Tillage, mulch and fertiliser impacts on soil nitrogen availability and maize production in semi-arid Zimbabwe

Full text

(1)

Tillage,

mulch

and

fertiliser

impacts

on

soil

nitrogen

availability

and

maize

production

in

semi-arid

Zimbabwe

Esther

N.

Masvaya

a,b,

*,

Justice

Nyamangara

c

,

Katrien

Descheemaeker

b

,

Ken

E.

Giller

b

a

InternationalCropsResearchInstitutefortheSemi-AridTropics,MatoposResearchStation,P.O.Box776,Bulawayo,Zimbabwe b

PlantProductionSystemsGroup,WageningenUniversity,P.O.Box430,6700AK,Wageningen,TheNetherlands c

DepartmentofEnvironmentalScienceandTechnology,ChinhoyiUniversityofTechnology,P.Bag7724,Chinhoyi,Zimbabwe

ARTICLE INFO Articlehistory: Received27May2016

Receivedinrevisedform28November2016 Accepted17December2016

Availableonlinexxx Keywords:

Agronomicefficiency ApparentNrecovery Cropresidueretention

ABSTRACT

Conservationagriculturehasbeenpromotedwidelyinsub-SaharanAfricantocushionsmallholder farmersagainsttheadverseeffectsofsoilfertilitydecline,stabilizecropyieldsandincreaseresilienceto climate changeandvariability. Ourstudyaimedto determineif aspectsofCA,namely tillageand mulchingwithmanureandfertiliserapplication,improvedsoilmineralNrelease,plantNuptakeand maizeyieldsincroppingsystemsonpoorsoilsinsemi-aridMatobo,Zimbabwe.Theexperiment,runfor threeseasons(2012/13–2014/15),wasasplit-splitplotdesignwiththreereplicates.Tillage (animal-drawnploughingandripping)wasthemainplottreatmentandresidueapplicationwasthesubplot treatmentwithtwolevels(100%residuesremovedorretainedafterharvest).Fivefertilityamendments (mineralfertiliserat0,20and40kgNha-1,5tha-1manureonlyand5tha-1manure+20kgNha-1)were sub-subplottreatments.PloughtillagestimulatedN mineralisationby 4–19kgNha-1andmaizeN uptake13–23%morethantherippertillage.Whenmulchwasaddedtotheploughtillage,mineralisation wasslowedresultinginlesscropNuptake(by5–19%)comparedwithnomulchapplication.Nuptakewas highestinthemanuretreatments.NrecoveryandagronomicNefficiencybymaizewerehighlyvariable overthethree seasons, reflectingtheuncertaintycomplicatingfarmers’decisionmaking. Nitrogen recoveryinthemanuretreatmentswasgenerallypoorinthefirstseasonresultinginlowgrainyieldsin therange100–260kgha-1regardlessoftillage,thoughhigherinsubsequentseasons.Inthesecond seasonmanureapplicationgavethelargestgrainyieldsundertherippertillage,bothwithandwithout mulchaveraging1850and2228kgha-1respectively.Undertheploughtillage,the40kgNha-1treatment gavethehighestgrainyieldsof1985kgha-1.Inthethirdseasonyieldsweregenerallypoorunderall treatmentsduetolowandpoorly-distributedrainfall.TheCAprinciplesofminimumsoildisturbanceand maintenanceofapermanentmulchcoverresultedinreducedsoilmineralNavailabilityforcropuptake and poormaizeyields.Nutrientinputsthroughmineralfertilisersandmanurearekeyto ensuring productioninsuchinfertile,sandysoilswhichpredominateinsemi-aridregionsofsouthernAfrica.

©2016ElsevierB.V.Allrightsreserved.

1.Introduction

Smallholder farmers practicing rainfed agriculture in sub-SaharanAfrica(SSA) facea myriadofchallenges;theseinclude poorsoilfertility,lowincomes,labourandlandconstraints,andare furtherexacerbatedbyclimatevariability(Mupangwaetal.,2012; Ngomaetal.,2015;Rurindaetal.,2013).Differentapproachesto improve soil fertility have been proposed including biological nitrogenfixation,soilsurfaceresiduemanagement,fertiliseruse,

enhancedrecyclingofanimalmanureandconservationagriculture (Hobbs, 2007; Mupangwa et al., 2012; Ncube et al., 2009; Nyamangaraetal.,2005).

Conservation agriculture(CA)hasbeenpromotedthroughout sub-humid and semi-arid areas of SSA to cushion smallholder farmersagainsttheadverseeffectsofsoilfertilitydecline,cropyield decline, andclimate change and variability(Hobbs etal., 2008; Ngomaetal.,2015).CAisbasedonthreeprinciples:(i)minimumor notillagetominimisesoildisturbance,(ii)diversificationofcrop species(oftenwithlegumes)growninrotationand/orassociation, and(iii)maintainingsemi-permanentorpermanentsoilcover,for example by leaving at least 30% of crop residues (FAO, 2011; Stevensonetal.,2014).Thisthree-prongedapproachisreportedto * Correspondingauthorat:InternationalCropsResearchInstituteforthe

Semi-AridTropics,MatoposResearchStation,P.O.Box776,Bulawayo,Zimbabwe. E-mailaddresses:e.masvaya@cgiar.org,enmasvaya@gmail.com(E.N. Masvaya). http://dx.doi.org/10.1016/j.still.2016.12.007

0167-1987/©2016ElsevierB.V.Allrightsreserved.

ContentslistsavailableatScienceDirect

Soil

&

Tillage

Research

(2)

havethepotentialtoimprovefarmresourceuseefficiencyandcrop yieldsespeciallywheremoistureislimiting(Hobbsetal.,2008).

AlthoughCAispromotedasasoil-fertilityenhancingtechnology, applicationofcropresiduespoorinN,suchascerealstover,may result in prolonged immobilisationof mineral N(Giller et al.,1997). In CAsystems, reduced N availability has beenattributed to slow residuedecompositionandNlossesfromleachinganddenitrifi ca-tion(Angásetal.,2006;Verachtertetal.,2009).Nyamangaraetal. (2014)proposedthatlargerNinputsmayberequiredunderCAto offsettheNimmobilisationcausedby cerealstover.Acombinationof highqualitymanurecombinedwithlowqualitycropresiduemay reduceNleakageandincreasenutrientuseefciency(Kiharaetal., 2011).Qin etal.(2015) found thatlargerN inputsresultedin a positiveeffectofstrawmulchonmaizeyields.Insub-humidwest Africa,maizegrainyieldinano-tillsystemwasonlyincreasedby mulch when fertiliser was also applied (Lal, 1995). However, smallholderfarmersinSSAtypicallyapplyonlysmallamountsof Nwhichmaynotbeadequate.(Vanlauweetal.,2014)arguethat appropriatefertiliseruseshouldbeconsideredafourthprincipleof CAasfertiliserisrequiredtoenhancebothcropproductivityand producesufficientcropresiduestoensuresoilcover.

