BASE Biotechnol. Agron. Soc. Environ.201620(S1),257-272
Indicatorsofphosphorusstatusinsoils:significanceandrelevanceforcropsoilsinsouthernBelgium.AreviewMalorieRenneson,SophieBarbieux,GillesColinetUniversityofLiège-GemblouxAgro-BioTech.TERRA.AgricultureIsLife.PassagedesDéportés,2.BE-5030Gembloux(Belgium).UniversityofLiège-GemblouxAgro-BioTech.BiosystemsEngineering.Water-Soil-PlantExchanges.PassagedesDéportés,2.BE-5030Gembloux(Belgium).E-mail:[email protected]
ReceivedonApril3,2015;acceptedonFebruary19,2016.
Introduction.Phosphorus (P) isanessentialelement forplantgrowth.Therefore, it isessential toaccuratelyevaluate itscontentinthesoil.ThisrequiresreliableindicatorsofsoilPstatus.Literature.ThispaperreviewsliteratureregardingtheindicatorsofPstatusinsoils.Manyindicatorscanbefound,includingsingle extractions (soluble, available, or total P), which are the most common indicators used worldwide. Over time,increasinglycomplexPindicatorshavebeendevelopedassequentialextractionswhichcharacterizethevariousformsofP,degreeofPsaturation,diffusivegradientsinthinfilms,biologicalextractions,isotopicmethods,ormorecomplexmodels.Tomakeachoiceamongthem,differentcriteriashouldbeapplied,includingrelevance,costandtime,easeofinterpretation,and,mostimportantly,theobjectiveoftheanalysis.Itisalsonecessarytoanalyzetheappropriatenesstosoilandclimate.Firstly,thispaperdescribesthevarioustypesofindicatorspresentintheliterature,andproposesaclassificationsystem.Secondly,allcitedindicatorsareevaluatedandcompared.Finally,thePindicatorsmetinWallonia,southernBelgiumarediscussed.Conclusions. Each P indicator presents advantages and disadvantages. This review highlights the importance of carefulconsiderationofindicatorchoice,andtheestablishmentofinterpretationthresholds.Keywords. Soil,phosphorus,indicators,fertility,soilanalysis,Belgium.
Indicateurs de l’état du phosphore : signification et pertinence dans les sols agricoles en Wallonie (synthèse bibliographique)Introduction.Lephosphore(P)estunélémentessentielpour lacroissancedesplantesdont ilestnécessairedeconnaitreprécisémentlateneurdanslesol.Celapasseinévitablementparl’utilisationd’indicateursadéquats.Littérature.CetarticleprésenteunerevuebibliographiquedesindicateursduPcontenudanslessols.Dansla littérature,il existe unemultituded’indicateurs de l’état duPdans le sol.On retrouvenotamment différentes extractions chimiques(Psoluble,disponibleoutotal)quisontlesindicateurslesplusutilisésdanslemonde.Aufildutemps,desméthodespluscomplexesontétédéveloppéestelleslesextractionsséquentiellesquicaractérisentlesdifférentesformesduPdanslesol,letauxdesaturationenP,lesrésineséchangeusesd’anions,lesextractionsbiologiques,lesméthodesisotopiquesoucertainsmodèlespluscomplexes.Pourfaireunchoixparmicesdifférentsindicateurs,différentscritèresdoiventêtreprisencomptedontlapertinence,lecoutetletempsd’analyse,lafacilitéd’interprétationmaissurtoutl’objectifdel’indicateur.Ilestégalementnécessaired’évaluerl’adéquationaveclesoletleclimat.Premièrement,lesdifférentstypesd’indicateursduPretrouvésdanslalittératureontétédécritsetuneclassificationaétéproposée.Ensuite,lesdifférentsindicateursontétéévaluésetcomparésentreeux.Enfin,lasituationenRégionwallonneaétédiscutée.Conclusions.TouslesindicateursduPprésententdesavantagesmaiségalementdesinconvénients.Cetterevuebibliographiquesoulignel’importancedebienréfléchirauchoixdesindicateursetdedisposerdeseuilsd’interprétationcorrespondant.Mots-clés.Sol,phosphore,indicateurs,fertilité,analysedesol,Belgique.
1. INTRODUCTION
Phosphorus(P)isanessentialnutrientforplantgrowth,and is therefore a critical part of the fertilizationrequirementsforcropproduction.However,excessive
bioavailablePinputscanleadtotheeutrophicationofsurfacewaters, which represents amajor concern intheworld.Increasingenvironmentalissuesandriseinfertilizerpriceshaveledtothereconsiderationofcertainagriculturalpractices.AccordingtotheNationalUnion
258 Biotechnol. Agron. Soc. Environ. 201620(S1),257-272 RennesonM.,BarbieuxS.&ColinetG.
of French Fertilizer Industries (UNIFA; www.unifa.fr), the consumption of phosphate mineral fertilizerhasdecreasedbynearly80%over thepast 30years.ThisdeclinethreatenstheavailabilityofPinsoilsoverthelongterm,andincreasestheriskofPdeficiencyinthemostsensitivesoils.Thus,theidentificationofsoildeficiencyrisk,concomitantwiththeminimizationofenvironmentallossesbyerosionandrun-off,requiresthe development of relevant indicators ofP status insoils.
Insoil,totalPcanbeseparatedintodifferentpoolsoforganicandinorganicfractions.InorganicPincludesprimary P minerals (apatite, strengite, variscite);secondaryPminerals(Ca,Mg,FeorAlphosphates);Padsorbedontotheedgesofclayminerals;Pboundtoorganicmatterthroughmetalliccations;anddissolvedP(H2PO4
-,HPO42-,PO4
3-).DissolvedPrepresentsmorethan 96% of the P taken up by plants according toBecketal.(1994)andcanbeconsideredasthePformdirectly available to plants, the quantity dependingon the time of ions exchange (Fardeau, 1993). Themobility and bioavailability are controlled by thelow levels of dissolved P in soil solution, which ismainly governed by high rates of adsorption and/orprecipitation of dissolved P with positively chargedsoilcompounds,includingmetalcations(Ca,Mg,Fe,Al)andFe-,Al-oxyhydroxide(Hinsinger,2001).So,Pionsconcentrationinsoilsolutionisthereforelargelycontrolledbycationicactivity insoil solution,whichis influenced by soil pH and environmental factorsas redox potential andmineral solubility (Pierzynskietal.,2005).However,Pestimatedavailableforplantscan highly differ from P really taken off by cropsbecauseitdependsonplantspeciesandsoilconditions.
Organic fraction is defined as P bound with C(organic matter and biological compounds such asDNAandphospholipids)(Condronetal.,2005).Theproportion of this fraction depends on factors suchas land use and pedo-climatic conditions, and canvary from 25-30% to 75-80% of total P (Fardeauetal.,1994).SoilPflowsoccurbetweenorganicandinorganicpoolsviaimmobilizationandmineralizationprocesses mediated largely by soil microorganismactivity (Oberson et al., 1996). Soil conditions,includingsoilmoisture,temperature,pH,andsurfacechemical properties are integral factors promotingthesereactions.
Duetotheprofusionofexistingindicators,itcanbedifficulttoidentifythemostappropriateindicatorforagivensetofconditions.Themanagementofagivenenvironmentrequires indicatorswhichareadaptedtolocalconditions,indicatorswhichareeasytomeasure,andindicatorswithknownthresholdsandlimitations.ItisnotthepurposeofthisarticletoprovideanexhaustivelistofexistingmethodsthatcurrentlyservetoindicateP status in soil. Rather, the goal of this study is to
evaluatethemostcommontypesofmethodsinorderto highlight the advantages and limitations of each.ThisevaluationwasalsoappliedmorespecificallyinWallonia,aregionsituatedinsouthernBelgium,thoughresultscanbeextrapolatedtofacilitatebothagronomicandenvironmentalmanagementofPinotherregions,providedthatthespecificedaphicpropertiesofagivenregionareconsidered.
2. MAIN METHOD CLASSES FOR P CHARACTERIZATION IN SOIL
Manydiagnostictoolshavebeendevelopedtoevaluatesoilfertility.Accordingtotheliterature,theavailabletoolsarebasedonthefollowingvariables:–theobjective(agronomic,environmental,orboth);–the scale of the study (e.g., a cultivated field orwatershed),or
–themeasurementprinciple.
Theseclassificationsaredescribedinthefollowingsectionsandinfigure 1.
2.1. Chemical extractions
Various chemical methods have been developed toestimatethecapacityofasoiltoprovidethelevelsofPnecessaryforplantgrowth.Currently,severaldozenmethods exist, with varying degrees of complexity.Mostofthesemethodsweredevelopedunderspecificconditions(soilorcultivationsystem),andwerefixedwithrespecttointerpretationreferences.Consequently,it is difficult to apply a uniformmethodworldwide,despite regulatory efforts within Europe for theestablishmentofastandardizedsetofmethods(Proix,2013).Therefore,definingspecificthresholdsbecomesimportant.
