chemical fractionation to characterize changes in sulphur and carbon in soil caused by management

$: Chemical fractionation to characterize changes in sulphur and carbon in soil caused by management$
Chemical fractionation to characterize changes insulphur and carbon in soil caused by management

B HUP INDER PAL - S I NGHa , M . J . H EDLEY

a , S . S AGGARb & G . S . F RANC I S

c

aInstitute of Natural Resources, Massey University, Private Bag 11222, Palmerston North, bLandcare Research, Private Bag 11052,

Palmerston North, and cCrop and Food Research, Private Bag 4704, Lincoln, Canterbury, New Zealand

Summary

Classical chemical fractionation of soil sulphur (S) into HI-reducible S and carbon-bonded S does not

separate S in soil into fractions that have differing mineralization potentials. Other techniques are needed

to separate organic S into more labile and less labile fractions of biological significance, irrespective of

their bonding relations. We have sequentially fractionated soil S and carbon (C) into their ionic forms

released onto ion-exchange resins and organic S and C extracted in alkali of increasing concentration.

We evaluated the technique on pasture and arable soils that had received various fertilizer and cultivation

treatments. Total S and C were greater in the soil of the fertilized pasture than in that of the unfertilized

pastures. Continuous arable cropping decreased total soil S and C, whereas restoration to pasture caused

an accumulation.

Resin, 0.1M NaOH, 1M NaOH and residual fractions accounted for between 1–13%, 49–69%, 4–16%

and 19–38% of total soil S and between 5–6%, 38–48%, 5–7% and 46–53% of total soil C, respectively.

Among different S and C fractions, the size of the 0.1M NaOH and residual fractions changed more with

the change in land use and management. The 0.1M NaOH fraction had a narrower C:S ratio (50–75:1) than

did the residual fraction (96–141:1). The significant degree of change in these two fractions, caused by

differences in land management, indicates that they may be useful indicators of change in ‘soil quality’.

Introduction

Maintaining and improving soil quality is essential if agricul-

ture and environmental quality are to be sustained. The

organic matter content of the soil has been widely used as a

measure of soil quality because it serves numerous functions

important to crop production and is an important reservoir of

the essential macronutrients such as nitrogen (N), sulphur (S)

and phosphorus (P). It also contributes significantly to the

formation and stabilization of soil structure. Losses in organic

matter content caused by intensive tillage and reduced input of

crop residues are well documented. Identifying changes in

organic matter content and quality is important in defining

sustainable agricultural practices over the longer term. Short-

term changes in the quantity of total organic matter in soil

induced by management are difficult to detect because a large

spatial variation obscures small changes in time. Measurement

of smaller, more labile fractions of organic matter in soil have,

however, been proposed to identify early changes caused by

management (Gregorich, 1996; Haynes, 2000).

The labile component of soil organic matter is a primary

source of mineralizable N, P and S, and therefore plays a

prominent role in nutrient dynamics and nutrient sustainabil-

ity of agricultural systems. As with C, changes in the contents

of these elements associated with the labile organic matter can

be important indicators of soil degradation or restoration.

More accurate estimates of ‘potentially’ dynamic fractions of

organic matter are required if changes in soil quality are to be

characterized over time. In contrast to N and P, information

on the effect of short- and long-term changes in land use and

management on labile soil organic matter fractions that supply

S to plants is scarce.

The classical chemical fractionation of S in soil according to

reactivity with reducing agents into HI-reducible S and

C-bonded S (Freney et al., 1969) does not characterize organic

fractions of S that have distinct mineralization potentials

(McLaren & Swift, 1977; Ghani et al., 1992). Some workers

have developed physical (Anderson et al., 1981) and a mixture

of chemical and physical (Bettany et al., 1979; Eriksen et al.,

1995) separation techniques to characterize the complex nature

of S in soil irrespective of its bonding relations to C. Again, theCorrespondence:Bhupinderpal-Singh.E-mail: [email protected]

Received 8 August 2002; revised version accepted 5 March 2003

European Journal of Soil Science, March 2004, 55, 79–90 doi: 10.1046/j.1365-2389.2003.00552.x

# 2003 Blackwell Publishing Ltd 79

fractionations were not entirely satisfactory, perhaps because

of the heterogeneity of S substrates in the organic matter. The

ideal solution to these problems might be provided by other

techniques that divide the organic S into a few biologically

meaningful labile fractions important for short- and long-term

supply of S to plants.