Inthecurrentsmallholderfarmingsystem,theperturbationby tillagestimulatesaflushofmineralN(the“Bircheffect”)withthe startoftherains(Chikowoetal.,2004;Gilleretal.,2011),whereas soil organic matter may have otherwise been protected from degradation.This“Bircheffect”isreportedtobeshort-livedand thedecompositionratesmayfallbacktoratessimilartothatofan undisturbedsoil (Anderssonand Giller,2012). Minimum tillage promotedinsmallholderareasinSSAunderCAismainlyfocused onthehandhoeplantingbasins(Gilleretal.,2011;Nyakudyaand Stroosnijder,2015).Farmerswithlimitedaccesstodraughtpower andusinghandhoespreparetheirfieldsinthedryseasoninorder tospreadlabourrequirementsforlandpreparation,allowingfor early planting(Nyamadzawoet al., 2012).However, the useof animaldrawnconservationtillagemethodssuchastheripperand directseederprovidesanopportunitytoreducethelabourdemand associatedwithlandpreparationusinghandhoes.Mechanisation canincreaseproductivityperunitareabyimprovingtimelinessof farmoperationsincludingplanting.Earlyplantingmaycoincide withthe“Bircheffect” whichisbeneficialtothecrop(Chikowo etal.,2004).MinimumsoildisturbanceinCAsystems,however, results in slower mineralisation compared with conventional tillagebecauseoftheminimumdisturbance(Chivengeetal.,2007) leadingtopreservationofsoilorganicmatterfromdecomposition. Duetothisslowmineralisation,(Lal,2007)suggestedthatresource poorfarmerswould bebetteroffploughingtheirsandysoilsto enhancemineralisationofwhateversoilorganicmatterpresentto enhancenutrientsupplyintheshortterm.Thereare,however,no detailedstudiesonseasonalmineralNavailabilityinthesemi-arid areasunderCApracticessuchasminimumtillageandcropresidue retentionparticularlyonsoilsofpoorfertilitythataretypicalin smallholderagricultureinSSA.

Wehypothesisethatforcroppingsystemsinsemi-aridclimates andonpoorfertilitysoils,thebenetsofCAandaddedmulchare in immobilising N so that it is not lost from the system and becomes available later for crop growth. The study specifically aimedatdeterminingifandhowtillage, mulching,manureand fertiliserapplicationandtheirinteractionsimprovedsoilmineral Nrelease,plantNuptakeandultimatelymaizeyields.

2.Materialsandmethods

2.1.Sitedescription

The study was carried out in Nqindiward, Matobo district, MatabelelandSouth,Zimbabwe(2039.580S,2815.580E;900masl).

Matobodistrict liesinAgroecological ZoneIV, characterisedby semi-arid climate typical of south west Zimbabwe. Rainfall is unimodalwithadistinctwet(November–March)anddry(April– October) season. The wet seasonreceives 450–650mm annual rainfall with a long term average annual rainfall of 580mm. Droughts are frequent as are severe dry spellsduring the wet season. There is only a 45–65% probability of rainfall between Octoberand Aprilexceeding500mm (Vincentet al.,1960).The dominantsoilsareEutricArenosols derivedfromgranite (WRB, 2006).Thesesandysoilsconstitute>15%ofthetotallandareain

Zimbabwe (Hartemink and Huting, 2008). The smallholder

farming system in Matabeleland is characterised by privately managedarablefields andcommunally-managedgrazing lands. Thearablefieldsarealsocommunallygrazedduringthedryseason unlessifsecurelyfencedoff.Matobodistrictislargelyrural(99.4%) with Nqindi ward having a total population of 3507 persons (ZIMSTAT,2012).

2.2.Triallayoutandtreatments

AfieldexperimentwassetupinDecember2012onaslightly sloping (<2%) farmer’s field and run for three seasons. The experimentwassetupasasplit-splitplotwithplotsarrangedina randomised complete block design with three replicates. The tillagesystemwasthemainplottreatmentwithtwolevels

(ox-drawn ploughing and animal drawn ripping) and the mulch

managementwas thesub plottreatment withtwolevels(100% residueremovedand100%residuesretainedafterharvest).The mulchsub-treatmentwasnotappliedinthe2012/13seasonasthis wasthefirstseason.Fivefertilityamendments(mineralfertiliserat 0,20and40kgNha1,5tha1manureonlyand5tha1manure+ 20kgNha1) were randomised as the sub-sub plot treatment. Plotsmeasured35m12mwithbordersmeasuring5m15m. ThethirdCAprinciple,croprotation,wasnotincludedinthestudy becauseoftheshortdurationoftheexperiment.

2.3.Trialmanagement

A basal application of compound D fertiliser (14kgNha1, 12kgPha1and12kgKha1)ormanurewasappliedinplanting furrows.The0Nfertilitytreatmentreceivedabasalapplicationof single super phosphate (12kgPha1) and muriate of potash (12kgKha1). All plots were planted to a short season maize variety (SC403) at 0.9m0.3m spacing to achieve a plant population of 37 000 plants ha1. The remainder of the N requirement(6,20,26kgNha1forthe20kgNha1,manure+20 kgNha1and40kgNha1treatmentsrespectively)wasappliedas atopdressing,usingammoniumnitrate(whichcontained34.5%N) at six weeks after planting and when there was enough soil moisturefor topdressing.In thesecond seasonincessantrains caused waterlogging such that top dressing application was delayed until nine weeks after planting which coincided with thecropfloweringstage.Theplotswerekeptweedfreebyaninitial applicationofglyphosate[N-(phosphono-methyl)glycine] herbi-cide soon after planting and thereafter by hand hoeing when required.Atharvest,inplotswheremulchwastobeapplied,the stoverwasleftonthesoilsurface.Theamountofmulchretained onthesurface wasdependentonthebiomass producedin the previousseasonsandrangedfrom1to2tha1afterthe2012/13 seasonand2–4tha1afterthe2013/14seasonwhichtranslatedto approximately 10–20% and 20–30% surface cover at planting respectively(basedonvisualassessmentofgroundcover).Inthe plough tillage treatment, the mulch was ploughed in at land preparation.Theplotswerelocatedinafencedoffsectionofthe farmandthusnolivestockcouldfeedoncropresiduesduringthe drymonths.