Single extractions of P content. Main existingindicators estimate the levels of available (orexchangeable)Pbasedon“thesumofPimmediatelyavailabletoplantsandofPthatcanbeconvertedintoanavailableformthroughphysical(desorption),chemical(dissolution), or biological (enzymatic degradation)processes in nature during a growing season”(Boström et al., 1988). Some analytical methodsutilizeextractantstomimictheactionofrootsthroughdissolution,desorption,orchelationreactions.Of theseveraldozenchemicalextractionmethodsreportedinliterature,thesedifferdependingontheextractantused,theextractiontime,andtheratioofsoiltoextractant.Someextractionmethodsarealsospecifictoacountryor region, while other methods are more universal(e.g., themethodsofOlsen,Mehlich3, orBray; seetable 1). Within the same country, several methods
Indicatorsofphosphorusinsoils 259
canalsocoexist.ThisisparticularlyevidentinFrance,where levels of available P are commonly evaluatedaccordingtothemethodsofOlsen,Joret-Hébert,andDyer(Table 1),explainedpartlybydiversityofsoils.
Rather than providing an exhaustive survey ofthe existing methods, the following discussion willbe limited to the most commonly cited methods inthe literature (Table 1). Proix (2013) proposed thatextractionmethodscanbeclassifiedintofourtypes(I–
IV),dependingonwhichsoilbindingmodeisaffectedbytheextractionperformed(Table 2).
TypeI analysismethods that detect soluble P arenotcommonlyusedasagronomicalindicators.Instead,these are routinely used as environmental indicatorsin somecountries.To formulate the riskofP loss,Pis extracted with distilled water (Sissingh, 1971).This method is used in The Netherlands, Austria,and Switzerland, among others; it estimates the
Figure 1. Classificationofthedifferentindicatorsofphosphorusstatusinsoils—Classification des différents indicateurs du phosphore dans les sols.
Diagnostic tools for the status
of soil phosphate
CHARACTERIZATION MODELIZATION
Input data
IsotopesBiological extractions
Phosphorus index for risk of P loss (PI)Phosphorus export
diagnostic tool (ODEP)
Various modelsModelisation
of phosphorusdynamic transfer
to the soil solution
Predictivemodeling
based on the function of Freundlich
Kineticisotope
exchange
Compartmentalrepresentationof P availability
Sequentialextractions
HedleyChang & Jackson
etc.
Degree ofphosphorussaturation
Singleextraction ofphosphorus
content
Water extractionOlsen
AA-EDTAAcetate-Lactate
Joret-HébertBray
Mehlich 3, etc.
Oxalate Mehlich 3
Chemical extractions
Anion exchangeresins
Table 2. ClassificationofthechemicalextractionsproposedbyProix(2013)—Classification des méthodes d’extraction chimique proposée par Proix (2013).
State of P in the soil How it works Extraction methodsTypeI Soluble Dissolutionofsolubleelementsis
performedandthisisusedasthesoilsolution
Water,CaCl2
TypeII Linkedtoclay-humiccomplexes
Pismainlyreleasedfromclayandorganicmatter
Neutralsalts,bufferingeffect(Olsen,etc.)
TypeIII Adsorbedorprecipitatedonoxyhydroxides
Pisreleasedfromironandaluminiumoxyhydroxides
Neutralsalts,bufferingeffects,chelatingmolecules(AA-EDTA,acetate-lactate,etc.)
TypeIV Totalorpseudo-totalstock TotalPconcentrationisdeterminedinordertoperformapedologicalcharacterizationofagivensoil
Triacid,Aquaregia
260 Biotechnol. Agron. Soc. Environ. 201620(S1),257-272 RennesonM.,BarbieuxS.&ColinetG.Ta
ble 1
.Characteristicsofsingleextractionsofphosphoruscontentmostcom
monlycitedintheliterature(com
pletedfrom
Jordan-M
eilleetal.,2012)—
Car
acté
rist
ique
s de
s pri
ncip
ales
mét
hode
s chi
miq
ues d
’ext
ract
ion
du p
hosp
hore
renc
ontr
ées d
ans l
a lit
téra
ture
(com
plét
é de
Jor
dan-
Mei
lle e
t al.,
201
2).
Met
hod
Che
mic
al e
xtra
ctan
tsEx
trac
tion
time
and
ratio
of s
oil
to e
xtra
ctan
t *
Cou
ntri
esSo
il ty
pes
Indi
cato
r ty
pes
Ref
eren
ces
Ammonium
acetateand
EDTA
0.5Mammoniumacetate+0.5M
aceticacid+0.025MEDTA
,pH4.65
1:10(w:v),
30minsh
aking
Belgium
(Wallonia),
Switzerland
Acidictoneutralsoils
ALakanenetal.,
1971
Acetatelactate
0.1Mammoniumlactate+0.4M
aceticacid,pH3.75
1:20(w
:v),
2hshaking
Belgium
(Flanders),
Sweden,N
orway
Allsoils
AEgneretal.,
1960
Bray1
0.03M
NH
4F+0.025M
HCl,pH
3.75
1:10(w
:v),
5minsh
aking
USA
,Spain,Italy,
Greece,India
Acidicsoilsrichiniron,not
inneutraltocalcareousso
ils,
notinclayeyso
ilswithhigh
degreeofexchangeablebases
ABrayetal.,
1945
CaCl 2Soluble
0.01MCaCl 2
1:10(w
:v),
2hshaking
TheNetherlands
A,E
Houbaetal.,
2000
CAL
0.05M
lactatedeCa+0.05M
calcium
acetate+0.3Maceticacid,pH4.1
1:20(w
:v),
2hshaking
Germany
Allsoils
ASchüller,1969
Dyer
Citricacid2%
,pH2
1:5(w:v),
2hshaking
France,B
elgium
(forest
soils)
Acidictoneutralsoils
ADyer,1894
NF31-160,
1999
Joret-H
ébert
0.2Moxalated’NH
4,pH
71:20(w:v),
2hshaking
France
Neutraltocalcareousso
ilsA
Joretetal.,
1955
NFX31-161,
1999
Mehlich1
0.0125M
H2SO
4+0.05MHCl
1:4(w:v),
5minsh
aking
USA
Acidictoneutralsoils
AMehlich,1953
Mehlich3
0.015MNH
4F+0.2M
CH
3COOH+
0.25M
NH
4NO
3+0.013M
HNO
3+
0.001MEDTA
1:10(w
:v),
5minsh
aking
Canada,USA
Allsoils
AMehlich,1984
Morgan
0.72M
sodium
acetate+0.52Macetic
acid,pH4.8
1:5(v:v),
15minsh
aking
USA
AMorgan,1941
Morgan
modified
0.62M
NH
4OH+1.25MCH
3COOH
1:5(v:v),
15minsh
aking
Ireland,USA
AcidicsoilswithaCEC
below
20cmol
+ . kg-
1A
McIntosh,1969
Olsen
0.5MNaH
CO
3,pH
8.5
1:20(w
:v),
30minsh
aking,
at20°C
France,England,U
SA,
Spain,Tunisia,South
Africa,Greece,India,
Iran,etc.
Calcareoussoilsextendedto
allsoilsexceptedhighlyacidic
soils
AOlsenetal.,
1954
ISO
11263:1994
./..
Indicatorsofphosphorusinsoils 261
amount of P present in the soil solution, similar tothe extraction of Pwith 0.01MCaCl2. The latter isgenerallycharacterizedbyvalueswhicharetypically2to3timessmallerthantheformer.Generally,calciumisthefirstcationtoformcomplexesinsoil,andthus,0.01M CaCl2 typically represents the average ionicstrengthofasoilsolution(Houbaetal.,2000).LevelsofsolublePmayalsovaryaccordingtoseason,withlevelsof extracted solublePbeinghigherduring thewetterperiodsthanthedryerperiods,whichisduetotheextentofmineralisation.
TypeII and III indicators correspond with thelevels of available (or exchangeable) P. These areagricultural indicators which are commonly usedworldwide.Themethodsvary,andcanbecustomizedfordifferentactionprocesses.Forexample,theOlsenmethod extracts P by ligand exchange, whereas theBraymethod extracts P by forming complexes withcalciumoraluminiumphosphates(Honsetal.,1990).Therefore,somemethodsaremoresuitableforacidicsoils, while others are more suitable for calcareoussoils(Table 1).
The properties of a soil can influence the resultsobtained from chemical extraction methods, inparticular, clay, organic matter, or pH (Tran et al.,1985). Moreover, the presence of carbonates caninfluencetheextractioncapacityoftheBraymethod,andthepresenceofclaycandisruptthefiltrationphaseof awater extraction.Extractionmethodswhich useammoniumacetateandethylenediaminetetraaceticacid(EDTA)arealsoaffectedbypH(Lakanenetal.,1971).FewcorrelationswithyieldhavebeenobservedfortheextractionofPviathelattermethodincalcareoussoils,sincethemechanismsofPextractionarenotsuitableinthesesoils.Incontrast,thesedifferentpropertieshaveonlyanegligibleinfluenceontheuptakeofPbyplants(Beaudin,2006).