To characterize dynamic soil organic matter, Murata et al.

(1995) and Murata & Goh (1997) fractionated soils using

various pre-extractions such as shaking with cold water and

hot water followed by seven shakings with a mixture of acids

(0.1MHCl and 0.3MHF). They considered hot-water-extractable

C an indicator of labile organic matter, whereas HCl–HF-

extractable C represented an indicator of change in more stable

organic matter. They subsequently separated unextracted C into

fulvic acids, humic acids and humin using a series of five 0.1M

Na4P2O7 and three 0.5M NaOH extractions, and considered the

final residual C fraction (humin) to include a mixture of stable

organic forms suchashumic acidmaterials, andalso the remainsof

crop residues and dead microorganisms. By using this detailed,

multistep, fractionation, the scientists hoped to produce quality

indicators to compare the effects of (i) long-term superphosphate

application to irrigated pasture (Murata et al., 1995), and

(ii) cropping systems (Murata & Goh, 1997) on the quantity and

quality of soil organic matter. Long-term superphosphate applica-

tion to irrigated pasture increased hot-water-soluble C, but it had

no significant effects on the distribution of fulvic and humic acid

fractions of C. In the sequence of cropping and pasture rotations

studied by Murata & Goh (1997), the hot-water-soluble C

decreased as a proportion of the total C after 6 years of continuous

cultivation. In thepasturephase, however, over the sameperiod, all

C fractions – hot water soluble, HCl–HF, fulvic and humic acids

(Na4P2O7 or NaOH) and humin – increased in some soils. So

despiteusing thesepretreatments in a soil organicmatter extraction

scheme,Murata et al. (1995) andMurata&Goh (1997) found that

labile soil organic matter still appeared in both the extracted and

unextracted fractions.

The sequential extraction procedure for soil P, introduced by

Hedley et al. (1982), is being successfully used to characterize

changes in labile organic P fractions with change in soil

management and cropping (Tiessen et al., 1983; Huffman

et al., 1996) and plant uptake (Hedley et al., 1994). It relies on

differing strengths of reagents to extract soil organic phosphate

of differing lability. Recently, Lilienfein et al. (2000) also used a

sequential P fractionation, similar to that of Hedley et al.

(1982), to characterize the influence of different forms of land

use on P and S pools in savannahOxisols. They found that land

use influenced the amounts of P and S in the ‘plant available

fractions’ (0.5M NaHCO3, 0.1M NaOH). Nguyen & Goh

(1992) have reported that the alkaline solutions (0.5M

NaHCO3 (pH 8.5) and 0.1M NaOH) could extract consider-

able proportions of total soil S. Therefore these extractants

might be suitable for use in a sequential fractionation of S and

for assessing their ability to characterize the availability of S in

organic matter.

We have investigated whether a scheme similar to the P

fractionation of Hedley et al. (1982) can be used to character-

ize changes in organic S and C fractions on a range of pasture

and arable soils with varied management histories. If this was

possible then data on the dynamics of P and S in the soil could

be obtained simultaneously.

Materials and methods

Soil samples were obtained from three experimental sites

representing a range in fertilizer applications of S (pasture

soils, Mt Thomas), period of cultivation (Kairanga) and pasture

restoration crops (Wakanui) (Table 1). A detailed description of

these sites, experimental treatments, management history and

soils is presented elsewhere (Giddens et al., 1995; Francis et al.,

1999; Shepherd et al., 2001). Table 1 summarizes the soil types,

agronomic treatments and fertilizer applications, and we briefly

describe the more important aspects below.

