Chemical fractionation of phosphorus, sulphur, and molybdenum in Brazilian savannah Oxisols under...

$: Chemical fractionation of phosphorus, sulphur, and molybdenum in Brazilian savannah Oxisols under different land use$
Ž .Geoderma 96 2000 31–46

Chemical fractionation of phosphorus, sulphur, andmolybdenum in Brazilian savannah Oxisols under

different land use

Juliane Lilienfein a,), Wolfgang Wilcke a, Miguel Angelo Ayarza b,Lourival Vilela c, Samuel do Carmo Lima d, Wolfgang Zech a

a Institute of Soil Science and Soil Geography, UniÕersity of Bayreuth,D-95440 Bayreuth, Germany

b CIAT-Laderas, Apartado 1410, Tegucigalpa, Hondurasc EMBRAPA, Caixa Postal 08223, Planaltina-DF, Brazil

d Federal UniÕersity of Uberlandia, 38406-210 Uberlandia-MG, Brazilˆ ˆ

Received 6 October 1998; received in revised form 12 July 1999; accepted 7 December 1999

Abstract

In soils, P, S, and Mo occur organically bound or as oxyanions. In Oxisols, oxyanions may bestrongly sorbed due to the high positive surface charge. The objective of this work was to comparechemical fractions of P, S, and Mo in differently used Oxisols with similar properties in an

Ž .on-farm experiment. Soil samples 0–0.15 m were taken from three replicate plots of each ofŽ . Ž . Ž .conventional tillage CT and no-till NT maize–soybean and conventional tillage sugarcane SC

Ž . Ž . Ž .cropping systems, degraded DP and productive pastures PP , Eucalyptus EU and PinusŽ . Ž . Ž .reforestations PI , and native savannah CE . The samples were sequentially extracted with 1

Ž . Ž . Ž . Ž .0.5 M NaHCO , 2 0.1 M NaOH, 3 1 M HCl, 4 hot concentrated HCl, and 5 concentrated3Ž .HClO rHNO . In the extracts, inorganic P and total P, S, and Mo were determined. Organic P4 3 i

Ž . y1P was calculated as total PyP . Total concentrations were 333–567 mg P kg , 231–284 mg So i

kgy1, and 3.2–3.9 mg Mo kgy1. The most important fractions in all studied systems were theŽ . Ž .NaOH fraction for P 38–49% of total P , the HClO rHNO fraction for S 27–35% of total S4 3

Ž .and the concentrated HCl fraction for Mo 86–90% of total Mo . The proportion of the moreeasily extracted NaHCO qNaOH fractions decreased along the line S)P)Mo. Fertilisation3

Žincreased plant-available P and S fractions in CT and NT whereas recalcitrant fraction con-.centrated HCl and HClO rHNO remained unchanged. The P rP ratios in NT and CT were4 3 i o

) Corresponding author. Tel.: q49-921-552-340; fax: q49-921-552-246.Ž .E-mail address: [email protected] J. Lilienfein .

0016-7061r00r$ - see front matter q 2000 Elsevier Science B.V. All rights reserved.Ž .PII: S0016-7061 00 00002-1

( )J. Lilienfein et al.rGeoderma 96 2000 31–4632

higher than in CE because fertiliser P mainly accumulated in inorganic P fractions. The pasturesoil had lower P concentrations indicating export by grazing. Thus, 12–20 years of land use hadmarked effects on P, smaller ones on S, and almost none on Mo concentrations and chemicalfractions. q 2000 Elsevier Science B.V. All rights reserved.

Keywords: Cerrado; Eucalyptus reforestation; Hedley fractionation; pasture; Pinus reforestation

1. Introduction

Ž .Large parts of the Brazilian savannah region the Cerrados are covered bynutrient-poor Oxisols with low P, S, and Mo concentrations. As a consequenceof the positive charge of hydrous oxides of Fe, which occur in high concentra-tions in Oxisols, P is strongly sorbed resulting in low plant-available P concen-

Ž .trations Goedert, 1983 . In soils, inorganic S and Mo also mainly exist asoxyanions. Therefore, it is likely that these plant nutrients also are strongly

Ž .retained in Oxisols. Mottavalli et al. 1993 , e.g., did not detect any KCl-extrac-table sulphate in a Cerrado Oxisol. We did not find any data on plant-availableMo concentrations of Oxisols in literature.

In strongly weathered Oxisols, only organic and occluded P forms in hydrousŽ .oxides of Fe are found in more than trace amounts Smeck, 1985 . In Colombian

soils under natural savannah about half of total P was found in recalcitrantfractions extractable with strong acids. Only 11% could be considered as

Ž .plant-available Friesen et al., 1997 . Similar results were reported by TiessenŽ .and Moir 1993 for Brazilian Oxisols.