Although it is difficult to compare the variousmethodsofPextraction,manyauthorshaveattemptedto do so (Homsy, 1992; Pote et al., 1999; Hoodaet al., 2000; Maguire et al., 2002; Neyroud et al.,2003). Consequently, some authors have proposed aclassification of extractants based on the amount ofPthatisextracted(Honsetal.,1990;Neyroudetal.,2003).However,theseclassificationsarenotuniversalforalltypesofsoil.Thus,itisessentialtoaccountforthespecificsofananalyticalmethodpriortoselectingitasanindicator.Inaddition,itisoccasionallydifficulttochooseoneuniquemethodduetodifferencesintheedaphicpropertiesofacountryorregion.Amorerobustmethodcanpotentiallybeappliedtoawiderangeofsoils, such asMehlich3 in Quebec (Beaudin, 2006)or theOlsenmethod. Interestingly, the lattermethodwasdevelopedincalcareoussoils,whiletheresultsofthismethodology are also satisfactory in acidic soils(Moreletal.,2000).Ta
ble
1 (c
ontin
ued)
.Characteristicsofsingleextractionsofphosphoruscontentm
ostcommonlycitedintheliterature(completedfromJordan-Meilleetal.,
2012)—
Car
acté
rist
ique
s des
pri
ncip
ales
mét
hode
s chi
miq
ues d
’ext
ract
ion
du p
hosp
hore
renc
ontr
ées d
ans l
a lit
téra
ture
(com
plét
é de
Jor
dan-
Mei
lle e
t al.,
201
2).
Met
hod
Che
mic
al e
xtra
ctan
tsEx
trac
tion
time
and
ratio
of s
oil t
o ex
trac
tant
*
Cou
ntri
esSo
il ty
pes
Indi
cato
r ty
pes
Ref
eren
ces
Triacid
HCl+HF+HClO
4Hotextraction
Allsoils
TNFX31-147,
1996
Truog
0.002MH
2SO
4+(N
H4)2SO
4,pH
3.0
1:200(w:v),
30minsh
aking
Neutraltocalcareousso
ilsA
Truog,1930
WaterSoluble
H2O
1:60(w
:v),
22hincubation
and1hshaking
TheNetherlands,
Germany
Allsoils
A,E
Sissingh,1971
1:10(w
:v),
5minsh
aking
Allsoils
Pansuetal.,
2003
*w:v:w
eight/volum
eratio—
rapp
ort p
oids
/vol
ume;v:v:volum
e/volumeratio—
rapp
ort v
olum
e/vo
lum
e;indicatortypes—
type
s d’in
dica
teur
s:A:agronom
ic—
agr
onom
ique,E:
environm
ental—
env
ironn
emen
tal,T:total—
tota
l.
262 Biotechnol. Agron. Soc. Environ. 201620(S1),257-272 RennesonM.,BarbieuxS.&ColinetG.
Finally, total P content (typeIV) is occasionallyused as an indicator of theP reserve present in soil.VariousacidmethodsexistandconsistofusingacidssuchasHF,HClO4,HNO3,HCl,orH2SO4separatelyorincombination,withvariableefficiency.ThisindicatormaybeusefulinevaluatingthePcontentofsedimentlost by erosion, or in assessing the pedologicalcharacterization of a given soil. Recently, total PcontentwasusedtoestimatePreservesboundtoparentmaterialsinWallonia(Rennesonetal.,2013).
Sequential extractions. Complementary to theapproachofestimatingavailableP,someauthorshavedevelopedsequentialextractionmethodstoidentifyPpoolsofvaryingsolubilityinsoil.Differentextractionmethodsarecombinedsequentiallyinordertodepletea soil of its content from decreasingly availablefractions.ResidualPisdeterminedbythedifferenceofPformscomparedtototalP.
Manyprotocolsforsequentialextractionhavebeenproposed (Table 3), and two major types have beendistinguished.Onemajor type involves classificationon the basis of the type of bonding,while the otherinvolves classifications on the basis of the degreeof availability to plants. The most commonly usedprotocols were originally developed by Chang et al.(1957)andHedleyetal.(1982).TheoriginalpurposeoftheformerworkwastodistinguishdifferentformsofPdependingonthetypeofbondingtosoilminerals(e.g.,Pboundtoironvsaluminiumvscalcium,etc.).However, the selectivity of the extractant in thisprotocolwasoftenquestioned.Fractionationschemesareunabletoisolatediscretemineral,astheysolubilizegroupsofmineralsusuallydefinedasPassociatedwithAl,Fe,Ca,orresidualforms(Pierzynskietal.,2005).Nevertheless, thisprotocol ledto thedevelopmentofasecondtypeofprotocols.Inparticular,Hedleyetal.(1982)incorporatedtheconsiderationofPavailability
for plants, without specifying the forms of bondingwithin each fraction, thereby assuming the samechemicalformispresentindifferentfractions.
Manystudieshaveattemptedtocompareavarietyofextractionmethodsusingthesamesetofsoilsamples(Levyetal.,1999;Taoufiketal.,2004).However,theresultsaredifficult togeneralizedue to the influenceofsoiltypes.Williamsetal.(1967)demonstratedthatthemethod outlined byChang et al. (1957)was notapplicableforcalcareoussoilsorsediments.Accordingto Tiessen et al. (1993), the fractionation of Hedleyet al. (1982) is the only method that can be usedwithmoderatesuccessfortheevaluationofavailableorganicP.Inacidsoils,somepoolsarenotcompletelyseparated (bicarbonate and hydroxide-extractable Pi)and represent a continuum of Fe- andAl-associatedP extractable (Tiessen et al., 1993). The Hedleyfractionation is also used in tropical soils, followingprotocol modifications (use of resin P fraction andgenerallylessPfractions)(Negassaetal.,2009).
Overall,whilethesemethodsprovideanextensivecharacterization of the different forms of P presentin a soil, they are expensive and time consuming toimplement,therebypreventingtheirroutineuse.
Degree of P saturation.Giventheimportanceoftheenvironmental issues related to P, and the influenceof edaphic properties on the results of chemicalextractions,anenvironmentalindicatorwhichaccountsforthedistinguishingcharacteristicsofaparticularsoilwas developed. This indicator evaluates the degreeof soil P saturation by measuring the proportion ofpotential binding sites in soil which are actuallyoccupiedbyP,withmainbindingsites involving theoxide and hydroxide groups of iron and aluminium(vanderZeeetal.,1988).Thus,theindicatoraccountsforboth thebindingcapacityand thefixedPcontentofthesoil.Leinweberetal.(1999)havedemonstrated
Table 3. SequentialP extractionsmost commonlycited in the literature—Principales méthodes de fractionnement du phosphore retrouvées dans la littérature.Method Chemical extractants ReferenceChang&Jackson NH4Cl,NH4F,NaOH,H2SO4,Na2S2O4-citrate Changetal.,1957Williamsetal. NH4Cl,NH4F,NaOH,Na2S2O4-citrate,NaOH,HCl Williamsetal.,1967Hieltjes&Lijklema NH4Cl,NaOH,HCl Hieltjesetal.,1980Hedley Anionexchangeresins,NaHCO3,NaOH,dissolvedHCl,concentratedHCl,
H2SO4
Hedleyetal.,1982
Bozongoetal. HydrogenperoxideandmethodofChang&Jackson Bozongoetal.,1989Ruttenberg MgCl2,CDB,acetate,HCl,hotHCl Ruttenberg,1992Paludan&Jensen H2O,bicarbonate-dithionite,NaOH,HCl Paludanetal.,1995Golterman Distilledwater,Ca-EDTA/dithionite,Na-EDTA,H2SO4,H2SO4inautoclave Golterman,1995Rydin&Welch NH4Cl,Na2S2O4/NaHCO2,NaOH,HCl Rydinetal.,1998
Indicatorsofphosphorusinsoils 263
thatthisparametercanbecalculatedaccordingtothefollowingequation:
Degree of P saturation = Poxα Alox +Feox( )
×100 (Eq .1)
where the degree of P saturation is expressed asa percentage, and Pox, Alox, and Feox represent theamounts of P, aluminium (Al), and iron (Fe) thatare extracted with ammonium oxalate (mmol.kg-1),respectively, while α represents a scaling factor(generally0.5).
Thismethodwasoriginallydevelopedintheacidicsandy soils ofTheNetherlands, and since has beenappliedtootherregionsandcountries(e.g.,Flanders[northern Belgium], Quebec, etc.). In Quebec, thisindicatorhasbeendefinedastheratiobetweenPandAl,asdeterminedbytheMehlich3extraction(Khiariet al., 2000). In 1999, Beauchemin et al. identifiedthevarious formulas thathadbeenused tocalculatethedegreeofPsaturationintheliterature.InastudyconductedinWallonia(Rennesonetal.,2015),itwasdemonstratedthattheequationhasfirsttobeadaptedto thepedologicalandgeologicalcontextof thesoilunderinvestigation.
Moreover, it was shown that the degree of Psaturation is correlatedwith the concentrations of Ppresentinrun-off(Poteetal.,1999)andindrainagewater(Leinweberetal.,1999),therebyindicatingthatit represents a relevant environmental indicator. InTheNetherlandsandinFlanders,thisindexhasbeenincorporated into legislation,whereby adegreeofPsaturationthatexceeds25%isdefinedasunacceptabledue to the riskofP transfer to the soil solution andwater(Breeuwsmaetal.,1995).