Description of field sites

Mt Thomas site (fertilized pasture). The fertilizer treatments

on this initially S-deficient permanent pasture were SO42–-S

applied as single superphosphate (SSP) or SSP and gypsum

(SSP–G), a mix of sulphate and elemental S in ‘sulphur super

extra’ (SSX), and elemental S0 granulated with bentonite

(Ben28, Ben56) applied tri-annually over 6 years (Table 1).

There were four replicates of each fertilizer treatment. Soil

samples (0–7.5 cm depth) were collected with a 2.5-cm diameter

probe with 20 soil cores from each replicate, which were bulked.

All S treatments produced marked growth responses in legume

yield in the first 4 years of the trial (Craighead et al., 1990).

Kairanga site (short- to long-term cultivation). This site car-

ried a permanent pasture (PP) and several continuous cultivation

systems with different times since last in pasture, i.e. 4 years of

maize cultivation (M4Y), 11 years of maize cultivation (M11Y)

and 30years of barley cultivation (B30Y). Also included was a

site that had been ploughed for 11 years for maize and then

returned to permanent pasture for 10years (M11YþP10Y)

(Table 1). Soil samples were taken from 0–10 cm depth of soil

(from each paddock) with a 2.5-cm diameter probe.

Wakanui site (short-term pasture cropping). This site had a

short-term mixed cropping trial, which included a range of

permanent pastures, annual pastures and arable crops started

in 1989 on a Wakanui silt loam soil after 11 years of arable

cropping (Francis et al., 1999). Four treatments providing dif-

ferences within and between the annual (ryegrass grazed and

conventionally cultivated, ARc; ryegrass grazed and direct

drilled, ARd) and perennial (ryegrass grazed, PRg; ryegrass

mown, PRm) pastures (Table 1), replicated four times, were

included in the present study because different crops in the

trial confound any other comparisons. In the sixth year of the

80 Bhupinderpal-Singh et al.

# 2003 Blackwell Publishing Ltd, European Journal of Soil Science, 55, 79–90

Table1Soilclassificationanddescriptionoftreatm

ents

fortheMtThomas,KairangaandWakanuisites

Soilgroupa

Applicationrate

/kgha�1application�1

Soilname

(Application)

New

Zealand(G

enetic)

FAO

Treatm

ent

Treatm

entcode

SO

42–

S0

MtThomas

Allophanic

brownsoil

DystricCambisol

Control

Ct

00

siltloam

(HighcountryYellow

Single

super

phosphate

andgypsum

SSP–G

28

0

(Triannual)

BrownEarth)

Bentonitic

S0

Sulphursuper

extra

Bentonitic

S056

Sulphursuper

extra(reapplied)

Ben28

SSX

Ben56

SSXr

0

19 0

27

28

37

56

91

Kairangasilty

Typic

orthic

gleysoil

Eutric

Gleysol

Permanentpasture

PP

22b

0b

clayloam

(RecentGley)

Maize4years

M4Y

35b

0b

(Annual)

Maize11years

Barley

30years

Maize11years

followed

by10years

pasture

M11Y

B30Y

M11YþP10Y

NA

b

NA

b

44b

NA

b

NA

b

0b

Wakanui

Immature

pallic

soil

Gleyic

Cambisol

Annualryegrass

grazedandconventionallycultivated

ARc

17

0

siltloam

(Yellow

GreyEarth)

Annualryegrass

grazedanddirectdrilled

ARd

17

0

(Annual)

Perennialryegrass

grazed

Perennialryegrass

mown

PRg

PRm

17

17

0 0

aHew

itt(1992).

bTheSregim

esgiven

are

those

applied

only

duringthesamplingyear.AlthoughShasbeenpreviouslyapplied

inthesetreatm

entseither

asammonium

sulphate

orasSSP,thedata

forthose

applicationswerenotavailable

from

thefarm

ers.

