureplansandpartnerships.Mechanizedlandpreparationcouldsavelabourcostsbutlandisgenerallytoosteep.Theweedproblemisseriousinallcommunesandalmostallfarmersspendalargeamountofpersondaysperyearonmanualweeding.Increasedherbicideusageforweedcontrolcouldreducelabourcostsbutitisdifficulttocarryliquidherbicideupsteepslopes.Higheryieldscouldpotentiallybegainedthroughmoreappropriate fertiliser formulationandmoderateincreasesinapplicationrates.Higheryieldingcassavavarietiesarelikelytohavethemost potential for increasing yields and improving farmer livelihoods and present the leastchallengesforadoption.Decliningyieldsandcassavaprices,andthefactthatcassavaonlyaccountsforasmallproportionoffarmerlivelihoodsmeansthatbenefitsofnewtechnologiesmustbeverysignificantinordertoencourageanywidespreadadoptionStrategyforengagementwithvaluechainstakeholdersforadoptionNewVarietiesThemainpriorityforinterventionexpressedbyfarmersinSonLawasnewvarietiesofcassava.Farmerprioritieswerevarietieswith;(i)higheryieldthanthecurrentvarietiesplantedinSonLa;

43

(ii)resistancetodisease,andinparticularresistancetoWitchesBroom;(iii)frosttolerance;(iv)earlyorlateharvestinginordertogainbettermarketprice;and(v)goodrootquality.Themainentrypoint/partnerforaninterventionintroducingimprovedvarietiesinthecassavavaluechaininSonLacouldbetheMaiSonStarchFactory.Thereisasignificantincentiveforthestarchfactorytopromotehigheryieldingvarietiesleadingtohigherrawmaterialsupplyinordertomore effectively use the increased capacity resulting from the investments in productivecapacitymadesincethetakeoverbyFOCOCEV.ThetechnologycharacteristicsofnewvarietiesandthecommunitycharacteristicsinSonLameanthatthepotentialpeakadoptionlevelofnewvarietiesbyfarmersinSonLaisrelativelyhigh.While FOCOCEV have a strong incentive to support the dissemination and adoption of newvarieties,theylackstronglong-termlinksthoughthevaluechain.Largertraderssupplyingthefactory(includingNguyenThiHainThuanChau)havestrongupstreamlinksinthevaluechainbacktofarmers,buthavelittleincentivetopromotehigheryieldingvarieties.Inordertofacilitateengagementoftradersandwidespreaddisseminationofvarieties, largerscale traders and associated small traders at commune level need to be incentivised toparticipate. Incentives could include subsidising the sale of stakes to larger traders andsupporting large traders and commune level traders tomultiply plantingmaterial for sale tofarmers. Initial technical support could come from the project, but financial support forsubsidisingplantingmaterialshouldcomefromthefactory.Moreeffectivefertilisertreatments.Themainentrypoint/partnerforaninterventionintroducingmoreeffectivefertilisertreatmentsinthecassavavaluechaininSonLacouldbefertiliserproductioncompaniesactiveinSonLaandtheirassociatednetworksofagriculturalinputsupplyshops.ThereisasignificantprofitincentiveforfertilisercompaniestopromotethewidespreaddisseminationandadoptionoffertiliserforcassavaproductionaslessthanhalfofcassavaproducersinSonLausefertiliserandfarmerswhodo use fertiliser are using relatively small quantities and the formulations used are eitherinappropriateorarenotknownbyfarmers.Thelinkagesoffertilisercompaniestofarmersarestrongduetotheirdistributionnetworksthroughinputsupplyshopsdowntothelocallevel.Whiletheengagementanddisseminationincentivesarehigh,thepotentiallevelofadoptionoffertiliseriscurrentlylowduetothenon-availabilityofappropriateformulationsoffertiliserforcassavaproduction.Oneof thekey investments in facilitationof theadoptionof fertiliser forcassavaproductionwillbeworking togetherwith fertilisercompanies todevelopappropriateformulationsbasedontrialresults.