Agronomic and environmental thresholds. To berelevant as an agronomic indicator, the amount ofextractablePmustbecloselyrelatedtocropresponse,suchasplantgrowthoruptakeofP (Figure 2).Thecorrelationbetweensoilandplantscanbeconductedingreenhouse,agrowthchamber,oronthefield.Testcalibrationsmustbeconductedoverabroadrangeofsoils in order to define fertility classes according tosoil properties, such as texture or pHvalues (Genotetal.,2011;Jordan-Meilleetal.,2012).
Similarly,environmentalthresholdcanbedefinedifextractablePiscorrelatedwithPlostbyrun-offandleaching.Thethresholdtypicallycorrespondsto:–the content of P which is tolerated in water uponlegislation,or
–the“changepoint”ofthecurverelatingsoillosstosoilPcontent,whichisthepointofthecurvewheretheslopeincreases(Poteetal.,1999)(Figure 2).
Generally, environmental threshold is higher thanagronomicthreshold,providingacontrolleverforthemanagementofP.
2.2. Anion exchange resins and diffusive gradients in thin films (DGT)
Anionexchangeresinsweredevelopedforwaterandsedimentsamplesinthe1930stoassessthepresenceof labile P in soil samples or the presence of P insoil solutions.Upon contactwithwater and soil, theanionexchangeresinsactasasinkforP.Sincethen,analytical protocols designed tomimic the effects ofrootshaveevolved(Qianetal.,2002).
In the1990s, theuseofDGTwaspreferredoveranionexchangeresinsforestimatesofPavailabilityinsoilstoobviatethedisadvantagesofthelatter(Chardonetal.,1996).Thesedisadvantagesincluded:–modificationof thephysicochemicalbalanceof thesoilexamined;
–theabsenceofan infinitebindingcapacity, therebyresultinginthepotentialfornon-maximaldesorption;
–difficultyinseparatingtheresinfromthesoil;–lackofspecificityintheadsorptionanddesorptionofcertainanions;
–theinfluenceofsulphateornitrateconcentrationsonthequantitiesofPextracted.
Diffusivegradients inthinfilms(DGT)areapassivesampling technique which has been successfullyapplied to aquatic systems for measuring P (Zhanget al., 1998), and more recently, for predicting cropresponse to applied P in soil (Mason et al., 2010).Basedonthesameprincipleasanionexchangeresin,DGTiscomposedofalayerofferrihydritebindinggelwithastrongaffinityforPbehindadiffusivehydrogellayer and an overlying protective filter membrane(Six et al., 2012). Diffusive gradients in thin films
Figure 2. Representationof agronomicandenvironmentalthresholds determination — Représentation de la détermination des seuils agronomique et environnemental.
P lo
st b
y ru
noff
Cro
p y
ield
s
Low Medium High
Environmental classes
Agronomicthreshold
Environmentalthreshold
264 Biotechnol. Agron. Soc. Environ. 201620(S1),257-272 RennesonM.,BarbieuxS.&ColinetG.
can be placed directly onto a saturated soil paste,allowingforfieldmeasurements.Diffusivegradientsin thin films provide a better estimate of P uptakeby plants compared to certain chemical extractionmethods(Zhangetal.,2014),andarelessvulnerableto potential chemical constraints, such as anionicinterferencesorpH(Masonetal.,2008).Moreover,alowcoefficientofdeterminationfortheregressionfitbetweenDGTandresinmeasurementwasobservedbyMasonetal.(2008).Despitetheresultsappearinghopeful for tropical soils, additional studies arenecessary to expand the types of plants, soil, andclimatic conditions that can be tested (especiallyin European soils), and to correctly interpret DGT-derivedresults(Zhangetal.,2014).
2.3. Biological extractions
To overcome the disadvantages of chemicalextractants,ithasbeenproposedthatplantscouldbeusedtoevaluatethebioavailabilityofP.
In bioassays, algae are grown inwater inwhichP is a limiting factor, for which the only source isthesoilsample.Unicellularalgaearegenerallyusedin these tests, includingSelesnatrum capricornutumor Scenedesmus quadricauda. The experiments aregenerally performed aerobically for 2 to 4weeks,and the algae are renewed weekly (Ekholm et al.,2003). Phosphorus availability is calculated onthe basis of the algae biomass present, and resultscan be extrapolated for longer periods of time todetermine the long-term availability of P. Severalstudies,includingthatofBoströmetal.(1988),haveshownthattheresultsobtainedusingsuchbioassaysare consistent with results obtained using chemicalextractions. However, bioassay results are onlyrelevant for the experimental conditions tested, andcannot be extrapolated to the natural environment.Furthermore, numerous species of algae exist innaturalenvironments,and theycanadapt tovariousformsofP.
Other biological methods have been used toestimatethecapacityofasoiltosupplyPtoaplant,including theuseofmicro-cultures (Stanford et al.,1957). In these cultures, plant growth is used toanalyze the amount of absorbedP.Unlike chemicalmethods,thistechniquemoreaccuratelyaccountsforallofthefactorsaffectingplantfood.
However,bioassaysaredifficulttoestablish,moretime-consumingtoperform,andtheexperimentsmustberepeatedtoachievetherepeatabilityandaccuracyofresults.Therefore,thesebioassaysdonotrepresenta substitute for chemical extraction methods, butshouldratherbeamethodperformedtocomplementchemicalextractionmethods.
2.4. Isotopic methods
TimeplaysasignificantroleindeterminingPavailability(Fardeau, 1993). Methods involving radioactiveisotopes of P were developed in France by Fardeau(1993),andmorerecentlybyMoreletal.(2000)andMoreletal.(2014).Isotopes,unlikechemicalreactions,enabletheobservationofsoilbehaviorwithregardtoPwithoutaffectingthebalancebetweentheformsofPpresentinsoil.Themainisotopesused,32Pand33P,haveahalf-lifeof14.3daysand23days,respectively.
Ingeneral,isotopicmethodsconsistofaninjectionwithadefinedamountofradioactivityintoasolution,followed by subsequent measurements of theradioactivityremaininginthesolutionafteradefinedperiodoftime.TheabilitytodetectPisbasedonthreeprinciples:–the concentration of phosphate ions in solution isconstantovertime;
–the isotopic tracer is instantly and uniformlydistributedinthesoilsolution;
–the flow of labelled P ions is equal to the flow ofunlabelledPions(Némery,2003).
The kinetics of the isotopic exchange methodalso enables the development of a compartmentalrepresentation of soil P reserves, as shown infigure 3 (Fardeau,1993).Thisdiagramillustrates theheterogeneityofphosphateionsthathavethepotentialto reach the soil solution over variable periods oftime.Moreover,thisrepresentationcanalsobelinkedto chemicalmethodsused todetermine the availableand total reservesofP (Figure 3).Theshapesof theextracted P vary depending on the extractant used(Figure 3).
Additional studies have demonstrated a range ofpotential applications for P isotopes. For example,in 2000, Morel used isotopic methods to model thedynamics ofP ion transfer between soil and the soilsolutionasafunctionofthedurationofPtransferandtheconcentrationofPinsolution.Morerecently,Moreletal. (2014)determined the relationshipbetween thephosphatebalancesheetandPexchangesusinglong-term test parcel data. Finally, P isotopes have beenusedinthestudyofplantgrowthandtomeasuretheefficiency of plant growth following the addition ofisotope-labelledfertilizers(Frossardetal.,1996).
However, there are disadvantages associatedwith the use of isotopes. In addition to the lack ofinformation regarding the organic fraction of P, thismethodcannotberoutinelyused,andthemanipulationof radioelements is extremely delicate. Furthermore,it is difficult togeneralize the results obtainedunderexperimentalconditionstoeventsoccurringnaturallyin soils. To address the latter point, scientists haveattemptedtorelatetheparametersofkineticequations
Indicatorsofphosphorusinsoils 265
to the physicochemical characteristics of soil (Moreletal.,2014).
2.5. Complex models
P indices for risk of P loss. Knowledge of a soil’sstatus is not sufficient for estimating the risk of Pexporttosurfaceorsubsurfacewatersincethelossesthatoccurareinfluencedbybothsourceandtransportfactors.IndicesforriskofPlossrepresentmanagementtoolsatthescaleofaparcelwhichareusedtoidentifycritical source areas of P loss and farming practicesthat increase the risk.Critical source areas of P losscanbedefinedby theircoincidingsource (soil,crop,andmanagement inducinghighP loss) and transport(runoff, erosion, and proximity to water course orbody)factors(Sharpleyetal.,2014).ThisindicewasdevelopedtosolveproblemsoflocalizedexcessofP.
Source factors represent theamountofP thatcanpotentially be mobilized, as well as the conditionspredisposingit toaccumulation(e.g.,soil testP,rate,method,timingofapplicationofbiologicalormineralfertilizers,andleachingofPfromplantresidues),whiletransportfactorsareessentiallyhydrologicalinnatureand affect the transport of P to rivers (e.g., erosion,surface run-off, subsurface drainage, connectivity).Phosphorus indices (PI) are simple models derivedfrom the results of more complex experiments ormodels(Buczkoetal.,2007).