Land-use impacts on P concentrations and chemical fractions in both, temper-ate and tropical soils, have been studied by various authors. Generally, inunfertilised cropping systems total P concentrations decrease and organic P

Žfractions are the major source for plant-available P Schoenau et al., 1989; Beck.and Sanchez, 1994; Zhang and MacKenzie, 1997a,b . In fertilised systems,

Ž .labile inorganic P fractions NaHCO -P and NaOH-P increase and fertiliser-P3 i i

is slowly transformed into more strongly bound P forms whereas organic PŽconcentrations remain unchanged Beck and Sanchez, 1994; Friesen et al., 1997;

.Zhang and MacKenzie, 1997a,b .Information on S and Mo chemical fractions in tropical soils and on land-use

effects on S and Mo concentrations and chemical fractions is scarce. We did notfind any study on the influence of land use on Mo chemical fractions. More

Ž .work has been done on temperate soils. Zucker et al. 1986 found up to 84% oftotal S in KH PO -extractable forms in acid mineral soil horizons. Curtin and2 4

Ž .Syers 1990 extracted almost all added sulphate with 1 M NaCl in agriculturallyused soils of the UK and Ireland. Both results indicate that sulphate is lessstrongly sorbed in soils than phosphate. The converse may be true for Mo. Afterincubation for 3–10 months of several Australian soils to which Mo was added,

( )J. Lilienfein et al.rGeoderma 96 2000 31–46 33

only -4% of total Mo was leachable with water. More than half of total Mowas stored in organic fractions and up to 40% in recalcitrant ones. Most of therecalcitrant Mo could be assigned to bonding to Fe oxides extractable with

Ž .dithionite-citrate Smith and Leeper, 1969 . The authors also reported that Moslowly is transformed from more bioavailable fractions to more strongly boundfractions in a few months after Mo application.

Few studies deal with the impact of S fertilising on S forms in soils.Ž .Castellano and Dick 1990 found most of applied fertiliser-S in the salt-extrac-

Ž .table sulphate fraction in a soil from Oregon U.S.A. . In contrast, Sakadevan etŽ .al. 1993 reported that most of applied fertiliser-S remained in organic form in

a New Zealand soil. In a Cerrado Oxisol, detectable KCl-extractable sulphatewas only found after liming to reduce positive surface charge or addition oforganic materials, which compete with sulphate for sorption sites and thus

Ž .reduce sulphate sorption Mottavalli et al., 1993 . Most information about theland-use impact on P and S were obtained in controlled studies on research

Ž .stations ‘‘on-station’’ . The results may not be readily transferable to ‘‘on-farm’’conditions.

Ž .The objectives of this work are 1 to determine P, S, and Mo concentrationsŽ .and chemical fractions in differently used savannah Oxisols and 2 to asses the

land-use influence on P, S, and Mo pools in an on-farm experiment. We used asequential extraction procedure developed to partition P in soils also to extract Sand Mo because of the chemical similarity of the elements. Similar considera-

Ž .tions led McLaren et al. 1998 to study chemical fractions of As, which in soilalso occurs as oxyanion, with a method developed to partition P in soils.

2. Materials and methods

2.1. Study sites

ŽThe study area is located southeast of Uberlandia Minas Gerais, about 400ˆ. 2km S of Brasilia, Fig. 1 . Within an area of about 100 km , three spatially

separated plots of each of eight systems representing the most importantŽ .land-use systems of the study region were selected: 1 conventional tillage

Žmaize–soybean rotation CT, annual ploughing with plug and disk harrow, oneŽ . Ž .harvest per year: maize Zea mays L. –soybean Glycine max L. Merr.

Ž . Žrotation, 2 no-till maize–soybean rotation NT, two harvests per year, firstŽsoybean, second maize the two maize–soybean systems are annually fertilised

with about 100 kg P hay1, 30 kg S hay1, 0.1 kg Mo hay1, the second crop iny1 y1 y1. Ž .NT received additional 50 kg P ha , 15 kg S ha , 0.05 kg Mo ha , 3

Žconventional tillage sugarcane SC, Saccharum officinalis L., initial fertilisationy1. Ž . Žof 100 kg P ha , 4 degraded pasture DP, Brachiaria decumbens Stapf,

partly open soil, some invasive Cerrado plants, initial fertilisation with 40 kg P,


Fig. 1. Locations of the study sites.

y1. Ž . Ž65 kg K, 32 kg N, and 1 mg of dolomite ha , 5 productive pasture PP,Brachiaria decumbens Stapf, pure grass, initially fertilised at the same rates as

.DP, then with 17 kg P and 33 kg K in 1996r97 , there were additional fertiliserŽ .applications to PP at unknown rates and application times prior to 1996, 6