(3)

2.4.Rainfalldata

Thefarmerhostingthetrialwasprovidedwitharaingaugeand recordbook torecord dailyrainfall for the2012/13 to2014/15 seasons. The rainfallrecords werecompared and verified with records from the nearby Matopos Research Institute weather station.

2.5.Soil,manureandplantsamples

Initial soil samples were collected from each block at

incremental depths of 0.10m upto 1m. The samples for each depthwerebulked,mixedandanalysedseparately.Soilswereair dried and sieved through a 2mm sieve and analysed for pH, texture, total and mineral N, Olsen P and organic C (Table 1). Aerobically composted manurewas collected from one source (MatoposResearch Institute’sBeefProductionSection)toavoid variability.Themanurewasdugoutfromthekraalstwomonths priortoapplication.ManuresampleswereanalysedfortotalC,N and P using themodified Walkley-Blackmethod withexternal heating,micro-KjeldahlandthemodifiedOlsenmethods respec-tively (Anderson and Ingram, 1993). The manure contained 199.0gCkg1, 9.8gNkg1 and 2.6gP kg1 of manurein 2012/ 2013; 201.1gCkg1, 9.5gNkg1 and 2.7gPkg1 of manure in 2013/2014 and 196.0gCkg1, 10.0gNkg1 and 2.6gPkg1 of manurein2014/15.

2.5.1.Nmineralisationmeasurements

Detailed field measurements of inorganicN dynamics were madeusingin-situincubationofundisturbedsoilcores(Anderson andIngram,1993; MurwiraandKirchmann,1993;Raisonetal., 1987)throughoutthe2013/14growingseason.Sixtubes(0.35m longwithaninternaldiameterof5cm)wereinsertedrandomlyin eachplot,soonafterlandpreparation,toadepthof0.30m.Threeof thetubeswereremovedimmediatelyandthesoilbulkedforeach plot.InitialmineralN(NH4+andNO3)wasdeterminedfromthese samples.Atthesurface,thetubeswerecoveredwitha polyethyl-enesheettopreventwaterfromentering.Thepolythenesheets

were removed every morning for aeration. These tubes were

removedandreplacedatfourweekintervals(days28,56,84and 112afterplanting)untilharvesting.Nwasextractedfromthesoil samplesbyshakingthefieldfreshsamplein0.5MK2SO4andthe NH4+-N and NO3-N content was determined using methods described in Anderson and Ingram (1993). The net amount of mineralised N was calculated as the difference in mineral N betweentwopointsintime(timei+1–timei).

2.5.2.Plantsamplesandcropyield

Atharvestingmaizegrainandstover(above-groundbiomass minusgrain) yieldsweredeterminedfromnetplotsmeasuring 4.5m3m.Grainandstoversamplesweresubsampledandgrain yieldswereadjustedto12%moisturecontent.Plantsampleswere collected at four-week intervals (28days) from planting until harvestinthe2013/14seasonandatharvestattheendofevery season.Theplantsamplesweredriedat70Cfor48handanalysed for total N content (Bremnerand Mulvaney,1982) tocalculate

aboveground N uptake. The apparent N recovery (ANR) was

calculatedasfollows:

ANR¼NuptakeFNuptakeC

FertilizerNapplied X100%

ThegrainyieldswereusedtocalculateagronomicNefciency (AE)asfollows:

AE¼GrainyieldFGrainyieldC

FertilizerNapplied kggrainkg

1N

whereFandCdenote“fertilisedcrop”and“unfertilisedcontrol” respectively(Mengeletal.,2001)

2.6.Statisticalanalysis

All parameters were tested for normality and found to be normallydistributedusingtheShapiro–WilkWtest(Shapiroand Wilk, 1965). Soil chemical characteristics were square root

transformed to homogenise variances (Gomez and Gomez,

1984). Analysis of variance was conducted on the soil N

mineralisation, N uptake, ANR, AE, grain and stover yield. The means of the treatments were separated by least significant difference (LSD) at 5% level of significance. All analyses were conductedusingGenstat14thEdition(VSN,2011).

3.Results

3.1.Rainfalldistribution

Rainfall amount, distributionand intensitywereerratic and highlyvariableacrossthethreeseasons(Fig.1).Severemid-season dry spells characterisedthe first (2012/13) and third (2014/15) seasons.Thefirstandthirdseasonsreceivedrainfalltotalsof272

and 432mm that were 54 and 27% lower than the long term

averageforthesite(590mm)respectivelyandwerealsobelowthe lowerlimit(450mmperannum)for theagro-ecologicalregion. Bothseasons wereclassified asdroughts andthewater-limited cropyieldpotentialwasnegativelyaffected.Inthesecondseason

Table1

SelectedsoilcharacteristicsofthestudysiteinNqindiward,Matobodistrict,south-westZimbabwe. Depth(cm) pH(CaCl2) TotalN

(gkg1) OrganicCarbon (gkg1) C:Nratio MineralN (kgha1) OlsenP (mgkg1)

Particlesizeanalysis

%Sand %Silt %Clay

0–10 4.7 0.51 6.6 13.2 23.3 11.1 84 16 0 10–20 4.7 0.40 4.2 10.5 26.4 5.5 90 10 0 20–30 4.6 0.30 3.5 11.7 33.9 2.6 90 10 0 30–40 4.8 0.30 2.5 8.3 15.7 1.3 88 10 2 40–50 4.8 0.20 1.4 7.0 6.9 ND* 90 10 0 50–60 4.8 0.30 1.1 3.7 11.3 ND 94 6 0 60–70 4.9 0.20 0.7 3.5 15.0 ND 90 10 0 70–80 5.0 0.18 0.5 3.1 21.4 ND 92 8 0 80–90 5.1 0.08 0.3 3.8 11.3 ND 92 8 0 90–100 5.0 0.08 0.3 3.8 16.9 ND 86 14 0 ND–notdetected.

(4)

(2013/14),therainfalltotalof872mmexceededthesiteaverageby 48% and several high intensitystorms resulted in intermittent waterloggingatthestudysite.