AnumberofP indices,allofwhicharebasedontheoriginalPIofLemunyonetal.(1993)(Equation2),have been developed according to the regionalcharacteristicsofagivenstateand/orcountry.SeveralP indices use the Pennsylvania PI. Currently, the PIapproach is routinely used in 47U.S. states, someCanadian provinces, and was adopted by severalEuropeancountries,includingFinland(2001),Ireland
(2003), Sweden (2005), Norway (2005), Denmark(2006),andGermany(currentlyinprogress).
Site vulnerability = 1.5*soil erosion + 1.5*irrigationerosion + 0.5*runoff class + 1*soil p test +0.75*P fertilizer application rate + 0.5*P fertilizerapplicationmethod+1*organicP source applicationrate+1*organicPsourceapplicationmethod(Eq.2)
wherethevaluesforeachcharacteristicarespecifiedinLemunyonetal.(1993),accordingtolevel.
TheprimaryadvantageofusingaPIisinitsspeedand ease of use. However, some parameters, such asmode of farm management, are difficult to map. Inaddition, a PI can accommodate correctivemeasures.Therefore,aPIisnotonlyanindicatorthatcanprovideanintegrativeapproach,butitcanalsobeeasilyadaptedtolocalconditions(Buczkoetal.,2007).Correspondingly,Pindiceshavebecomehighlypopulartoolsfrombothscientificandpoliticalstandpoints.Phosphorusindicesrequirereadilyavailabledata,andtheycanbeautomatedusingcomputersoftwarewhichlinksthemtoadatabaseofinterest.
Modelling of P loss.Hydrologicalmodelsofincreasingcomplexityhavebeengeneratedtoquantifyvolumesofrun-offwater,aswellassedimentlossandPloss,whilealsoaccountingforlocalspecificities(e.g.,connectivityofparcelswithariverandtopography,amongothers).Withdifferentmodelsavailable,itmaybecomedifficulttoselectthemostappropriatemodel.Itisimportanttoconsiderthescaleofstudyforeachofthemodelsbeingconsidered,aswellastheirdisadvantages.Hydrologicalmodels are very time-consuming and difficult toimplement. In particular, they require extensivecalibration,andare thereforenot typicallysuitableforroutineuse.Moreover,thesemodelscanoftenprovetobe
Figure 3.Compartmental representationof the available formsof phosphorus (P) extracted according to isotopicmethod(adaptedfromFardeau,1993)—Représentation compartimentale des formes du phosphore (P) disponible extraites selon la méthode isotopique (adapté de Fardeau, 1993).
Liquid phase Solid phase
Ions in a soil
solution = CP
Capacity = PL/CP
Instant ion exchange = PL
Ion exchange between 1 min and 1 day
Ion exchange between 1 day and 3 months
Ion exchange between 3 months and 1 year
Ion exchange over 1 year
266 Biotechnol. Agron. Soc. Environ. 201620(S1),257-272 RennesonM.,BarbieuxS.&ColinetG.
inadequateduetogroundconditionswhicharedifficulttoforeseeandtransferprocesseswhicharenotalwayscomplete.ThesemodelsarebasedonthehypothesisthatPtransferisaresultofrunoffanderosion.However,inpractice,thesituationismorecomplex.
Nevertheless, some regions have developed thesemodelsandusethemtoevaluatetheexportofPfromagricultural parcels. In particular, Quebec employsa Phosphorus Export Diagnostic Tool (ODEP) thatisbasedon theSWATmodel, and is able to integratedataregardingthetopography,soiltype,drainage,andagronomicmanagementofaparcel.Consequently,thissoftwaretoolcanbeusedtoquantifyPlosses,toidentifythe factors responsible for the losses, and to simulatedifferent scenarios of P management. In Wallonia,Dautrebandeetal.(2006)usedtheEPICgridmodeltoestimatePexportfromparcels.
3. EVALUATION OF DIFFERENT INDICATORS
Many indicatorsofP levelsexistworldwide,and it isoftendifficult to select themostappropriate indicator.Moreover,many criteria exist to evaluate indicators,and eachof the existingmethodspresent advantagesanddisadvantages(Table 4).Forthisstudy,indicatorswere evaluated according to the framework of the“Sustainability Assessments of Farming and theEnvironment” (SAFE) hierarchical frameworkdeveloped by Van Cauwenbergh et al. (2007). Thisframework has defined the following six criteria forenvironmentalindicators:–abilitytodiscriminateintimeandspace,–analyticalaccuracy,–costandtimeofanalysis,–easeofinterpretation,–relevancetoregulations,–theabilitytoextrapolatetheresultsobtainedusingagivenindicator.
Themain indicators were evaluated according tothesecriteria(Table 5).
One of the most important considerations is theobjectiveoftheanalysistobeperformed,andtherefore,the ability to discriminate in time and space. Threetypes of indicators exist: agronomic, environmental,and agro-environmental (Table 1). An indicator canbe considered as having an agronomic interest ifthe relationshipwithplant uptakeor yieldshasbeenproven. Similarly, a parameter can be considered asan environmental indicator when it presents a goodrelationship with P transferred to aquatic systems.Environmental indicators generally require mildextractants such aswater or calciumchloride,whichsimulate desorption or solubilisation of P from soilto solution. Naturally, there is a link between these
two objectives.When P in soil solution increases, Pyield can increase, though P loss is also susceptibleto increase. That is why some indicators as thedegreeofPsaturationpresentbothanagronomicandenvironmentalinterest.
Secondly, according to SAFE criteria, indicatorsmustbeeasytoanalyzewithinareasonablecost,andanalysisshouldbeaccurate.Someanalysesare time-consuming and expensive to implement, or requirespecific conditions (Table 5). For example, isotopicmethods, while being extremely accurate can onlybe performed by specialized laboratories. Moreover,although main extraction methods consist of a soil-extract suspension using a ratio which is seldomrepresentative of soil conditions, ratios are oftenmorerepresentativeoftheaquaticenvironment.EachmethodextractsdifferentformsofPandthechemicalextractants often mobilize significant amounts ofunavailableformsalongwithplant-availableP(Frossardetal.,2004).Additionally,thedeterminationofPcanbecomeunreliableinsomesoilsifPconcentrationsintheextractareclosetodetectionlimits,asobservedinwaterorCaCl2extractions.
Thirdly,Tiessenetal.(1993)statedthatavailablePmeasures“apoolofsoilPthatissomehowrelatedtothatportionofsoilPwhichisplantavailable”.Therefore,results should not be interpreted independently, butaccording to regional standards, which are definedaccordingtoexperimentationwithplants.Tothisend,Jordan-Meilleetal.(2012)comparedcurrentmethodsusedforrecommendingPfertilizersinEurope,despitethe large number of analytical methods employed.Twodifferentsoilsweretested,andtherecommendedPdosagevariedbetween0 and89kgP.ha-1 forbothsoils,dependingonthecountryandmethodused.
Lastly, extrapolatingmethods and results toothersoils isnotasimpletask.Indeed,nooneindicator issuitableforallsoils.Forexample,levelsofavailablePcanbedeterminedusing themethodof Joretet al.(1955). However, this method is more suitable forcalcareous soils, whereas the Dyer method providessatisfactory results in soils ranging from acidic toneutral(Table 1).
An ideal indicator should be scientificallyvalidated, be relevant in relation to the stakeholdersandobjectives,discriminateagainstexpectedchanges,haveaninterest/costratiogreaterthan1,bebasedonreadilyavailabledata,andbeconsistentwithexistingregulations. However, in practice, indicators rarelyfulfillallof thesecriteria.Therefore, theselectionofanindicatorisbasedonacompromisebetweenfixedobjectivesandcostinordertocharacterizeacomplexand often problematic phenomenon. Based on thecriteriaestablishedbyVanCauwenberghetal.(2007),itisevidentthatwaterextractionofPwasoneofthemost advantageous indicators available (Table 5).
Indicatorsofphosphorusinsoils 267
In contrast, other methods, such as sequentialfractionation,areoflittletonointerest.
Inpractice, indicatorsareoftencorrelatedtoeachother.PhosphorusavailabilityisdeterminedbytotalPcontentandbufferingcapacity.Currently,theselectionof indicatorsforagivencountryorregionisaresultof historical origin. It is very difficult to introducechangesintothesemethodologies,primarilyduetothenecessityofensuringthestandardizationandcontinuityofdatabases,andtheuseofspecificthresholds.
4. WHICH INDICATORS CAN BE USED IN WALLONIA?
In Wallonia (Belgium), there are no measurescurrently scheduled to assess themanagement statusof P. Incidentally, according to the Organisation forEconomic Co-operation and Development (OECD),Belgium is considered as the first European countryto possess a surplus of P delivered to agriculturalparcels.Thissurplus,amountingto21kgP.ha-1.yr-1,is
Table 4. Advantagesanddisadvantagesofvarioustypesofmethodsofphosphoruscharacterizationinsoil—Avantages et inconvénients des différents types de méthodes de caractérisation du phosphore du sol.Indicators Methods Countries using
these indicatorsAdvantages Disadvantages
Chemicalextractions AvailableP(Olsen,AA-EDTA,Dyer,etc.)
Allcountries Quick,inexpensive,easytouse
Specifictocertaintypesofsoil;specificthresholdsforeachcountry;modifiesthephysicochemicalconditionsofthemedium
WatersolubleP TheNetherlands Quick,inexpensive,easytouse
Mayunderestimatetheavailablephosphorus
DegreeofPsaturation(ammoniumoxalate,Mehlich3,etc.)