Ž . Ž .Pinus caribaea Morelet plantation PI , 7 Eucalyptus grandis W. Hill exŽMaiden plantation both forest plantations were initially fertilised with 80-g

y1.Superphosphate per plant corresponding to 13 kg P ha . Native savannah soilsŽ . Ž .Cerrado, CE were used as references 8 . All study sites are continuously used

Ž . Ž . Ž .for the same purposes for 20 PI, EU , 12 DP, PP, CT , or 1–6 years SC andpassed directly from natural vegetation to the current land-use system except forthe NT soils, which were used as conventionally tilled cropping systems for9–11 years and the SC soils, which were used as pastures 6–11 years prior tothe present land use. We only chose plots at comparable topographic positionswith macroscopically similar soils. A prerequisite for statistical evaluation is theindependence of the replicates. Although a completely randomised distribution


of the study plots could not be realised, because we had to use the existingsystems, we consider the statistical prerequisites for variance analysis to be metas the minimum distance between two replicates of the same system is 300 m.

Ž .Composite soil samples 0–15 cm consisting of five subsamples were takenfrom 10=10-m2 plots of each of the three replicates per system at the end of

Ž .the rainy season in April 1997. All samples were air-dried 408C and sieved to-2 mm.

2.2. Chemical and physical characterisation

Particle-size distribution was determined after removal of Fe oxides withNa-dithionite and organic matter with H O at 908C and following dispersion2 2

with hexametaphosphate. Coarse and fine sand, 250–2000 and 50–250 mmparticle-size diameter were separated by sieving. Silt and clay concentrations

Ž .were determined with the pipette method Gee and Bauder, 1996 .The following soil properties were determined: pH with a glass electrode in

H O and 1 M KCl, soilrsolution ratio 1:2.5, total concentrations of C, N, and S2

by dry combustion and gas chromatographic separation with a CNS analyserŽ . Ž .Elementar Vario EL , effective cation-exchange capacity ECEC by extraction

Ž .with 1 M NH -acetate Sumner and Miller, 1996 , Al in poorly ordered Fe4Ž .oxides with the oxalate-buffer method Al , Schwertmann, 1964 , crystalline Feo

Ž . Žoxides with the cold dithionite-citrate-buffer DCB method Fe , Holmgren,d.1967 . Metal concentrations in the extracts were determined with flame atomic

Ž .absorption spectrometry Varian SpectrAA 400 .

ŽFig. 2. Fractionation scheme Tiessen and Moir, 1993, modified; P , S , Mo s inorganic forms,i i i.P , S , Mo sorganic forms .o o o


Fig. 3. Average portions of P, S, and Mo of the total concentration in the fractions of theŽ . Ž .sequential extraction and standard deviations ns3 in soils of conventional tillage CT and

Ž . Ž . Ž .no-till NT maizersoybean rotations, conventional tillage sugarcane SC , Eucalyptus EU andŽ . Ž . Ž . Ž .Pinus PI plantations, degraded DP and productive pastures PP , and natural savannah CE .


2.3. P, S, and Mo fractionation

Ž .A modified version of the fractionation scheme of Hedley et al. 1982 wasused to sequentially fractionate soil P, S, and Mo into five fractions. Thefractionation scheme and an approximate characterisation of the obtained frac-tions are given in Fig. 2.

Total P and S concentrations in the extracts were determined with inductivelyŽ .coupled plasma-atomic emission spectroscopy ICP-AES, GBC Integra XMP .

Total Mo was measured with inductively coupled plasma-mass spectroscopyŽ .VG Plasmaquad PQ2 TurboqMS, VG Elemental .

2.4. Calculations and statistical analyses

The sum of the P, S, and Mo concentrations in the sequential extracts wasŽ .taken as total concentration Fig. 3 . Total S concentrations were determined

independently as given above.Main and interactive effects were tested by using Tukey’s honest significant

Ž . Ž .difference HSD mean separation test Hartung and Elpelt, 1989 . StatisticalŽanalysis was conducted with STATISTICA for Windows 5.1 StatSoft, 1995,

.Loll and Nielson, Hamburg, Germany . The significance level was set atp-0.05.

3. Results and discussion

3.1. Comparability of the soils

Differences in concentrations and chemical fractions of P, S, and Mo betweenland-use systems can only be interpreted as caused by land-use practices if thesoils were comparable prior to the beginning of land use. To test this pre-requisite, we compared soil characteristics which are not or only little influencedby land use like soil type, particle-size distribution, and Al and Fe concentra-o d

tions.ŽAll soils are anionic Acrustoxes with about 70–80% of clay Soil Survey

.Staff, 1997 . Concentrations of clay, Fe , and Al do not differ significantlyd oŽ .between the land-use systems Table 1 . Additionally, the standard deviations of

ŽFe and Al concentrations, between the replicate plots of one system Fe : 9.1,d o d. ŽAl : 0.27, ns3 are similar to those between all plots Fe : 9.2, Al : 0.33,o d o

.ns24 . We, therefore, assume that the soils had comparable properties prior toland use and differences can be interpreted as result of land-use practices.