3.2.MineralN,Nuptake,apparentNrecoveryandagronomicN

efficiency

3.2.1.SeasonalmineralNdynamicsandNuptake

TherewasaninitialincreaseinmineralNinthefirst28days after planting, mineralisation then stabilized until day 84 (vegetative and reproductive stages) and then increased until harvesting (maturity stage) (Fig. 2). N mineralised during the growingseasonrepresented0.71–2.3%ofthetotalNpoolinthe0– 0.3msoildepth.Tillageandfertilityamendmentshadasignificant effectonNmineralisation(P<0.05)throughouttheseason.The ploughtillageresultedinmoreNmineralisationthantheripper tillage by 4–19kgNha1. There was also more mineralisation whereNwasappliedeitherasmineralfertiliserorcattlemanure comparedwiththecontrol(0N)whichgenerallyhadthelowest cumulativeNmineralisation.Exceptwhenploughingand mulch-ingwereapplied, thehighest cumulative Nmineralisation was obtainedwiththe40Ntreatment.Mulchingonitsowngenerally

had no significant effect (P>0.05) on N mineralisation but its interactions with tillage and fertility amendments significantly influencedNmineralisation.Whenmulchwasploughedin,there

was a decrease in mineralisation compared with ploughing

withoutmulchby8–15kgNha1(1340%)acrossfertility amend-menttreatments,indicatingN immobilisation.Undertheripper tillage, however,mulchingresultedin a16–100%higher miner-alisationcomparedtonomulching.Nmineralisationfollowedthe trendplough(nomulch)>ripper+mulch>plough+mulch>ripper (nomulch)regardlessoffertilityamendment.

Nuptakegenerallyincreaseduntilday84whentotalplantN levelled off. Maize that received N fertility amendments had significantly greaterN uptake (P<0.05)throughout the season than the 0N control (Fig. 2). Except for the plough+mulch

treatment, N uptake was highest in the manure treatments.

Tillagehadasignificanteffect(P<0.05)ontotalNuptakeatday84 andday112.MaizeNuptakeundertherippertillagewas13–23% lowerthantheploughtillage.Mulchinghadasignificanteffecton Nuptakebymaize(P<0.05).Incorporation ofmulchunderthe ploughtillageresultedinsignificantlylowerNuptake(P<0.05)by 5–19%whencomparedwithploughwithoutmulch.Nuptakeand followedthetrendplough(nomulch)>ripper+mulch>ripper(no Fig.1.Cumulativerainfall(a–c)inNqindiward,Matobodistrict,Zimbabwefromthe2012/13–2014/15.Thedashedline(a–c)isthelongtermseasonalaveragefortheregion. Thesolidarrowsshowtheplantingdateformaize.

Fig.2.CumulativeNMineralisationestimatedfromtheinsituincubationtechnique(0–30cm)(a–d)andNetNuptakebymaize(abovegroundparts)(e–h)inthe2013/14 seasoninfluencedbysoilfertilityamendmentsunderploughonly(aande),plough+mulch(bandf),ripperonly(candg)andripper+mulch(dandh)onasandysoilin Nqindiward,Matobodistrict,south-westZimbabwe.Barsrepresentstandarderrorsofthedifferenceofthemeansfortheeffectoftheinteractionoftillage,mulchandfertility amendmentoncumulativeNmineralisationandNuptakeondays28,56,84and112.

(5)

mulch)>plough+mulch.TheeffectofthetopdressingNfertiliser, appliedat48daysafterplanting,was not apparentin both the cumulativeNmineralisationandtheNuptake.

3.2.2.ApparentNrecovery(ANR)andagronomicNefficiency(AE)

TheANRandAEwereveryvariableacrossthethreeseasonsand treatments(Table2).Onlyfertilityamendmentshadasignificant effect(P<0.05)onboththeANRandAEinthefirstseasonandon

ANR in the second season. In the second season, fertility

amendment significantly affected ANR with recovery highest

under the 20N treatment regardless of tillage and mulch

application. Tillage and mulching had no signicant effect on bothANRandAEinallthreeseasons.Theinteractionsofmulchand fertilityamendmentsandofallthreetreatments(tillage,mulch andfertilityamendments)hadsignificanteffects onANRin the secondand thirdseasons althoughthesedidnotfollow similar trendsbetweenthetwoseasons.ANRwashighestundertheripper (nomulch)+20Nfertilitytreatmentinthesecondseasonwhilstin thethirdseason,ANR washighest theplough+mulch+manure only.ThethirdseasonhadthepoorestandhighlyvariableANRand AE.AlltreatmentshadANRintherange7.0–40.4%andaveraged 15%.ThisseasonhadtheleastAEthatranged(4.7)1.3kggrain kg1Nandaveraged(0.16)kggrainkg1N.Therewere,however, noindividualtreatmenteffectsbutinteractioneffectsofmulchand fertilityamendmentsandtillage,mulchandfertilityamendments onNrecoveredbythemaizecropinthethirdseason.Therewere nosignificanteffectsoftillage,mulch,fertilityamendmentsand theirinteractionsonAEinthesecondandthirdseasons.

3.3.Maizeyieldsinresponsetotillage,mulchandfertilityamendment

application

Grainyieldsinthefirstseasonwerelowandintherange105– 720kgha1underbothtillagetreatmentsandacrossallfertility amendments(Fig.3).Inthisseason,ploughinggavehighergrain

(200–720kgha1) and stover yields (15002100kgha1) com-paredwithripping(105–454kgha1grainand860–1500kgha1 stover)althoughnotsignificantfor grain(P<0.05).Theharvest indices,althoughnotsignificant,weregenerallyhigherunderthe plough tillage compared with ripping by 1–10%. There were significantdifferencesingrainyieldsresultingfromtheapplication

of different fertility amendments. Both manure treatments

resulted in low grainyieldsunder both tillagetreatments. The highestgrainyieldswereobtainedwhenmineralNfertiliserwas added(boththe20Nand40Ntreatments).Theeffectsoftillage, mulch,fertilityamendmentsandtheirinteractionsonstoveryields andtheharvestindiceswerenosignicant.