TheNetherlands,Belgium(Flanders),Quebec
Indicatorisbothagronomicandenvironmental;presenceofthresholds
Sometimesdifficulttointerpret;theanalysisismoretime-consumingthanothers
Biologicalextractions Bioassays None Reproducesthebehaviourofplants
Time-consumingandimpossibletouseroutinely
Anionexchangeresins Brazil Mimicstheeffectofroots;suitableforallsoils
Affectsthephysico-chemicalequilibria;doesnotprovideaninfinitebindingcapacity
Isotopes None ProvidesamodelofPiontransferkinetics;suitableforawiderangeofsoils;itsresultscanbeextrapolatedtolongerperiodsoftime;itpermitsobservationwithoutchangingequilibria
Cannotbeusedroutinely;manipulationofradioelementsisextremelydelicate;alackofinformationabouttheorganicfraction;experimentaldilutionconditionsdonotrepresentnaturalconditions
Morecomplexmodels PI,ODEP,etc. Quebec,USA,Norway,Finland,Denmark
Comprehensive;takesintoaccountbothsourceandtransportfactors;predictstheamountsandformsofP
Greatcomplexity;significantamountofrequireddata;oftenimperfect;moderatelyoperational
268 Biotechnol. Agron. Soc. Environ. 201620(S1),257-272 RennesonM.,BarbieuxS.&ColinetG.Ta
ble
5. AnevaluationofdifferentindicatorsaccordingtoSAFE
requirementsthatagoodindicatorshouldmeet—
Éva
luat
ion
des
diffé
rent
indi
cate
urs
sur
base
des
ex
igen
ces S
AFE
que
doit
rem
plir
un
bon
indi
cate
ur(V
anCauwenberghetal.,2007)
Indi
cato
rsA
bilit
y to
di
scri
min
ate
in ti
me
and
spac
e
Ana
lytic
al a
ccur
acy
Cos
t and
tim
e of
ana
lysis
Tran
spar
ency
an
d ea
se o
f in
terp
reta
tion
Rel
evan
ce to
re
gula
tions
Tran
sfer
and
ex
trap
olat
ion
of
resu
ltsIndicatorsofP
bioavailability
AvailableP
accordingto
Lakanen-Erviö
+++
++Fast
++Agronom
ic
thresholds
-Suitablefor
Walloonso
ilsbutnot
forcarbonatedsoils
AvailableP
byothermethods
++Ingeneral
+Av
erage
+Thresholds
+Regional
thresholds
-Notsu
itableforall
soils
SolubleP
inwater
++M
ayvary
dependingonthe
season
+++
Veryfast,
lowcost
+Thresholds
+Regulations
withinso
me
countries
+Suitableforall
soils,notsensitiveto
soiltexture
DegreeofPsaturation
++Takesinto
accountsoil
properties
++Fast
-DependsonP
andsoilbinding
capacity
+Environm
ental
thresholdof25%
-Developedfor
acidicsandysoils
Sequentialextractions
+Dependson
agriculturalpractices
-Canbedifficultto
achieverepeatability
-Time-
consum
ing
andnotused
routinely
-Diffi
cultdueto
themanyforms
ofP
~None
~
Anionexchangeresins
+~Dependsonthe
presenceofN
O3and
SO4
+Simple
+~
+Suitableforall
soils
Biologicalextractions
~-P
oorrepeatability
-Time-
consum
ingand
notroutinely
used
~~None
+Suitableforall
soils
Isotopes
+DependsonP
inputs
+Cannotbeused
anyw
here
-Delicate
manipulations
needtobe
performedbyan
accreditedlab
+Kinetic
modelling
~None
+Notboundtoso
il
Indiceso
friskofPloss
++Takesinto
accountsoil
properties
-Calibrationrequired
~Rapidonce
calibrated,but
time-consum
ing
andcostly
++Provideslosses
andrisks,possible
simulations
+InNordic
countries
-Calibratedforeach
soil
++:verygood—
très
bon;+:good—b
on;~:m
ixed—
moy
en;-:bad—
mau
vais.
Indicatorsofphosphorusinsoils 269
largelytheresultofpigfarmsbeingprimarilylocatedinFlanders.However,Walloniacontainslowersurplus,despite significant regional disparities (Genot et al.,2009;Rennesonetal.,2015).
InWallonia, the soils predominantly range fromneutral toacidic,andPextractionhasbeenroutinelyperformedbysoilanalysislaboratoriessince1990usingammonium acetate and EDTA (pH 4.65) (Lakanenetal.,1971).Ammoniumacetateisusedtodissolveanyaluminiumphosphatepresent,while theacidity leadstoareversiblereleaseofPfromiron,aluminium,andpartlycalcium(Honsetal.,1990;Woodardetal.,1994).Thechelatingagent,EDTA,aidsinthepreventionofnewly released P from binding to iron (Dao, 2004).Thus, ammonium acetate facilitates the dissolutionofasubsetofPassociatedwithironoxidemoleculesthatareeithernotcrystallizedorslightlycrystallized,correspondingwith the different forms of P that areavailable to plants. However, ammonium acetate islargely unable to dissolve calcium phosphates, thusrenderingitunsuitedtocarbonatedsoils.InWallonia,carbonatedsoilsrepresentlessthanonepercentoftheterritory.
According to table 5, available P from Lakanen-Erviö (PAA-EDTA) presents different advantages.However, to be an adequate agronomic indicator, itshould be related to P uptake or yields. Hons et al.(1990), Homsy (1992) and Woodard et al. (1994)studiedtherelationshipbetweenPexportedbyplantsand yields. PAA-EDTA was correlated with P fertilizerlevel P (R²= 0.89), primarily in slightly acidic soils(Hons et al., 1990). In a study by Woodard et al.(1994), PAA-EDTA predicted yields and P concentrationresponses in plants more accurately than Olsen P.ThresholdshavebeendefinedforWalloniaaccordingtoPplantlevelsandsoilpropertyresults(Genotetal.,2011).
Significant correlations between PAA-EDTA andotherPextractionmethodshavebeenobserved,moststronglywiththemethodsofBray,Olsen,andSissingh(Ryser et al., 2001). Hons et al. (1990) and Homsy(1992) foundR2 values (between ammonium acetateand EDTA extraction and others) varying between0.66and0.91andbetween0.64and0.90,respectively,depending on the soil.Therefore, theLakanen-Erviömethod is well correlated with other predominantmethods used worldwide. Neyroud et al. (2003)classified the amount of extractable P according to16methods.TheLakanen-Erviömethodextracts lessPthanMehlich3orBraymethods,butmorethantheOlsenmethod.
The northern region of Belgium consists of soils(loamy and sandy texture)which exhibit a relativelylimited P binding capacity and relatively high levelsof available P. Extraction of P by the method ofLakanenetal.(1971)hasbeenperformedinWallonia
inanagriculturalcontextwithoutconsiderationforitsimpactontheenvironment.Celardin(2003)discovereda significant relationship between Lakanen-Erviö Pcontentandwater-extractableP,whichisrepresentativeofPlossrisk(R²of0.625and0.47forpH4.6-6.5andpH 6.6-8.6, respectively). Houben et al. (2011) andRennesonetal.(2015)haveevaluatedthepotentialforusing thedegreeofP saturationasanenvironmentalindicator in the Walloon region. The latter mayrepresentapromisingindicatorofPstatusinWallonia,provided that the existing equations are adapted tothe soil characteristics beforehand (Renneson et al.,2015).Moreover,Rennesonetal.(2015)showedthatextraction of P using the method of Lakanen et al.(1971)correlated(R=0.78)with theextractionofPperformedwithammoniumoxalate,whichisusedforcalculatingthedegreeofPsaturation.
OtherindicatorssuchasPindexforriskofPlosscouldbedeveloped.However,relativelyfewdataarecurrently available regarding the amount of P lossoccurringintheagriculturalparcelsofWallonia.
5. CONCLUSIONS
AprofusionofPindicatorstypescanbefoundintheliterature.ThemajorityofindicatorscurrentlyusedaredesignedtocharacterizethestatusofthesoilbasedontheirPcontent,whereasother,morecomplexindicatorsaredesignedtomodelPflowtotheenvironmentandestimate the riskofP loss.The latter assessesP losswhilesimulatingcontributionsto,ormanagementof,Pcontent.Thisapproachhasacertainadvantage,yetthesemethodscanbetime-consumingdifficulttoimplement.Consequently,mostcountriesusesimplermeasurementindicatorsofsoilP,involvingvarioustypesofanalysesrangingfromsingleextractionmethodstotheuseofPisotopes.Severalmethodsalsouseionexchangeresinsorbiologicalextractions.Moreover,someauthorshavecombined single extractions tomeasureP forms intothesoil(sequentialextraction).
Correlations can often be found between thedifferentindicatorsreportedintheliterature.However,theseresultscanrarelybeappliedgenerallyanddependfromonesoiltoanother.SomeauthorshaveproposedaclassificationsystembasedontheamountofPthatisextracted.However, thisisnotuniversalforall typesofsoil.