()

J.Lilienfein

etal.r

Geoderm

a96

200031

–46

38

Table 1Ž .Average properties of the study soils 0–15 cm

a b cC N pH pH ECEC Fe Al Clayd oy1 y1 y1 y1 y1 y1w x w x Ž . Ž . w x w x w x w xg kg g kg H 0 KCl mmol kg g kg g kg mg kg2 c

Ž .Cerrado CE 24.8 1.42 4.71 3.99 4.8 42.23 3.45 680Ž .Degraded pasture DP 23.0 1.35 4.89 4.10 8.0 39.37 3.62 820Ž .Productive pasture PP 24.6 1.43 5.38 4.40 14.7 34.17 3.50 780

Ž .Eucalyptus EU 23.8 1.39 5.12 4.22 11.5 41.63 3.47 690Ž .Pinus PI 20.1 1.11 4.74 4.53 3.6 40.60 4.68 730

Ž .Conventional tillage sugarcane SC 23.3 1.40 5.69 4.57 18.2 40.58 3.50 810Ž .No-till maize–soybean NT 21.8 1.38 5.91 5.02 35.1 40.50 3.20 720

Ž .Conv. tillage maize–soybean CT 21.1 1.33 6.16 5.41 39.6 45.63 3.12 690

aEffective cation-exchange capacity.bDithionite-citrate extractable Fe.cOxalate-extractable Al.


3.2. Total concentrations and chemical fractions of P, S, and Mo

Total concentrations of P, S, and Mo range between 333 and 567, 231 andy1 Ž .284, and 3.2 and 3.9 mg kg , respectively Table 2 . Compared with concentra-

Žtions found in agriculturally used surface soils of the temperate regions P:y1 y1 .500–3000 mg kg , S: 400–500 mg kg , Sanchez, 1976 , P and S concentra-

tions are low, but P concentrations are about two times higher than those FriesenŽ . Ž .et al. 1997 reported for Colombian Oxisols. Tiessen and Moir 1993 found

between 130 and 150 mg P kgy1 in native and cultivated Oxisols in Brazil, Sranged between 78 and 319 mg S kgy1 in differently textured Cerrado OxisolsŽ .Kliemann, 1987; Lilienfein et al., 1998 . The Mo concentrations of the study

Ž y1 .soils are similar to concentrations 0.5–10 mg kg , Amberger, 1988 inŽ .temperate surface soils Amberger, 1988 and are within the normal range for

Ž y1 .Mo concentrations in tropical soils 0.1–40 mg Mo kg , Landon, 1991 .Total S concentrations are not significantly different between the sum of the

sequential fractions and the dry combustion and gas chromatographic detectionwith the CNS analyser indicating a quantitative and reproducible S extraction

Ž .with the sequential extraction scheme Table 2 .The distribution of P, S, and Mo among the fractions of the Hedley sequence

Ž .is element-specific Fig. 3 ; for each of the elements, it is similar in all studiedland-use systems. The largest P fraction is extracted with NaOH, followed bysimilar proportions extracted with concentrated HCl and HNO rHClO . No P3 4

was detected in the dilute HCl fraction because Ca-phosphate is absent in manyŽ .strongly weathered soils Tiessen and Moir, 1993 . The NaHCO -extractable3

Ž .proportion of total P concentrations in the study soils 4–5% is small comparedwith those in soils of the temperate region but also compared with those in other

Ž . Ž .Oxisols as reported by Tiessen and Moir 1993 and Friesen et al. 1997 whodetected about 8–20 % of total P in the NaHCO -fraction.3

The largest S fraction is the HNO rHClO fraction. The proportions of3 4

extracted S decrease along the line NaOH)concentrated HCl)NaHCO )3

dilute HCl. Sulphur is the only one of the studied elements, which was found inthe dilute HCl fraction. We were not able to measure the sulphate concentrationsin the extracts because they were below the detection limit of our analyticalequipment. Therefore, it was not possible to distinguish analytically betweenorganic and inorganic S in the extracts. We also did not determine the extractedS species. Thus, we may only tentatively assign the S extracted with particularfractions of the sequential extraction scheme to chemical S forms. Solutions of

Ž0.5 M NaHCO are known to extract free sulphate completely Kilmer and3.Nearpass, 1960; Johnson and Todd, 1983; Alewell and Matzner, 1996 . Freney

Ž . Ž .1961 and Freney et al. 1971 reported that NaHCO may also dissolve3Ž .organic S forms. However, Alewell 1993 studied the sulphate release of 10

different synthetic organic sulphur compounds and found that only from thearomatic ester sulphate 8-hydroxyquinoline sulphate was released to a larger