Although there were nosignificant differences in grain and stoveryieldsresultingfromtillage,ripping,ingeneral,significantly depressed grain yields by 4–60% in the second season when comparedwithploughing.Theeffectofmulchonlyongrainand stoveryieldswasnotsignificant(P<0.05).Mulchingsignificantly increasedgrainyieldsinallfertilitytreatmentsundertheripper tillageexceptforthe0Nandmanureonlytreatments.However, mulching reduced grainyieldswhen combined withploughing under the 20N, 40N and manure+20N fertility treatments by between2and60%.Inthesecondseason,fertilityamendments alonehadasignificanteffectongrainandstoveryields.Manure applicationgavethehighestgrainyieldsundertherippertillage, bothwithandwithoutmulch(averaging1850and2228kgha1 respectively).Undertheploughtillage,the40Ntreatmentgavethe highestgrainyieldsof1985kgha1onaveragewhilstthemanure treatments resulted in the highest stover yields. The harvest indices, werenot significantfor tillage, mulch, fertility amend-mentsandtheirinteractions

In thethirdseason,theeffects oftillage, mulchand fertility amendment treatment on grain and stover yields and harvest indiceswerenotsignificant.Grainandstoveryieldsweregenerally poorduetolowandpoorlydistributedrainfall.Grainyieldswerein therange200–270kgha1,whilststoveryields400900kgha1. Table2

Effectoftillage,mulchingandfertilityamendmentonapparentNrecoveryandagronomicNuseefficiencyoverthreeseasons(2012/13–2014/15)inNqindiward,Matobo district,south-westZimbabwe.

Tillage Mulchtreatment Fertilitytreatment %ApparentNrecovery AgronomicNefficiency (kggrainkg1 N) 2012/13 2013/14 2014/15 2012/13 2013/14 2014/15 Plough Nomulch 20N 67.3 109.2 12.2 19.9 19.2 0.5 40N 19.4 26.1 15.1 5.7 2.0 0.7 Manureonly 22.0 83.6 14.2 1.1 27.9 1.1 Manure+20N 16.6 83.4 11.5 1.9 21.6 0.6 Mulch 20N – 107.7 7.8 – 21.3 4.7 40N – 82.8 7.1 – 25.5 2.1 Manureonly – 35.4 36.4 – 27.7 1.6 Manure+20N – 23.4 14.7 – 8.6 0.7 Ripping Nomulch 20N 42.9 136.1 5.6 8.9 3.6 1.7 40N 8.3 104.8 5.9 8.7 4.8 0.5 Manureonly 13.9 110.1 18.9 2.9 12.4 0.2 Manure+20N 12.2 59.9 16.2 0.2 13.2 0.7 Mulch 20N – 107.2 2.6 – 8.4 0.3 40N – 43.6 26.7 – 44.2 0.7 Manureonly – 73.7 23.8 – 18.8 1.3 Manure+20N – 2.4 8.0 – 18.1 0.4

Factor PvalueSED PvalueSED PvalueSED PvalueSED PvalueSED PvalueSED

Tillagetype 0.46313.37 0.1992.34 0.5348.49 0.903.96 0.8043.61 0.5740.72

Mulch – 0.34818.19 0.7764.55 – 0.30511.28 0.8811.53

Fertilityamendment 0.007*

10.95 0.01416.21 0.1655.95 <0.012.72 0.1818.53 0.0410.79

TillagexMulch – 0.94518.33 0.2719.63 – 0.96211.85 0.4041.69

TillagexFertilityamendment 0.81318.93 0.49619.99 0.09511.19 0.0565.25 0.58811.38 0.5691.24

MulchxFertility – <0.00126.92 0.0318.59 – 0.15915.62 0.4411.83

TillagexMulchxFertility – <0.00133.54 0.01514.11 – 0.37819.32 0.0742.21

*

(6)

The grain yields followed the same trends as in the previous seasons.

4.Discussion

4.1.Nmineralisationpatterns,uptakeandrecovery

Ingeneral,theConservationAgricultureprinciplesofminimum soil disturbance and maintenance of a permanent soil cover (mulch)resultedin reduced soilmineralN availabilityfor crop uptakeandultimatelylowyieldsfromapoorfertilitysoilin semi-aridZimbabwe.Ploughingenhancesthemineralisationoforganic matterinthesoilbyexposingpreviouslyunexposedsoilsurfaces toattackbymicrobesandbyprovidingthelatterwithnewsources ofenergy(Peignéetal.,2007).Comparedwithploughing,ripper tillagecauses lesssoil disturbanceand exposesthesoil organic matterlesstotheclimaticelementsandsoilmicrobes,withslower mineralisationas a result.Our results concurwith studiesthat reportedslowermineralisationoforganicmatterunderreduced tillagecomparedwithconventionaltillagebecauseoflesstopsoil disturbance(Bationoetal.,2007;Chivengeetal.,2007).Thislow mineralisationassociatedwiththerippertillageresultedinlowN uptakeand consequently decreased grainyields(in thesecond season)whencomparedwiththeploughtillagewhennomulch wasappliedunderbothtillagetreatments.

When mulch was ploughed in, mineralisation was slowed

comparedwithnomulchapplication.SoilinorganicNmayhave beenimmobilisedduringthedecompositionoftheappliedmulch witha largeC:Nratioaveraging53:1whilstthesoil’sC:Nratio averaged12:1.TheC:Nratioofcropresiduesisagoodindicatorof whethermineralisation orimmobilisationwilldominateduring decomposition. N release patterns are regulated by the initial compositionofthecropresiduesandthestoichiometric require-mentsofthedecomposers(Manzonietal.,2008;Palmetal.,2001). Mineralfertilisers providea readysourceof nutrientsandmay stimulatedecompositionofcerealstover(Sakalaetal.,2000).

We observed substantialN mineralisation at theend of the season,whichproducedapoolofmineralNthatmaybeheldinthe soiloverthedryseasonduetotheabsenceofleachingandreduced Nuptake byplants(Rasouliet al.,2014).Perturbationbytillage

furtherstimulatesaflushofNmineralisationwiththestartofthe rains.CropsarenotabletostoresurplusNtoimproveyieldsand grainqualityinlatervegetativestages(Verhulstetal.,2014).This impliesthattheNmineralisedattheendoftheseason,aswellas thatwhichbecomesavailablethroughthe“Bircheffect”,maywell belostduetothefirstrainsaheadofplantingorbeforethenext cropisestablishedandgrowingactively.Intheirstudy,(Chikowo et al., 2003) concluded that there is an inherent problem in managingNoriginatingfrommineralisationasitaccumulatesat thebeginningoftheseason,wellaheadofpeakdemandandroot developmentbycrops,andissusceptibletoleaching.