44

Detailed Tables Table47:AverageHouseholdIncomesfromvariousSources(VND/Year),byCommune

Average HouseholdIncomes from variousSources(VND/Year)

Nameofcommune BoMuoi ChiengChan NaOt PungTra TotalFreshrootincome 6,472,462 6,885,469 13,715,156 1,896,875 7,239,494DryChipincome 1,406,154 - 3,216,856 - 1,156,727TotalCassavaIncome 7,878,615 6,885,469 16,932,013 1,896,875 8,396,221Paddy rice productionvalue

16,030,769 14,503,906 2,092,188 17,245,938 12,482,062

upland rice productionvalue

18,462 - 2,973,438 337,500 829,183

IncomefromMaize 10,753,692 61,650,781 1,180,813 447,656 18,478,062Income from all otherannualcrops

923 93,750 1,056,250 - 286,615

Incomefromcoffee 140,769 171,875 6,024,375 4,075,000 2,593,424Income from all othertreecrops

210,769 490,625 - - 175,486

CroppingIncome 35,034,000 83,796,406 30,259,075 24,002,969 43,241,054Non-Cassava CroppingIncome

27,155,385 76,910,938 13,327,063 22,106,094 34,844,833

CattleIncome 4,823,077 3,339,063 1,265,625 9,589,063 4,754,475BuffaloIncome 14,000,001 8,563,031 - 5,296,875 6,992,350GoatIncome 838,462 1,981,250 421,875 1,540,625 1,194,163PigIncome 830,769 1,365,625 175,000 2,912,188 1,318,988ChickenIncome 512,615 226,563 462,500 1,490,625 672,451DuckIncome 85,385 117,188 - 49,219 63,035OtherLivestockIncome 23,077 40,625 20,313 406,250 122,179fishIncome - - 15,873 - 4,000TotalLivestockIncome 21,113,385 15,633,344 2,360,938 21,284,844 15,121,533On-farmIncome 56,147,385 99,429,750 32,620,013 45,287,813 58,362,587Off-farmWages - 562,500 1,370,313 1,275,000 798,833Irregular non-farmincome

3,392,308 14,394,531 3,428,125 14,015,625 8,786,576

SalaryIncome 3,428,615 11,988,125 3,428,438 4,840,625 5,911,751NTFPincome - 467,188 679,688 234,375 343,969FishingIncome - 395,161 - - 96,078OtherIncome 2,005,538 6,357,969 943,750 9,687,500 4,738,016Off-farmIncome 8,826,462 33,590,625 8,480,000 28,778,125 19,875,642TotalIncome 64,973,847 133,020,375 41,100,013 74,065,938 78,238,229

45

Table48:AverageHouseholdIncomesfromvariousSources(VND/Year),byIncomeQuartile

AverageHouseholdIncomesfromvariousSources(VND/year)

IncomeQuartile

Q1 Q2 Q3 Q4 Total

Freshrootincome 5,112,500 7,536,462 7,123,438 6,930,938 6,679,183

DryChipincome 842,188 2,675,674 281,250 804,063 1,156,727

TotalCassavaIncome 5,954,688 10,212,136 7,404,688 7,735,000 7,835,910

Paddyriceproductionvalue

4,103,125 8,052,308 14,527,344 23,314,688 12,482,062

uplandriceproductionvalue

1,194,531 1,330,000 573,438 210,938 829,183

IncomefromMaize 895,000 7,226,308 20,637,375 45,329,375 18,478,062

Incomefromallotherannualcrops

938 0 0 1,150,000 286,615

Incomefromcoffee 1,240,000 2,372,308 2,975,781 3,789,063 2,593,424

Incomefromallothertreecrops

10,938 93,846 48,438 550,000 175,486

CroppingIncome 13,399,219 29,286,905 46,167,063 82,079,063 42,680,742

Non-CassavaCroppingIncome

7,444,531 19,074,769 38,762,375 74,344,063 34,844,833

CattleIncome 78,125 3,269,231 4,979,688 10,714,063 4,754,475BuffaloIncome 0 1,207,692 8,266,157 18,585,938 6,992,350GoatIncome 62,500 536,923 1,539,063 2,648,438 1,194,163PigIncome 168,750 332,769 2,139,063 2,650,781 1,318,988ChickenIncome 95,313 552,615 328,125 1,715,625 672,451

DuckIncome 0 17,692 56,250 178,906 63,035OtherLivestockIncome

0 20,000 375,000 95,313 122,179

46

fishIncome 0 0 0 15,625 3,891TotalLivestockIncome 404,688 5,936,923 17,683,344 36,604,688 15,121,533