()

J.Lilienfein

etal.r

Geoderm

a96

200031

–46

40

Table 2Average P, S, and Mo concentrations in the fractions of the sequential extraction. Values followed by different letters are significantly different between

Ž .the land-use systems at P -0.05 Tukey’s HSD test

Ž .a Phosphorus

aNaHCO -P NaHCO -P NaOH-P NaOH-P 1 M HCI-P HCl -P HNO rHClO -P P3 i 3 o i o conc 3 4 toty1w xmg kg

Ž .Cerrado CE 6.7a 10.9a 41.5a 101.2a nd 120ab 108a 388aŽ .Degraded pasture DP 5.8a 11.1a 40.0a 88.9a nd 114ab 83b 341aŽ .Productive pasture PP 6.2a 14.7a 40.5a 96.6a nd 97b 78b 333a

Ž .Eucalyptus EU 9.6ab 10.7a 70.0a 110.9a nd 115ab 87b 403a( )Pinus PI 9.6ab 7.3a 47.9a 118.7a nd 95b 82b 361a

Ž .Conventional tillage SC 6.5a 7.6a 51.4a 103.4a nd 114ab 87b 370aŽ .No-till maize–soybean NT 19.1b 9.0a 159.5b 120.4a nd 172a 87b 567b

Ž .Conv. tillage maize–soybean CT 14.1ab 6.0a 119.9ab 117.9a nd 158ab 84b 500ab

Ž .b Sulphura bNaHCO -S NaOH-S 1 M HCI-S HCI -S HNO rHCIO -S S CNC-S3 conc 3 4 tot

y1w xmg kg

Ž .Cerrado CE 30.8ab 57.3a 20.4a 47.6a 83.2a 239a 247aŽ .Degraded pasture DP 36.8ab 57.2a 21.5a 50.6a 64.9a 231a 242aŽ .Productive pasture PP 36.6ab 64.6ab 21.6a 49.7a 68.9a 241a 263a

Ž .Eucalyptus EU 34.6ab 60.5a 20.6a 41.2a 80.4a 237a 249a

()

J.Lilienfein

etal.r

Geoderm

a96

200031

–46

41

NaHCO -P NaHCO -P NaOH-P NaOH-P 1 M HCI-P HCl -P HNO rHClO -P Pa3 i 3 o i o conc 3 4 tot

y1w xmg kg

Ž .Pinus PI 42.9ab 55.9a 20.6a 35.9a 80.4a 236a 236aŽ .Conventional tillage SC 27.8b 62.9ab 21.0a 44.5a 84.4a 240a 242a

Ž .No-till maize–soybean NT 59.3a 74.5b 23.7a 51.8a 75.2a 284a 277aŽ .Conv. tillage maize–soybean CT 29.5ab 64.4ab 25.2a 51.3a 69.1a 239a 247a

Ž .c MolybdenumaNaHCO -Mo NaOH-Mo 1 M HCI-Mo HCI -Mo HNO rHCIO -Mo Mo3 conc 3 4 tot

y1w xmg kg

Ž .Cerrado CE nd 0.29a nd 3.01a 0.17a 3.46aŽ .Degraded pasture DP nd 0.24a nd 3.16a 0.12ab 3.53aŽ .Productive pasture PP nd 0.28a nd 2.83a 0.10b 3.21a

Ž .Eucalyptus EU nd 0.32a nd 3.07a 0.12ab 3.50aŽ .Pinus PI nd 0.40a nd 2.92a 0.09b 3.41a

Ž .Conventional tillage SC nd 0.26a nd 2.98a 0.11b 3.35aŽ .No-till maize–soybean NT nd 0.41a nd 3.24a 0.12ab 3.77a

Ž .Conv. tillage maize–soybean CT nd 0.44a nd 3.37a 0.10b 3.92a

aP , S , Mo ssum of the fractions of the extraction sequence.tot tot totbCNC-Ss total S concentration determined with a CNS-analyser.


extent in 0.5 M NaHCO . As a result of their high water solubility, aromatic3

ester sulphates are strongly leached from the soil and thus generally do notsignificantly contribute to the NaHCO -extractable S in soils. Thus, the first3

fraction of the sequential extraction scheme probably almost exclusively consistsŽ .of free sulphate. Alewell and Matzner 1996 have shown that percolation with

0.5 M NaHCO through a column filled with disturbed field-fresh soil material3

extracts free sulphate exhaustively. Their method corresponds to a 10-h batchextraction at a soil:solution ratio of 1:10. As we only batched the soil samplewith 0.5 M NaHCO for 30 min at a soil:solution ratio of 1:30, some sulphate3

may not have been removed with the first extract and thus be found insubsequent fractions. Except for the NaHCO fraction, all other reagents of the3

sequential extraction scheme mainly dissolve organic S compounds. However, itis impossible to assign clearly defined chemical S forms to these fractions.There is also no possibility to relate the chemical S fractions of the sequentialextraction procedure to the fractions gained by the traditional HI reductionŽ .C-bonded S and ester-S; Freney, 1961 .