NA,notavailable.

Characterization of soil organic sulphur and carbon 81


trial, we sampled the topsoil (0–5 cm depth) by taking 10 cores

from each replicate using a 2.5-cm diameter probe.

Preparation of soil samples

Soil sampling depths for the root zone of each site varied for

each site as they were dictated by the previous researchers (see

above). Comparisons of C and S content in the soil across

different treatments were therefore restricted to within a single

site, where the depth of soil sampled was the same.

From these soil samples, organic material such as visible

roots was removed. The soils were sieved (< 2mm) and air-

dried at 30�C� 2�C. The air-dried samples of soil were then

finely ground for 10 s in a ring grinder (Rocklabs,Auckland,New

Zealand) for fractionation and other soil analyses. Unless other-

wise stated, all soil samples were analysed in duplicate in the

various extractions described below.

A sequential soil S and C fractionation technique

The procedure used to fractionate soil S and C sequentially is

summarized in Figure 1. It is similar to that of Hedley et al.

(1982) for soil P. As in that fractionation, the initial design of

the sequential fractionation was to determine first the

exchangeable ionic forms of S and C in the soil (both soluble

and adsorbed and simple organic esters) extracted on to ion-

exchange resins [resin-S(C)]. Strips of both anion- and cation-

exchange resin membranes were used because in soils of

variable charge, extraction of anions may be more effective

when cations contributing to surface positive charge are

removed (Saggar et al., 1992). Solid-phase non-exchangeable

forms of S and C were extracted in alkali extractions of two

strengths. Hence, conceptually ‘labile’ organic S and C fractions

would be extracted by 0.1MNaOH [0.1MNaOH-S(C)] thatwould

meet medium- to long-term plant nutrient availability, and less

‘labile’ fractions extracted by 1M NaOH [1M NaOH-S(C)]

(Hedley et al., 1982). The S and C fractions not extracted with

either resin (representing short-term nutrient availability) or the

NaOH extractants were termed as the residual fraction [residual-

S(C)]. The initial concept viewed the unextracted residual fraction

as a recalcitrant, chemically stable fraction.

The fractionation procedure for C in the soil was slightly

modified from that of S (Figure 1) in the following way. The

anion and cation resins were charged as in the procedure

described by Saggar et al. (1990), except that the anion-

exchange resin was charged as OH– with 0.1M NaOH instead

Soil

Elution in 10 mlResin-S(C)

2 g soil (ring ground) in 40 ml polypropylenecentrifuge tubes, add 18 ml deionized water and resins (both cation and anion). Shake for 2 hoursat room temperature.

Soil residue+ water

Ion-exchange resin strips

of 0.5 M NaCl or0.5 M NaNO3

0.1 M NaOH-S(C)

1 M NaOH-S(C)

Add 2 ml of 1 M NaOH (to generate 0.1 M NaOH), shake for 16 hours at room temperature, centrifuge and filter supernatant through Whatman No 42 filter paper or glass microfibre filter (GF,C).

Add 20 ml of 1 M NaOH, shake for 16 hours at room temperature, centrifuge and filter supernatantthrough Whatman No 42 filter paper or glass microfibre filter (GF,C).

Residual-S(C)

Dry the soil residue at 35°C and ring grind.

Remove strips, rinse with deionized water and elute them together.

Filtrate

Soil residue

Soil residue

Filtrate

Figure 1 A sequential fractionation technique for soil S and C.



of HCO3– resin (i.e. by using 0.5M NaHCO3). This would avoid

any interference in the estimation of C from HCO3– ions. Both

cation and anion resins were eluted in 0.5M NaNO3 (this eluted

the same amount of C from the cation- and anion-exchange resins

as was eluted in 0.5M NaCl) to avoid interference by the Cl– ion

during dichromate digestion for the estimation of total C in the

extract (see below). Also, alkaline extracts were filtered through

Whatman glass microfibre (GF,C) so that no cellulose came out in

the alkaline extracts from paper filters (Whatman No 42).