Therewasadecreaseinobservedmineralisationinthecritical vegetativestage.Forsemi-aridregions,(Piha,1993)suggestedsplit applicationsofNfertiliseratratesbasedonfutureexpectedrainfall to maximise fertiliser use and recovery efficiency. In our experiment,weappliedfertiliseratplantingandat48daysafter planting.ThesecondNapplicationdidnotresultinasignificant observedincreaseinNavailability.Itmayhavebeenbecauseatthe startoftheincubationexperiment,themineralNwasincreasedby the basal fertiliser application such that a further addition in mineralNlateronintheseasonmaynothavebeenapparent.The success of split N application may depend on the number of fertiliserapplicationsandtheirtimingandquantities,onweather conditionsand ontheamountofavailable soilmineralN(Piha, 1993;Verhulstetal.,2014).Inourexperiment,thetimingofthe secondapplicationinthesecondseasonwas affectedby water-logging.Wehadtowaituntilallthewaterhadinfiltratedbeforewe could apply the top dressing fertiliser. The second application coincidedwiththemaizefloweringstage,acriticalgrowthstage, whichmayhavecontributedtothehigheryieldsobservedinthis seasoninconjunctionwiththehigherrainfall.

The N recoveryand agronomic N efficiencyby maize were highly variable over the three seasons, which reflects the uncertainty complicating farmers’ decision making. Nitrogen recoverybymaizeinthemanuretreatmentswasgenerallypoor in the first season, though greater in subsequent seasons, as reportedearlier(Nyamangaraetal.,2004).Theauthorsattributed theslowNmineralisationfrommanuretotheCandNstabilisation

which occurred during aerobic decomposition of the manure

duringstorage.Nrecoveryfromthemanure+20Ntreatmentwas Fig.3.MaizegrainandstoveryieldsinNqindiward,Matobodistrict(2012/13-2014/15season).Barsrepresentstandarderrorsofthedifferenceofthemeansforfactors(a) Tillage,(b)Fertilityamendmentand(c)Mulching.

(7)

veryhighin thesecond season,which isattributedto residual fertilityfromthepreviousseason,which wasverydry.Farmers shoulduseresourcesavailabletothemasfertilityamendmentsto supplyNtothemaizecroppingsystem,eitherasorganicmanures orasmineralfertilisers.However,manureapplicationisapreserve of cattleowners and mineral fertilisers remainunaffordable to most smallholder farmers in SSA including Zimbabwe (Chianu etal.,2012;MazvimaviandTwomlow,2009;Mtambanengweand Mapfumo,2005).

Maizegrainandstoveryieldswereaffectedbythequalityofthe seasonsintermsofrainfallamountandseasonaldistribution.Two ofthethreeseasons(therstandthird)inwhichtheexperiment was conductedwereclassifiedas droughts,negatively affecting yields.Lowrainfallandpoorrainfalldistributionarea character-isticofandmajorchallengestocropproductioninsemi-aridareas of southern Africa (Nyamangara et al., 2014).Thierfelder et al. (2013) reported that CA systems showed great potential in mitigatingtheeffectsofseasonaldry-spells,asthehighinfiltration rates in CAwould lead togreater soil moistureavailability for crops.Thiswas,however,notthecaseinourexperiment(witha mulch cover of 0 and 2–4tha1 in the rst and third seasons respectively)asthetwoseasonswerecharacterisedbylongdry spellsmidwaythroughtheseasonlastingmorethan21days.

4.2.Methodologicallimitationsoftheinsituincubationmethod

One limitation of the in situ method in estimating N

mineralisationisthatinreplacingsoilcoresintothesoil,maize rootsmayhavebeenseveredandretainedwithinthetubes.These rootsmayhaveimmobilisedNduringtheirdecayleadingtoan underestimation of N mineralisation (Raison et al., 1987). ConverselyNmayhavebeenreleasedduringtheirdecomposition leading to an overestimation of N mineralisation (Khanna and Raison, 2013). Furthermore, in situ core methods promote the accumulationofNH4+N(bypreventinguptake).TheNH4+Nmay be nitrified leading to accumulation of NO3N which may increase denitrification (Raison et al., 1987). We measured mineralisationinthetop0.3mwhichistheploughlayeralthough maizerootsexploredeepersoilhorizons.HenceourN mineralisa-tionresultsmaynotaccountforalloftheNavailabletothecrop. Nevertheless,theinsituincubationmethodgaveusefulinsights intothedifferencesamongtreatments(Fig.2).

4.3.Inputaccess:drawbacksofsmallholderfarmers

Theprinciplesof CAminimumsoildisturbanceand mainte-nance of a mulch resulted in reduced mineralisation (Fig. 2) resulting in low yields whilst a combination of the principles

resulted in high mineralisation and yields comparable to

conventionalploughminusmulch (Fig.3).Fertiliserapplication resultedinyieldbenetsregardlessoftillageormulchapplication particularlywhenmoisturewasnotlimiting.Thepotentialbenefits ofthetillage,mulchapplicationandfertilityamendment applica-tiononcropproductivityandsoilNdynamicsarerelatedtothe managementintensityindividualfarmerscanachieve.Theuseof mineralfertilisersiskeytoproductionyettheirwidespreaduseby smallholder farmers remains low as availability in rural areas remainsakeyconstrainttotheiruseaswellastheircostremains prohibitive.Withover40%ofhouseholdsinsemi-aridZimbabwe havingaccesstodraughtanimals(ZimVac,2012),theuseofthe rippermayprovideasolutiontocurbthechroniclabourdemand associatedwithlandpreparationinmanualformsofin smallhold-ercommunities.However,onlyafewequipment manufacturers havebeenproducingrippertineattachmentsforthemouldboard ploughandtheserequirepre-financingforproduction.Thereare

poor market linkages between smallholder farmers market

linkages between smallholder farmers and local agro dealers, andagro-dealersandinputsuppliers.Atpresent,inputproducers areunwillingtoprovidefertilisersorequipmentsuchasripperson credittolocalfarmersindryareasasrisksofcropfailurefromthe erraticrainfallpatternarehigh.Thehighlikelihoodofcropfailure from the poor rainfall in the semi-arid areas is in itself a disincentive for farmers tothe purchase and useof fertilisers. Accesstoinputsstillremainsachallengeespeciallytopoorfarmers whoareoftenconcernedwithmeetingtheimmediatehousehold foodrequirementsovertheimprovementandmaintenanceofsoil fertilityonthesesandysoilsimportantinthelongterm. 5.Conclusions

Weobservedthatmineralisation,Nuptakeandcropyieldswere stimulatedunderploughtillagecomparedwiththerippertillage.

When mulch was added together with plough tillage, the

mineralisation of N, crop N uptake and maize yields were

decreasedcomparedwithnomulchapplicationunderthesame tillage.Whencombinedtogether,followingtwoprinciplesofCA, therippertillageandmulchapplicationdidnotreduce mineral-isationandresultedinyieldscomparabletothoseobtainedunder theplough tillagewithout mulch.Nitrogen containingfertiliser andmanureresultedinmoreavailableNforthecropandincreased grainyields.Weconcludethatnutrientinputsarekeytoensuring productionintheinfertile,sandysoilspredominantinsemi-arid regionsofsouthernAfrica.