On-farmIncome 13,803,906 35,223,828 63,850,407 118,683,750 57,802,276

Off-farmWages 1,381,250 863,077 703,125 246,875 798,833

Irregularnon-farmincome 1,557,813 3,449,231 10,667,969 19,554,688 8,786,576

SalaryIncome 305,625 1,238,462 4,656,563 17,519,375 5,911,751NTFPincome 457,813 353,846 275,000 289,063 343,969FishingIncome 0 0 70,313 500,000 142,023OtherIncome 426,563 2,016,923 4,546,250 12,051,719 4,749,689Off-farmIncome 4,129,063 7,921,538 20,919,219 50,161,719 20,732,840

TotalIncome 17,932,969 43,145,366 84,769,626 168,845,469 78,535,116

47

Table48:LabourCostsforVariousProductionActivities(VND/Year),byCommune

Nameofcommune BoMuoi ChiengChan NaOt PungTra TotalField EstablishmentHouseholdLabour

2,908,718 1,761,692 2,364,153 3,421,763 2,619,560

Field EstablishmentOutsideLabour

- 51,829 77,702 18,750 36,708

Land PreparationHouseholdLabour

5,092,234 1,107,647 3,623,502 8,824,368 4,681,636

Land PreparationOutsideLabour

176,923 21,164 194,104 203,646 149,393

Planting MaterialPreparation HouseholdLabour

932,967 531,387 654,577 1,415,885 886,177

Planting MaterialPreparation OutsideLabour

- 33,333 14,286 - 11,765

Planting StakesHouseholdLabour

2,850,037 1,596,170 1,976,979 5,169,866 2,906,793

Planting Stakes OutsideLabour

153,846 506,525 532,729 642,150 457,140

Fertiliser HouseholdLabour

24,359 105,643 213,379 35,156 93,850

FertiliserOutsideLabour - 4,233 - - 1,046First WeedingHouseholdLabour

4,544,982 1,858,711 2,592,297 7,185,826 4,061,687

First Weeding OutsideLabour

53,846 93,122 307,937 375,000 206,928

Second WeedingHouseholdLabour

3,249,524 727,786 924,244 5,759,115 2,681,883

Second WeedingOutsideLabour

30,769 8,466 13,605 341,146 98,917

Third WeedingHouseholdLabour

520,989 145,522 60,317 183,333 229,669

Third Weeding OutsideLabour

- - - - -

Harvesting HouseholdLabour

4,869,341 2,045,354 4,622,628 10,488,839 5,521,081

Harvesting OutsideLabour

1,615,128 1,945,307 895,314 707,515 1,291,072

TransportingHouseholdLabour

1,293,681 583,951 1,480,860 4,437,649 1,953,655

Transporting OutsideLabour

876,026 480,755 1,023,413 197,917 644,592

Chipping and DryingHouseholdLabour

759,744 510,582 565,029 3,214,621 1,266,206

48

Chipping and DryingOutsideLabour

73,846 - 14,966 156,250 61,737

Other post-harvestHouseholdLabour

191,685 130,915 71,605 256,696 163,321

Other post-harvestOutsideLabour

- - - 156,250 39,216

TotalLabour 30,218,645 14,250,093 22,223,626 53,191,741 30,064,030HouseholdLabour 27,238,260 11,105,359 19,149,570 50,393,118 27,065,517OutsideLabour 2,980,385 3,144,734 3,074,055 2,798,624 2,998,512

49

Table49:Detailedcostandlabour,soilmanagementtrialinPungTracommune(for1ha)

Unitprice(Vnđ) Cassava Cassava+

cowpeaCassava+mungbean

Cassava+Peanut

Cassava+grasstrip

Cassava+cotourline

Fertilizersandpesticides

NPK(kg) 4,200 0 80 80 80 40 0Ure(kg) 72,000 87 87 87 87 87 87Kaliclorua(kg) 10,000 80 80 80 80 80 80Suppelân(kg) 3,700 142 142 142 142 142 142Herbicide(liter) 65,000 6 6 6 6 6 6Pesticide (spraytimes) 600.000 1 1 1

Seeds Cassava(stems) 100 10,000 10,000 10,000 10000 10,000 10,000peanut(kg) 40,000 0 0 0 20 0 0cowpea(kg) 40,000 0 10 0 0 0 0mungbean(kg) 50,000 0 0 10 0 0 0Grass Panicum.sp(kg) 4000 0 0 0 0 1,000 0

Labour(workingdays)