Analysis of S in different fractions

Resin-extractable S. We analysed an aliquot of 0.5M NaCl

or NaNO3 elutant for resin-S by reducing S to H2S using HI

mixture and measuring it as bismuth sulphide turbidimetric

finish (Dean, 1966) using a Technicon II autoanalyser. Any

C-bonded S in this extract would not be measured.

Sulphur in NaOH extracts. The alkaline extracts were

digested with freshly prepared sodium hypobromite (NaOBr)

as in the procedure described by Tabatabai & Bremner (1970).

However, the amount of NaOBr used in this digestion was

4ml instead of the 3ml recommended in the original method.

The greyish digests were redissolved in deionized water and

analysed for SO42–-S on an autoanalyser as described above

for resin-S.

Total S in soil and residual fraction. The total S in soil and

residual fraction was oxidized to SO42–-S by the ignition

method of Steinberg et al. (1962). The digests were analysed

for SO42–-S on an autoanalyser as described above.

Analysis of C in different fractions

Total C in soil samples was determined by combustion on a

Leco FP-2000 CNS analyser. The oxidation and digestion

technique of Amato (1983) as described by Sparling et al.

(1991) was used to determine C in the 0.5M NaNO3 elutant,

0.1M and 1M NaOH extracts and the residual fraction of

Kairanga soil.

Chloride (Cl–) interference in the resin extract

Extracts from Kairanga soils contained various concentrations

of C. However, Cl– interfered (Quinn & Saloman, 1964) during

the dichromate digestion (for total C determination, Sparling

et al., 1991) of resin extracts from the Kairanga soils. A recov-

ery of 99.5% C of a glucose spike was obtained for Cl–-free

extracts. The interference effect from chloride was determined

to be linearly related to chloride concentration by spiking a

standard glucose solution with different concentrations of

chloride (0, 0.5, 1, 2, 3, 4 milliequivalent). A linear relation

was used to correct any chloride interference. The Cl– concen-

tration in a sample was determined with chloride analyser.

Statistics

The significance of a linear relation between amounts of S and

C in the soil and in extracts of soil (referred to as S and C

fractions) was expressed by the product-moment correlation

coefficient. Simple linear regression was used to determine the

rate of change of different soil S (or C) fractions with changes

in total soil S (or C) induced by management. The F test was

used to test the significance of difference between the rates of

change (slopes). The significance of differences between

treatment means in each replicated experiment (i.e. Mt Thomas

and Wakanui sites) was tested (at 5% level of significance) by

analysis of variance (ANOVA), and the treatment means for each

site were separated by Tukey’s Honest Significant Difference

procedure.

Results and discussion

Changes in total soil S and C

Improved permanent pasture. Fertilizer-S applied to the

S-deficient pasture soil at Mt Thomas over 6 years significantly

increased the amount of S in the soil (Figure 2). The greatest

accumulation of total S was in the SSXr treated soil, and the

amount decreased in the order of Ben56> SSX¼ SSP–G

>Ben28>Ct (Figure 2). This indicates a better conservation

of elemental S0 into organic S or an accumulation of residual

elemental S0. There was a concurrent increase in the total C

content with increasing rates of fertilizer-S (Figure 2). However,

the relationship (r¼ 0.43) between total soil S and C with

increasing rates of S application was not significant because

although SSXr treatment resulted in considerably more accu-

mulation of S than did the Ct treatment, the corresponding

increase in C was small. Nguyen & Goh (1990) also found

significantly greater organic C in the soil of irrigated, grazed

pastures receiving long-term application of superphosphate

than in the unfertilized pasture soil. They attributed this

increase in C to an increased biological fixation of N2 by

legumes that were no longer deprived of S. This led in turn to

an increase in pasture dry matter and hence greater organic C

input to soil. At the Mt Thomas site, S fertilization markedly

increased pasture production (Craighead et al., 1990), and this

has resulted in most of the applied S accumulating in organic

forms. After many years of pasture improvement, the rate of

accumulation of organic S in the soil decreases, and mineraliza-

tion becomes a significant source of S for pasture plants

(Sakadevan et al., 1993). Accumulated organic S therefore

improves the fertility and economic value of a pasture soil.