Eachofthemethodsavailablehascertainadvantagesand disadvantages, and no indicator is suitable forevery soil.Theappropriatenessof all indicator typeshasbeenevaluatedaccording tovariouscriteria.Theselection of an indicator is generally a compromisebetweenfixedobjectiveandothercriteria,suchasthecostandtimeofanalysis.However,tomakeachoiceinexistingindicators,aglobalanalysismustbemade.
270 Biotechnol. Agron. Soc. Environ. 201620(S1),257-272 RennesonM.,BarbieuxS.&ColinetG.
Someindicators,suchasisotopicmethodshavegreatpotential as P indicators, though their routine use isdifficult(necessityofspecializedlaboratories).Currentindicators are generally explained by their historicaluse. Some authors suggest that extraction methodsshouldbeharmonizedthroughoutEurope.However,itisdifficulttochangeamethodduetotheinadequacyinsomesoils,lackofthresholdineachcountry,andthenecessitytomaintainsoilqualitymonitoring.
At present, in Wallonia (southern Belgium), Pextraction has been performed using ammoniumacetateandEDTA(pH4.65).Thisextractionmethodiswell correlatedwith other extractionmethods andcropyields.AreflectionismadetostudytheinterestofotherindicatorsasthedegreeofPsaturationwhichisanenvironmentalindicator.
This review highlights the importance of carefulconsiderationofindicatorchoice,andtheestablishmentof interpretation thresholds. This review creates acomparisonofappropriateandregionalizedreferencevalues.
Bibliography
BeaucheminS. & SimardR.R., 1999. Soil phosphorussaturation degree: review of some indices and theirsuitabilityforPmanagementinQuebec,Canada.Can. J. Soil Sci.,79,615-625.
BeaudinI., 2006. Revue de littérature. La mobilité du phosphore. Québec, Canada: Centre de référence enagricultureetagroalimentaireduQuébec.
BeckM.A.&SanchezP.A.,1994.Soilphosphorusfractiondynamics during 18 years of cultivation on a typicpaleudult.Soil Sci. Soc. Am. J.,58,1424-1431.
BoströmB.,PerssonG.&BrobergB.,1988.Bioavailabilityof different phosphorus forms in freshwater systems.Hydrobiologia,170,133-155.
BozongoJ.C.,BertruG.&MartinG.,1989.Lesméthodesdespéciationduphosphoredanslessédiments:critiquesetpropositionspourl’évaluationdesfractionsminéralesetorganiques.Arch.Hydrobiol.,116,61-69.
BrayR.H.K. & KurtzL.T., 1945. Determination of total,organicandavailableformsofphosphorusinsoils.Soil Sci.,59,39-45.
BreeuwsmaA.,ReijerinkJ.G.A.&SchoumansO.F.,1995.Impact of manure on accumulation and leaching ofphosphate in areas of intensive livestock farming. In:SteeleK.,ed.Animal waste and the land water interface.NewYork,USA:Lewis-CRC,239-251.
BuczkoU.&KuchenbuchR.O.,2007.Phosphorusindicesas risk-assessment tools in theU.S.A. and Europe - areview.J. Plant Nutr. Soil Sci.,170,445-460.
CelardinF.,2003.EvaluationofsoilP-testvaluesofcantonGeneva/SwitzerlandinrelationtoPlossrisks.J. Plant Nutr. Soil Sci.,166,416-421.
ChangS.C. & JacksonM.L., 1957. Fractionation of soilphosphorus.Soil Sci.,84,133-144.
ChardonW.J., MenonR.G. & ChienS.H., 1996. Ironoxide impregnatedfilterpaper(Pi test):areviewof itsdevelopmentandmethodologicalresearch.Nutr. Cycling Agroecosyst.,46,41-51.
CondronL.M., TurnerB.L. & Cade-MenunB.J., 2005.Chemistryanddynamicsofsoilorganicphosphorus.In:SimsJ.T.&SharpleyA.N.,eds.Phosphorus: agriculture and the environment. Madison, WI, USA: AmericanSocietyofAgronomy,87-121.
DaoT.H.,2004.Ligandsandphytasehydrolysisoforganicphosphorusinsoilsamendedwithdairymanure.Agron. J.,96,1188-1195.
DautrebandeS. & SohierC., 2006. L’érosion hydrique et les pertes en sols agricoles en Région wallonne. Rapport analytique 2006 sur l’état de l’environnement wallon.Gembloux,Belgique:FacultéuniversitairedesSciencesagronomiquesdeGembloux.
DyerB.,1894.Ontheanalyticaldeterminationofprobablyavailablemineralplantfoodinsoils.J.Chem. Soc.,65,115-167.
EgnerH.,RiehmH.&DomingoW.R.,1960.Untersuchungenüber die chemische Bodenanalyse als Grundlage fürdie Beurteilung des Nährstoffzustandes der Böden. II.Chemische Extraktionsmethoden zur Phosphor undKalium-bestimmung. Kungliga Lantbrukshögskolans Annaler,26,45-61.
EkholmP.&KrogerusK.,2003.Determiningalgal-availablephosphorus of differing origin: routine phosphorusanalysesversusalgalassays.Hydrobiologia,492,29-42.
FardeauJ.-C., 1993. Le phosphore assimilable des sols:sareprésentationparunmodèlefonctionnelàplusieurscompartiments.Agronomie,13,317-333.
FardeauJ.C. & ConesaA.P., 1994. Le phosphore. In :BonneauB. & SouchierM., éds. Pédologie. 2.Constituants et propriétés du sol. Paris: MassonPublishing,649-658.
FrossardE.,SinajS.,ZhangL.-M.&MorelJ.-L.,1996.Thefateofsludgephosphorusinsoil-plantsystems.Soil Sci. Soc. Am. J.,60,1248-1253.
FrossardE., JulienP., NeyroudJ.-A. & SinajS., 2004.Le phosphore dans les sols. État de la situation en Suisse.Cahierdel’environnementn°368.Berne:Officefédéraldel’environnement,desforêtsetdupaysage.
GenotV., ColinetG., BrahyV. & BockL., 2009. L’étatde fertilité des terres agricoles et forestières en régionwallonne (adapté du chapitre 4 - sol 1 de «L’État del’Environnement wallon 2006-2007»). Biotechnol. Agron. Soc. Environ.,13,121-138.
Genotetal.,2011.Un conseil de fumure raisonné. Le cas du phosphore.Gembloux,Belgique:REQUASUD.
GoltermanH.L., 1995. The role of ironhydroxyde-phosphate-sulphide system in the phosphate exchangebetweensedimentsandoverlyingwater.Hydrobiologia,297,43-54.
Indicatorsofphosphorusinsoils 271
HedleyM.J., StewartJ.W.B. & ChauhanB.S., 1982.Changes in inorganic and organic soil phosphorusfractions induced by cultivation practices and bylaboratoryincubations.Soil Sci. Soc. Am. J.,46,970-976.
HieltjesA.H.M. & LijklemaL., 1980. Fractionationof inorganic phosphates in calcareous sediments.J. Environ. Qual.,9,405-407.
HinsingerP., 2001. Bioavailability of soil inorganic P inthe rhizosphere as affected by root-induced chemicalchanges:areview.Plant Soil,237,173-195.
HomsyS., 1992. Comparaison de quatre méthodes utilisées en routine dans les laboratoires européens pour l’appréciation de l’offre en phosphore disponible du sol. Mémoire: Faculté universitaire des SciencesagronomiquesdeGembloux(Belgique).
HonsF.M., LarsonvollmerL.A. & LockeM.A., 1990.NH4OAC-EDTA-Extractable phosphorus as a soil testprocedure.Soil Sci.,149,249-256.
HoodaP.S.etal.,2000.Relatingsoilphosphorusindicestopotentialphosphorusreleasetowater.J.Environ. Qual.,29,1166-1171.
HoubaV.J.G., TemminghoffE.J.M., GaikhorstG.A. &vanVarkW.,2000.Soilanalysisproceduresusing0.01Mcalciumchlorideasextractionreagent.Commun. Soil Sci. Plant Anal.,31,1299-1396.
HoubenD., MeunierC., PereiraB. & SonnetP., 2011.Predictingthedegreeofphosphorussaturationusingtheammonium acetate-EDTA soil test.Soil Use Manage.,27,283-293.
ISO 11263:1994, 1994. Soil quality - Determination of phosphorus - Spectrometric determination of phosphorus soluble in sodium hydrogen carbonate solution.Vernier,Switzerland:ISO.
Jordan-MeilleL. et al., 2012.An overview of fertilizer-PrecommendationsinEurope:soiltesting,interpretation,andfertilizerrecommendations.Eur. J. Soil Sci.,28,419-435.
JoretG.&HébertJ.,1955.Contributionàladéterminationdubesoindessolsenacidephosphorique.Ann. Agron.,2,233-299.
KhiariL. et al., 2000.An agri-environmental phosphorussaturation index for acid coarse-textured soils.J. Environ. Qual.,29,1561-1567.
LakanenE. & ErviöR., 1971. A comparison of eightextractants for the determination of plant availablemicronutrientsinsoils.Acta Agralia Fennica,123,223-232.