Acknowledgements

We thank the Netherlands Organization for International CooperationinHigherEducation(NUFFIC)forfunding.

References

Anderson,J.M.,Ingram,J.S.I.,1993.TropicalSoilBiologyandFertility:AHandbookof Methods.CABInternational.

Angás,P.,Lampurlanés,J.,Cantero-Martínez,C.,2006.TillageandNfertilization: effectsonNdynamicsandbarleyyieldundersemiaridMediterranean conditions.SoilTillageRes.87,59–71.

Bationo,A.,Kihara,J.,Vanlauwe,B.,Waswa,B.,Kimetu,J.,2007.Soilorganiccarbon dynamics,functionsandmanagementinWestAfricanagro-ecosystems.Agric. Syst.94,13–25.

Bremner,J.M.,Mulvaney,C.,1982.Nitrogen—Total.MethodsofSoilAnalysis.Part2. ChemicalandMicrobiologicalProperties.,pp.595–624.

Chianu,J.,Chianu,J.,Mairura,F.,2012.Mineralfertilizersinthefarmingsystemsof sub-SaharanAfrica.Areview.Agron.Sustain.Dev.32,545–566.

Chikowo,R.,Mapfumo,P.,Nyamugafata,P.,Nyamadzawo,G.,Giller,K.E.,2003. Nitrate-Ndynamicsfollowingimprovedfallowsandmaizerootdevelopmentin aZimbabweansandyclayloam.Agrofor.Syst.59,187–195.

Chikowo,R.,Mapfumo,P.,Nyamugafata,P.,Giller,K.E.,2004.Maizeproductivityand mineralNdynamicsfollowingdifferentsoilfertilitymanagementpracticesona depletedsandysoilinZimbabwe.Agric.Ecosyst.Environ.102,119–131. Chivenge,P.,Murwira,H.,Giller,K.,Mapfumo,P.,Six,J.,2007.Long-termimpactof

reducedtillageandresiduemanagementonsoilcarbonstabilization: implicationsforconservationagricultureoncontrastingsoils.SoilTillageRes. 94,328–337.

FAO,2011.SaveandGrow:APolicymaker’sGuidetotheSustainableIntensification ofSmallholderCropProduction.FoodandAgricultureOrganizationofthe UnitedNations.

Giller,K.E.,Cadisch,G.,Ehaliotis,C.,Adams,E.,Sakala,W.D.,Mafongoya,P.L.,1997. BuildingsoilnitrogencapitalinAfrica.Repl.SoilFertil.Afr.151–192. Giller,K.E.,Corbeels,M.,Nyamangara,J.,Triomphe,B.,Affholder,F.,Scopel,E.,

Tittonell,P.,2011.Aresearchagendatoexploretheroleofconservation agricultureinAfricansmallholderfarmingsystems.FieldCropRes.124,468– 472.

Hartemink,A.E.,Huting,J.,2008.Landcover,extent,andpropertiesofArenosolsin SouthernAfrica.AridLandRes.Manage.22,134–147.

Hobbs,P.R.,Sayre,K.,Gupta,R.,2008.Theroleofconservationagriculturein sustainableagriculture.Phil.Trans.R.Soc.BBiol.Sci.363,543–555. Hobbs,P.R.,2007.Conservationagriculture:whatisitandwhyisitimportantfor

futuresustainablefoodproduction?J.Agric.Sci.Cambridge145,127. Khanna,P.K.,Raison,R.J.,2013.Insitucoremethodsforestimatingsoilmineral-N

fluxes:re-evaluationbasedon25yearsofapplicationandexperience.SoilBiol. Biochem.64,203–210.

(8)

Kihara,J.,Bationo,A.,Mugendi,D.N.,Martius,C.,Vlek,P.L.G.,2011.Conservation tillage,localorganicresourcesandnitrogenfertilizercombinationsaffectmaize productivity,soilstructureandnutrientbalancesinsemi-aridKenya.Nutr.Cycl. Agroecosyst.90,213–225.

Lal,R.,1995.Tillageandmulchingeffectsonmaizeyieldforseventeenconsecutive seasonsonatropicalalfisol.J.Sustain.Agric.5,79–93.

Lal,R.,2007.Constraintstoadoptingno-tillfarmingindevelopingcountries.Soil TillageRes.94,1–3.

Manzoni,S.,Jackson,R.B.,Trofymow,J.A.,Porporato,A.,2008.Theglobal stoichiometryoflitternitrogenmineralization.Science321,684–686. Mazvimavi,K.,Twomlow,S.,2009.Socioeconomicandinstitutionalfactors

influencingadoptionofconservationfarmingbyvulnerablehouseholdsin Zimbabwe.Agr.Syst.101,20–29.

Mengel,K.,Kirkby,E.A.,Kosegarten,H.,Appel,T.,2001.PrinciplesofPlantNutrition. SpringerNetherlands.

Mtambanengwe,F.,Mapfumo,P.,2005.Organicmattermanagementasan underlyingcauseforsoilfertilitygradientsonsmallholderfarmsinZimbabwe. Nutr.Cycl.Agroecosyst.73,227–243.

Mupangwa,W.,Twomlow,S.,Walker,S.,2012.Reducedtillage,mulchingand rotationaleffectsonmaize(ZeamaysL.):cowpea(Vignaunguiculata(Walp)L.) andsorghum(SorghumbicolorL.(Moench))yieldsundersemi-aridconditions. FieldCropRes.132,139–148.

Murwira,H.K.,Kirchmann,H.,1993.Nitrogendynamicsandmaizegrowthina Zimbabweansandysoilundermanurefertilisation.Commun.SoilSci.Plan24, 2343–2359.

Ncube,B.,Dimes,J.P.,vanWijk,M.T.,Twomlow,S.J.,Giller,K.E.,2009.Productivity andresidualbenefitsofgrainlegumestosorghumundersemi-aridconditions insouth-westernZimbabwe:unravellingtheeffectsofwaterandnitrogenusing asimulationmodel.FieldCropRes.110,173–184.

Ngoma,H.,Mason,N.M.,Sitko,N.J.,2015.Doesminimumtillagewithplanting basinsorrippingraisemaizeyields?Meso-paneldataevidencefromZambia. Agric.Ecosyst.Environ.212,21–29.