Most of the Mo is extracted with the concentrated HCl fraction, followed bydecreasing proportions in the NaOH and HNO rHClO fractions. However, we3 4

did not detect any Mo in the plant-available NaHCO and dilute HCl fractions.3

We could not detect molybdate in the extracts because its concentrations were,like those of sulphate, lower than the detection limit of our analytical equipment.Even the tentative assignment of the extracted Mo in the various fractions tochemical Mo forms is difficult because little is known about chemical Mo forms

Ž .in soil. Shuman 1991 pointed out that there is nearly a complete lack ofselective Mo extractions.

The proportions of the NaHCO fraction considered to be directly plant-avail-3

able, on average, decrease along the line S)P)Mo, those of the medium-termplant-available NaOH fraction are similar for S and P and smaller for Mo. For

Ž .all studied soils, high proportions 50–90% of each element are detected in thestrongly bound fractions extracted with concentrated HCl and HNO rHClO ,3 4

Ž .consisting of P, S, and Mo occluded in Fe oxides. Friesen et al. 1997 extractedŽ .40–50% and Tiessen and Moir 1993 about 60% of total P in these recalcitrant

fractions, probably due to smaller concentrations of Fe oxides in the studiedcoarser-textured soils.

3.3. Land-use influence

The land-use influence is most apparent for P, it is smaller for S and Mo.Total P in NT and CT soils is 64% and 28% higher, respectively, than in CE dueto regular fertilisation. However, there are no differences in total P concentra-tions between SC and CE, because of the low fertilisation rate in SC and the

Žshort time the soils are used for surgarcane production. Pasture soils DP and.PP , in contrast, have about 12–15% smaller concentrations of total P than soils


in CE. This may be attributable to the removal by grazing if the assumption ofsimilar geogenic P concentrations is true.

The land-use influence is stronger on the plant-available P fractions extractedwith NaHCO and NaOH than on the more strongly bound fractions. In NT, CT,3

and EU, 104%, 68%, and 29% more P is extracted from soil in these fractions,respectively, than in CE. Phosphorus concentrations in the concentrated HClfraction of the NT and CT soils are higher than in CE soils. Although thedifferences are not significant, this may indicate the transformation of fertiliser P

Žinto less plant-available P fractions Friesen et al., 1997; Zhang and MacKenzie,.1997b . Phosphorus concentrations in the recalcitrant HNO rHClO fraction are3 4

not significantly different between the land-use systems. However, the soils inCE contain significantly higher P concentrations in this fraction than those in theother systems. Higher concentrations of recalcitrant P in CE may be explainedby geogenically higher P concentrations or may be the result of P depletion andsubsequent transformation of recalcitrant P to other P forms in all other systems.Similar P concentrations in the HNO rHClO fraction among all land-use3 4

systems except CE indicate similar geogenic P concentrations underlining theabove assumption that differences in total P concentrations between the pasturesoils and the other systems are attributable to removal by grazing.

The two plant-available fractions are the only ones where we detected organicP. The plant-available P fractions are higher in NT and CT than in CE. Thei

differences are significant between CE and NT only. Organically bound Pfractions, in contrast, do not differ significantly between the cropping systemsand CE. This resulted in a change of the P rP ratios. Whereas in CE about 70%i o

of P in NaHCO qNaOH-fractions is organic, the proportion in NT and CT is3Ž .-50%. Similar results have been reported by Beck and Sanchez 1994 , Friesen

Ž . Ž .et al. 1997 , and by Zhang and MacKenzie 1997a,b .Higher S concentrations are found in the NaHCO and NaOH fractions of NT3

than CE soils. The differences are significant for the NaOH fraction only.Concentrations of S in the dilute HCl fraction and of total S also tend to behigher in the cropping systems but differences are smaller than for P and notsignificant. This probably is attributable to a lower S fertilisation rate and to

Žstronger leaching of the applied S Mottavalli et al., 1993; Sakadevan et al.,.1993 . There are no significant differences in total S concentrations and in any S

fraction between CT and CE. This may be related with the lower S fertilisationrate in CT where no second crop is grown than in NT. Another reason may bestronger sulphate leaching in CT than in NT because CT is unvegetated for alonger period than NT, particularly at the end of the rainy season.

There are no significant differences in total Mo concentrations between thesystems. Like P, Mo concentrations in the HNO rHClO -fraction of the CE3 4

soils are significantly higher than in most of the other systems probably for thesame reasons as for P. In NT, NaOH-Mo concentrations in soil are slightlyhigher than in CE soils, however, the differences are not significant.