LeinweberP., MeissnerR., EckhardtK.U. & SeegerJ.,1999.Management effects on forms of phosphorus insoilandleachinglosses.Eur. J. Soil Sci.,50,413-424.
LemunyonJ.L.&GilbertR.G.,1993.Theconceptandneedforaphosphorusassessmenttool.J. Prod. Agric.,6,449-450.
LevyE.T. & SchlesingerW.H., 1999. A comparison offractionationmethodsforformsofphosphorusinsoils.Biogeochemistry,47,25-38.
MaguireR.O. & SimsJ.T., 2002. Soil testing to predictphosphorusleaching.J. Environ. Qual.,31,1601-1609.
MasonS., HarnonR., ZhangH. & AndersonJ., 2008.Investigating chemical constraints to themeasurementofphosphorusinsoilsusingdiffusivegradients in thinfilms(DGT)andresinmethods.Talanta,74,779-787.
MasonS., McNeillA., McLaughlinM.J. & ZhangH.,2010. Prediction of wheat response to an applicationof phosphorus under field conditions using diffusivegradients in thin films (DGT) and extractionmethods.Plant Soil,337,243-258.
McIntoshJ.L.,1969.BrayandMorgansoiltestextractantsmodified for testing acid soils from different parentmaterials.Agron. J.,61,259-265.
MehlichA.,1953.Determination of P, Ca, Mg, K, Na, NH4.Raleigh,NC,USA:DepartmentofAgriculture.
MehlichA., 1984. Mehlich3 soil test extractant: amodificationofMehlich2extractant.Commun. Soil Sci. Plant Anal.,15,1409-1416.
MorelC., TunneyH., PlenetD. & PellerinS., 2000.Transfer of phosphate ions between soil and solution:perspectivesinsoiltesting.J. Environ. Qual.,29,50-59.
MorelC. et al., 2014.Modeling of phosphorus dynamicsin contrasting agroecosystems using long-term fieldexperiments.Can. J. Soil Sci.,94,377-387.
MorganM.F., 1941. Chemical soil diagnosis by the universal test system.NewHaven,CT,USA:ConnecticutAgriculturalExperimentStation.
NegassaW.& LeinweberP., 2009. How does theHedleysequential phosphorus fractionation reflect impacts oflanduseandmanagementonsoilphosphorus:areview.J. Plant Nutr. Soil Sci.,172,305-325.
NémeryJ., 2003. Origine et devenir du phosphore dans le continuum aquatique de la Seine des petits bassins amont à l’estuaire : rôle du phosphore échangeable sur l’eutrophisation. Thèse de doctorat: Université ParisVI-PierreetMarieCurie(France).
NeyroudJ.A.&LischerP.,2003.Dodifferentmethodsusedto estimate soil phosphorus availability across Europegive comparable results? J. Plant Nutr. Soil Sci.,166,422-431.
NFX 31-147, 1996.Qualité des sols – Sols, sédiments - Mise en solution totale par attaque acide. La PlaineSaint-Denis,France:AFNOR.
NFX31-160, 1999.Qualité des sols – Détermination du phosphore soluble dans une solution à 20 g·l-1 d’acide citrique monohydraté – Méthode Dyer.LaPlaineSaint-Denis,France:AFNOR.
NF X 31-161, 1999. Qualité des sols – Détermination du phosphore soluble dans une solution d’oxalate d’ammonium à 0,1 mol·l-1 – Méthode Joret-Hébert.LaPlaineSaint-Denis,France:AFNOR.
ObersonA., BessonJ.M., MaireN. & SticherH., 1996.Microbiological processes in soil organic phosphorustransformationsinconventionalandbiologicalcroppingsystems.Biol. Fertil. Soils,21,138-148.
272 Biotechnol. Agron. Soc. Environ. 201620(S1),257-272 RennesonM.,BarbieuxS.&ColinetG.
OlsenS.R., ColeC.V.,WatanabeF.S.&DeanL.A., 1954.Estimation of available phosphorus in soils by extraction with sodium bicarbonate. USDA Circular No 939.Washington:USGovernmentPrintingOffice.
PaludanC. & JensenH.S., 1995. Sequential extraction ofphosphorus in freshwater wetland and lake sediment:significanceofhumicacids.Wetlands,15,365-373.
PansuM. & GautheyrouJ., 2003. L’analyse du sol minéralogique, organique et minérale.Paris:Springer.
PierzynskiG.M., McDowellR.W. & SimsJ.T., 2005.Chemistry,cycling,andpotentialmovementofinorganicphosphorus in soils. In: SimsJ.T. & SharpleyA.N.Phosphorus: agriculture and the environment. Madison,WI,USA:AmericanSocietyofAgronomy,53-86.
PoteD.H. et al., 1999. Relationship between phosphoruslevelsinthreeUltisolsandphosphorusconcentrationsinrunoff.J. Environ. Qual.,28,170-175.
ProixN., 2013. Revue des méthodes d’analysesagronomiques utilisées en Europe. In : Actes des 11e Rencontres de la fertilisation raisonnée et de l’analyse de terre, 20-21 novembre 2013, COMIFER-GEMAS, Poitiers, France.
QianP.&SchoenauJ.J.,2002.Practicalapplicationsofionexchange resins in agricultural and environmental soilresearch.Can. J. Soil Sci.,82,9-21.
RennesonM.etal.,2013.RelationshipsbetweenthePstatusofsurfaceanddeephorizonsofagriculturalsoilsundervariouscroppingsystemsandfordifferentsoiltypes:acasestudyinBelgium.Soil Use Manage.,29,94-102.
RennesonM.etal.,2015.Degreeofphosphorussaturationinagriculturalloamysoilswithanear-neutralpH.Eur. J. Soil Sci.,66,33-41.
RuttenbergK.C., 1992. Development of a sequentialextractionmethodfordifferentformsofphosphorusinmarinesediments.Limnol. Oceanogr.,37,1460-1482.
RydinE.&WelchE.B.,1998.Aluminiumdoserequiredtoinactivatephosphateinlakesediments.Water Res.,32,2969-2976.
RyserJ.-P.,WaltherU.&FlischR.,2001.Donnéesdebasepourlafumuredesgrandesculturesetdesherbages.Rev. Suisse Agric.,33,80.
SchüllerH.,1969.DieCAL-Methode,eineneueMethodezurBestimmungdespflanzenverfugbarenPhosphatesinBoden.Z. Pflanzenernahr. Bodenkd.,123,48-63.
SharpleyA. & WangX., 2014. Managing agriculturalphosphorusforwaterquality:lessonsfromtheUSAandChina.J. Environ. Sci.,26,1770-1782.
SissinghH.A.,1971.AnalyticaltechniqueofthePwmethod,usedfortheassessmentofthephosphatestatusofarablesoilsinTheNetherlands.Plant Soil,34,483-486.
SixL. et al., 2012. The performance of DGT versusconventionalsoilphosphorustestsintropicalsoils-anisotopedilutionstudy.Plant Soil,359,267-279.
StanfordG.&DeMentJ.D.,1957.Amethodformeasuringshort-termnutrientabsorptionbyplants:I.phosphorus.Soil Sci. Soc. Am. J.,21,612-617.
TaoufikM., KemmouS., IdrissiL.L. & DafirJ.E., 2004.Comparaison de deux méthodes de spéciation duphosphoredansdessédimentsdelapartieavaldubassinOumRabiaa(Maroc).Water Qual. Res. J. Can.,39,50-56.
TiessenH.&MoirJ.O.,1993.CharacterizationofavailableP by sequential extraction. In: CarterM.R., ed. Soil sampling and methods of analysis.BocaRaton,FL,USA:CanadianSocietyofSoilScience,LewisPublishers,75-86.
TranT.S. & GirouxM., 1985. Comparison of severalmethods of extracting available P in relation with thechemicalandphysicalproperties.Can. J. Soil Sci.,65,35-46.
TruogE., 1930. Determination of readily availablephosphorusofsoils.J. Am. Soc. Agron. J.,22,874-882.
Van CauwenberghN. et al., 2007. SAFE -A hierarchicalframeworkforassessingthesustainabilityofagriculturalsystems.Agric. Ecosyst. Environ.,120,229-242.
van der ZeeS.E.A.T.M. & van RiemsdijkW.H., 1988.Modelforlong-termphosphatereactionkineticsinsoil.J. Environ. Qual.,17,35-41.
WilliamsJ.D.H., SyersJ.K. & WalterT.W., 1967.Fractionation of soil inorganic phosphate by amodificationofChangandJacksonprocedure.Soil Sci. Soc. Am. J.,31,736-739.
WoodardH.J.etal.,1994.Apreliminarycomparisonoftheammonium acetate-EDTA soil phosphorus extractionmethod to the Bray-1 and Olsen soil phosphorusextractionmethods.Commun. Soil Sci. Plant Anal.,25,2909-2923.
ZhangC., DavisonW., GadiR. &KobayashiT., 1998. In situ measurement of dissolved phosphorus in naturalwatersusingDGT.Anal. Chim. Acta,370,29-38.
ZhangC. et al., 2014. Bioavailability assessment ofphosphorusandmetalsinsoilsandsediments:areviewof diffusive gradients in thin films (DGT). Environ. Monit. Assess.,186,7367-7378.
(82ref.)