Nyakudya,I.W.,Stroosnijder,L.,2015.Conservationtillageofrainfedmaizein semi-aridZimbabwe:areview.SoilTillageRes.145,184–197.

Nyamadzawo,G.,Nyamugafata,P.,Wuta,M.,Nyamangara,J.,Chikowo,R.,2012. Infiltrationandrunofflossesunderfallowingandconservationagriculture practicesoncontrastingsoils,Zimbabwe.WaterSA38,233–240. Nyamangara,J.,Piha,M.,Giller,K.,2004.Effectofcombinedcattlemanureand

mineralnitrogenonmaizeNuptakeandgrainyield.Afr.CropSci.11,300–389. Nyamangara,J.,Mudhara,M.,Giller,K.,2005.Effectivenessofcattlemanureand

nitrogenfertilizerapplicationontheagronomicandeconomicperformanceof maize.S.Afr.J.PlantSoil22,59–63.

Nyamangara,J.,Nyengerai,K.,Masvaya,E.,Tirivavi,R.,Mashingaidze,N., Mupangwa,W.,Dimes,J.,Hove,L.,Twomlow,S.,2014.Effectofconservation agricultureonmaizeyieldinthesemi-aridareasofZimbabwe.Exp.Agr.50, 159–177.

Palm,C.A.,Gachengo,C.N.,Delve,R.J.,Cadisch,G.,Giller,K.E.,2001.Organicinputs forsoilfertilitymanagementintropicalagroecosystems:applicationofan organicresourcedatabase.Agric.Ecosyst.Enviro.83,27–42.

Peigné,J.,Ball,B.,Roger-Estrade,J.,David,C.,2007.Isconservationtillagesuitablefor organicfarming?Areview.SoilUseManage.23,129–144.

Piha,M.,1993.Optimizingfertilizeruseandpracticalrainfallcaptureinasemi-arid environmentwithvariablerainfall.Exp.Agr.29,405–415.

Qin,W.,Hu,C.,Oenema,O.,2015.Soilmulchingsignificantlyenhancesyieldsand waterandnitrogenuseefficienciesofmaizeandwheat:ameta-analysis.Sci. Rep.5.

Raison,R.J.,Connell,M.J.,Khanna,P.K.,1987.Methodologyforstudyingfluxesofsoil mineral-Ninsitu.SoilBiol.Biochem.19,521–530.

Rasouli,S.,Whalen,J.K.,Madramootoo,C.A.,2014.Review:reducingresidualsoil nitrogenlossesfromagroecosystemsforsurfacewaterprotectioninQuebec andOntario,Canada:bestmanagementpractices,policiesandperspectives. Can.J.SoilSci.94,109–127.

Rurinda,J.,Mapfumo,P.,vanWijk,M.,Mtambanengwe,F.,Rufino,M.,Chikowo,R., Giller,K.,2013.Managingsoilfertilitytoadapttorainfallvariabilityin smallholdercroppingsystemsinZimbabwe.FieldCropRes..

Sakala,W.D.,Cadisch,G.,Giller,K.E.,2000.Interactionsbetweenresiduesofmaize andpigeonpeaandmineralNfertilizersduringdecompositionandN mineralization.SoilBiol.Biochem.32,679–688.

Shapiro,S.S.,Wilk,M.B.,1965.Ananalysisofvariancetestfornormality(complete samples).Biometrika591–611.

Stevenson,J.R.,Serraj,R.,Cassman,K.G.,2014.Evaluatingconservationagriculture forsmall-scalefarmersinSub-SaharanAfricaandSouthAsia.Agric.Ecosyst. Environ.187,1–10.

Thierfelder,C.,Mwila,M.,Rusinamhodzi,L.,2013.Conservationagriculturein easternandsouthernprovincesofZambia:long-termeffectsonsoilqualityand maizeproductivity.SoilTillageRes.126,246–258.

VSN,2011.GenStatforWindows, 14thed.VSNInternationalHemel,Hempstead,UK. Vanlauwe,B.,Wendt,J.,Giller,K.,Corbeels,M.,Gerard,B.,Nolte,C.,2014.Afourth principleisrequiredtodefineConservationAgricultureinsub-SaharanAfrica: theappropriateuseoffertilizertoenhancecropproductivity.FieldCropRes. 155,10–13.

Verachtert,E.,Govaerts,B.,Lichter,K.,Sayre,K.,Ceballos-Ramirez,J.,Luna-Guido, M.,Deckers,J.,Dendooven,L.,2009.ShorttermchangesindynamicsofCandN insoilwhencropsarecultivatedonpermanentraisedbeds.PlantSoil320,281– 293.

Verhulst,N.,Francois,I.,Grahmann,K.,Cox,R.,Govaerts,B.,Verhulst,N.,Francois,I., Grahmann,K.,Cox,R.,Govaerts,B.,2014.NitrogenUseEfficiencyand OptimizationofNitrogenFertilizationinConservationAgriculture.CIMMYT, MexicoDF.

Vincent,V.,Thomas,R.,Staples,R.,1960.AnAgriculturalSurveyofSouthern RhodesiaPart1.Agro-ecologicalSurvey.AnAgriculturalSurveyofSouthern Rhodesia.Part1. Agro-ecologicalsurvey.

WRBWorkingGroup,2006.Worldreferencebaseforsoilresources,Aframework forinternationalclassification,correlationandcommunication,2ndedn.World SoilResourcesReports,103.FoodandAgricultureOrganizationoftheUnited Nations,Rome.

ZIMSTAT,2012.Census2012:ProvincialReport(MatabelelandSouth)Zimbabwe PopulationCensus2012.CentralStatisticalOffice,Harare.

ZimVac,2012.Rurallivelihoodsassessment.In:Committee,Z.V.A.(Ed.),Rural LivelihoodsAssessment.SIRDC,Harare.

The author has requested enhancement of the downloaded file. All in-text references underlined in blue are linked to publications on ResearchGate. The author has requested enhancement of the downloaded file. All in-text references underlined in blue are linked to publications on ResearchGate.

Figure

Fig. 2. Cumulative N Mineralisation estimated from the in situ incubation technique (0–30 cm) (a–d) and Net N uptake by maize (above ground parts) (e–h) in the 2013/14 season influenced by soil fertility amendments under plough only (a and e), plough + mulc
Fig. 2. Cumulative N Mineralisation estimated from the in situ incubation technique (0–30 cm) (a–d) and Net N uptake by maize (above ground parts) (e–h) in the 2013/14 season influenced by soil fertility amendments under plough only (a and e), plough + mulc p.4

References

Updating...