4. Conclusions

Total P and S concentrations in the studied Oxisols are low compared withconcentrations found in many surface soils of the temperate regions but compa-rable with concentrations reported for other Oxisols. The concentrations of Moare comparable with those reported for temperate soils. No data is available tocompare our results with Mo concentrations from other Oxisols. The chemicalfractions of P, S, and Mo in the studied Oxisols are element-specific and similar

Žin all studied systems. The proportions of the plant-available fractions NaHCO3.and NaOH fractions decrease along the line S)P)Mo. The results probably

reflect differences in the ratios of organically to inorganically bound P, S, andMo and differences in the sorption characteristics of the respective oxyanions.However, it is not possible to exactly assign to well-defined chemical speciesthe chemical fractions of the used extraction scheme. The value of the appliedmethod consists in the simultaneous extraction of the oxyanions of P, S, and Moand thus time and cost savings in characterizing chemical fractions of thesenutrients. From literature, it is known that the Hedley extraction scheme issuitable for the characterization of P plant availability. We infer that it is likelyto be also suitable for characterizing S and Mo plant availability while thetraditionally used HI reduction method for S separates C-bonded S from ester-Sbut does not allow for the determination of S plant availability.

Land-use impacts on chemical fractions decrease along the line P)S)Mo.ŽLand-use impacts on P and S concentrations in plant-available fractions NaHCO3

.and NaOH are most apparent. The impacts decrease as plant availabilitydecreases. The land-use influence on chemical fractions of P in our on-farmexperiment is similar to that reported from on-station experiments but differ-ences between the land-use systems are smaller. The 12–20 year-long use of thestudied Oxisols for cropping resulted in increased P and S concentrationswhereas those of Mo are almost unaffected.

Acknowledgements

ŽWe gratefully acknowledge the funding of the DFG Ze 154r36-1: CE, PI,. Ž .EU, NT, CT, SC and MAS Managing Acid Soils programme managed by

Ž .CIAT Centro Internacional de la Agricultura Tropical: DP, PP . We thank A.C.Frascoli, and I. Lobe for their support. We are indebted to the Bayreuther

¨ ¨Ž .Institut fur Terrestrische Okosystemforschung BITOK whose Central Analyti-¨cal Department performed parts of the chemical analyses. We also thank H.L.A.

Ž . ŽBessa, C.R. Cage Fazenda Planalto Hirofume , J.T. Fonseca Fazenda Ran-. Ž .charia Alegre , Getulio Guimaraes Fazenda Bomjardim , A. Mauro Lucinda´ ˜

Ž . Ž . ŽFazenda Pinusplan , H. Fuzaro Fazenda Passarinho , W. Ribeira da Sa Fazenda´. Ž .Beija Flor , and A.F. Santes Fazenda Estancia Recanto das Flores .


References

Alewell, C., 1993. Effects of organic sulfur compounds on extraction and determination ofinorganic sulfate. Plant Soil 149, 141–144.

Alewell, C., Matzner, E., 1996. Water, NaHCOy, NaH POy, and NaCl-extractable SO2y in acid3 2 4 4

forest soils. Z. Pflanzenernaehr. Bodenkd. 159, 235–240.Amberger, A., 1988. Pflanzenernaehrung, UTB, Stuttgart, 264 pp.Beck, M.A., Sanchez, P.A., 1994. Soil phosphorus dynamics during 18 years of cultivation on a

Typic Paleudult. Soil Sci. 34, 1424–1431.Castellano, S.D., Dick, R.P., 1990. Cropping and sulfur fertilization influence on sulfur transfor-

mation in soils. Soil Sci. Soc. Am. J. 54, 114–121.Curtin, D., Syers, J.K., 1990. Extractability and adsorption of sulfate in soils. J. Soil Sci. 41,

305–312.Freney, J.R., 1961. Some observations on the nature of organic sulfur compounds in soil. Aust. J.

Agric. Res. 12, 424–432.Freney, J.R., Melville, G.E., Williams, C.H., 1971. Organic sulphur fractions labeled by addition

of 35 g S-sulphate to soils. Soil Biol. Biochem. 3, 133–141.Friesen, D.K., Rao, I.M., Thomas, R.J., Oberson, A., 1997. Phosphorus acquisition and cycling in

crop and pasture systems in low fertility tropical soils. Plant Soil 196, 289–294.Ž .Gee, G.W., Bauder, J.W., 1996. Particle-size analysis. In: Klute, A.K. Ed. , Methods of soil

analysis: Part 1. Physical and mineralogical methods. Soil Science Society of America,Madison, Wisconsin, USA, pp. 383–411.

Goedert, W.J., 1983. Management of the Cerrado soils in Brazil: a review. J. Soil Sci. 34,405–428.