Prepareland 60 60 60 60 60 60Planting 20 20 20 20 20 20Topdressing 2 2 2 2 2 2weeding1st 25 30 30 30 25 25weeding2nd 20 25 25 25 20 20Harvestcassava 73 73 73 73 73 73Intercroping Planting 0 10 10 10 0Praypesticide 4 4 4 Harvestintercrop 0 8 3 10 0Grasstrip Planting 10 Harvest 5 ControlCotourline 2Yieldandprice Cassava(tons) 1,400,000 14.37 14.22 15.3 15.02 14.77 14.384peanut(kg) 20,000 0 0 0 254 0 0Cowpea(kg) 30,000 0 240 0 0 0 0mungbean(kg) 30,000 0 0 80 0 0 0Grasstrip(kg) livestock 0 0 0 0 248 0

50

Table50:Detailedcostandlabourrequirement,FertilizertrialinPúngTra(for1ha)

Price(Vnđ) T0 T1 T2 T3 T4Fertilizersandhericides NPK(kg) 4,200 0 300 600 Ure(kg) 72,000 0 87 Kaliclorua(kg) 10,000 0 80 Suppelân(kg) 3,700 0 142 FDP(kg) 10.000 0 309Herbicide(liter) 65,000 7 7 7 7 7Seeds Cassava(stems) 100 10,000 10,000 10,000 10000 10,000Labour(workingdays) Prepareland 60 60 60 60 60Planting 20 20 20 20 20Topdressing 0 0 0 2 0weeding1st 25 25 25 25 25weeding2nd 20 20 20 20 20Harvestcassava 65 68 73 73 73Yieldandprice Cassava(tons) 1,400,000 12.2 16.28 16.67 22.37 18.7

51

Table51:Detailedcostandlaboutrrequirement,sSoilmanagementinBóMười(for1ha)

Price(Vnđ) cassava Cassava

+cowpeaCassava+mungbean

Cassava+Peanut

Cassava+grasstrip

Cassava+cotourline

Fertilizersandpesticides

NPK(kg) 4,200 0 80 80 80 40 0Ure(kg) 72,000 87 87 87 87 87 87Kaliclorua(kg) 10,000 80 80 80 80 80 80Suppelân(kg) 3,700 142 142 142 142 142 142Herbicide(liter) 65,000 6 6 6 6 6 6Pesticide (spraytimes) 600.000 1 1 1

Seeds Cassava(stems) 100 10,000 10,000 10,000 10000 10,000 10,000peanut(kg) 40,000 0 0 0 20 0 0cowpea(kg) 40,000 0 10 0 0 0 0mungbean(kg) 50,000 0 0 10 0 0 0GrassGhi-ne(kg) 4000 0 0 0 0 1,000 0Labour(workingdays)

Prepareland 60 60 60 60 60 60Planting 20 20 20 20 20 20Topdressing 2 2 2 2 2 2weeding1st 25 30 30 30 25 25weeding2nd 20 25 25 25 20 20Harvestcassava 73 70 71 71 73 73Intercropping Planting 0 10 10 10 0Praypesticide 4 4 4 Harvestintercrop 0 10 6 10 0Grasstrip Planting 10 Harvest 5 ControlCotourline 2Yieldandprice Cassava(tons) 1,400,000 15.47 12.93 13.01 13.76 14.06 15.26peanut(kg) 20,000 0 0 0 254 0 0Cowpea(kg) 30,000 0 240 0 0 0 0mungbean(kg) 30,000 0 0 80 0 0 0Grasstrip livestock 0 0 0 0 248 0

52

Table52:Detailedcostandlabourrequirement,FertilizertrialinBóMười(for1ha)

Price(Vnđ) T0 T1 T2 T3 T4Fertilizersandherbicides NPK(kg) 4,200 0 300 600 0 0Ure(kg) 72,000 0 0 0 87 0Kaliclorua(kg) 10,000 0 0 0 80 0Suppelân(kg) 3,700 0 0 0 142 0FDP(kg) 10.000 0 0 0 0 309Herbicide(liter) 65,000 7 7 7 7 7Seeds Cassava(stems) 100 10,000 10,000 10,000 10000 10,000Labour(workingdays) Prepareland 60 60 60 60 60Planting 20 20 20 20 20Topdressing 0 0 0 2 0weeding1st 25 25 25 25 25weeding2nd 20 20 20 20 20Harvestcassava 85 85 83 80 80Yieldandprice Cassava(tons) 1,400,000 23.43 23.22 22.2 18.28 17.1