Hartung, J., Elpelt, B., 1989. Multivariate Statistik. Oldenbourg-Verlag, Munchen, 815 pp.¨Hedley, M.J., Stewart, J.W.B., Chauhan, B.S., 1982. Changes in inorganic and organic soil

phosphorus fractions by cultivation practices and by laboratory incubations. Soil Sci. Soc. Am.J. 46, 970–976.

Holmgren, G.G.S., 1967. A rapid citrate-dithionite extractable iron procedure. Soil Sci. Soc. Am.Proc. 31, 210–212.

Johnson, D.W., Todd, D.E., 1983. Relationship among iron, aluminum, carbon, and sulfate in avariety of forest soils. Soil Sci. Soc. Am. J. 47, 792–800.

Kilmer, V.J., Nearpass, D.C., 1960. The determination of available sulfur in soils. Soil Sci. Soc.Am. Proc. 24, 337–340.

Kliemann, H.J., 1987. Contribuicao ao estudo da disponibilidade de enxofre em solos brasileiros.˜ESALQ, PhD Thesis, Piracicaba, Brazil, 231 pp.

Ž .Landon, J.R. Ed. , 1991. Booker Tropical Soil Manual. Longman, London, Great Britain, 474 pp.Lilienfein, J., Wilcke, W., Neufeldt, H., Ayarza, M.A., Zech, W., 1998. Land-use effects on

organic carbon, nitrogen, and sulfur concentrations in macroaggregates of differently usedOxisols. Z. Pflanzenernaehr. Bodenkd. 161, 165–171.

McLaren, R.G., Naidu, R., Smith, J., 1998. Fractionation and distribution of arsenic in soilscontaminated by cattle dip. J. Environ. Qual. 27, 348–354.

Mottavalli, P.P., Duxbury, J.M., de Souza, D.M.G., 1993. The influence of organic soil amend-ments on sulfate adsorption and sulfur availability in a Brazilian Oxisol. Plant Soil 154,301–308.

Sakadevan, K., Hedley, M.J., Mackay, A.D., 1993. Sulphur cycling in New Zealand hill countrypastures: II. The fate of fertilizer sulphur. J. Soil Sci. 44, 615–624.

Sanchez, P.A., 1976. Properties and Management of Soils in the Tropics. New York, 618 pp.Schoenau, J.J., Stewart, J.W.B., Bettany, J.R., 1989. Forms and cycling of phosphorus in prairie

and boreal forest soils. Biogeochemistry 8, 223–237.


Schwertmann, U., 1964. Differenzierung der Eisenoxide des Bodens durch Extraktion mitAmmoniumoxalat-Losung. Z. Pflanzenernaehr. Dueng. Bodenkd. 105, 194–202.¨

Ž .Shuman, L.M., 1991. Chemical forms of micronutrients in soils. In: Mortvedt, J.J. Ed. ,Micronutrients in Agriculture. Soil Science Society of America, Madison, WI, USA, pp.113–144.

Smeck, N.E. et al., 1985. Phosphorus dynamics in soils and landscapes. Geoderma 36, 185–199.Smith, B.H., Leeper, G.W., 1969. The fate of applied molybdate in acidic soils. J. Soil Sci. 20,

246–254.Soil Survey Staff, 1997. Keys to Soil Taxonomy. 7th edn. Pocahontas Press, Blacksburg, VI,

USA, SMSS technical monograph no. 19.Sumner, M.E., Miller, W.P., 1996. Cation-exchange capacity and exchange coefficients. In:

Ž .Sparks, D.L Ed. , Methods of soil analysis: Part 3. Chemical methods. American Society ofAgronomy, Madison, WI, USA, pp. 1201–1229.

Tiessen, H., Moir, J.O., 1993. Characterization of available P by sequential extraction. In: Carter,Ž .M.R. Ed. , Soil Sampling and Methods of Analysis. CRC Press, Boca Baton, FL, USA, pp.

75–86.Zhang, T.Q., MacKenzie, A.F., 1997a. Changes of phosphorus fractions under continuous corn

production in a temperate soil. Plant Soil 192, 133–139.Zhang, T.Q., MacKenzie, A.F., 1997b. Changes of soil phosphorus fractions under long-term corn

monoculture. Soil Sci. Soc. Am. J. 61, 485–493.Zucker, A., Helger, H., Zech, W., 1986. Schwefelformen saurer Boden unter Nadelwald in¨

Nordostbayern. Z. Pflanzenernaehr. Bodenkd. 149, 110–119.

Chemical fractionation of phosphorus, sulphur, and molybdenum in Brazilian savannah Oxisols under...

Documents

Transcript of Chemical fractionation of phosphorus, sulphur, and molybdenum in Brazilian savannah Oxisols under...