Cropping and pasture restoration phases. At the Kairanga

site (short- to long-term cultivation) the soil on the PP farm

contained more total S and C than the soils cultivated for

maize (M) or barley (B) (Figure 2). There was a greater

depletion of total S and C in the medium- to long-term culti-

vation treatments (M11Y and B30Y) than in the short-term



cultivation (M4Y). Decreases in soil S were significantly

(P< 0.01) related to decreases in soil C (r¼ 0.98). A similar

significant linear relationship between the change in the total

soil S and C was also observed in the short-term cropped

Wakanui site (Figure 2). The Kairanga soil, however, had a

narrower C:S ratio (67–83:1) compared with the wider C:S

ratio (93–105:1) of the Wakanui soil (Figure 2).

Long-term cultivation had considerably decreased organic

S and C in the soil, probably as a result of enhanced decompos-

ition of organic matter and decreased organic matter inputs

under cropping (Saggar et al., 2001). We think the enhanced

decomposition is due to mechanical stress breaking aggregates

bound by organic matter and exposing the organic matter to

microbial decomposers, which leads to faster mineralization in

warmer soil. Murata & Goh (1997) also found that increased

period of cultivation and cropping decreased total C and total

N, whilst under pasture following cultivation, total C and

total N, and aggregate stability, of soils increased with time

(Shepherd et al., 2001). We found in the present study that

conversion of 11-year arable cropping sites (Kairanga or

Wakanui) to perennial pasture management for 10 years

(Kairanga, compare M11Y with M11YþP10Y) or 6 years

(Wakanui, PRg and PRm treatments) caused a greater accu-

mulation of total S and C in the soil than in the annually

cultivated pasture (ARc and ARd treatments) (Figure 2).

The comparison of unit change in total S and C across all

the three sites (Figure 2) showed that changes in total S

content in relation to unit changes in C were not statistically

different. This suggests that although management changes

considerably, changes in S in the soil follow changes in C.

Sulphur and carbon fractions sensitive to changes in

management practices

The chemical fractionation appeared to recover between

90 and 115% of the total S in the soil and between 102

and 110% of the total soil C. Apparent recoveries greater

than 100% presumably result from either summation of

errors in the fractionation of total S or underestimation of

soil total S content by the NaHCO3–AgO oxidation (i.e. sum of

S fractions> total S).

In all the three experimental sites (Table 1), 0.1M NaOH-S

constituted the major portion (49–69%) of total soil S, fol-

lowed by residual-S (19–38%), 1M NaOH-S (4–16%) and

resin-S (1–13%) (Figures 3, 4a and 5). However, in the

Kairanga soil, which was also subjected to sequential C frac-

tionation, the major portion of total C was recovered in resi-

dual-C (46–53%) followed by 0.1M NaOH-C (38–48%), 1M

NaOH-C (5–7%) and resin-C (5–6%) (Figure 4b).

Resin-S, which constituted only a small part of total soil S

(1–13%), was greater in the recently fertilized pasture soils

(Mt Thomas) than in the soils under short- and long-term

cultivated (Kairanga) and short-term pasture cropping (Wakanui)

(cf. Figures 3, 4a and 5). This might reflect the fact that the soil at

6

Total soil C /%

200

300

400

500

600

700

To

tal s

oil

S /m

g k

g–1

PP

M4Y

M11Y+P10Y

B30YM11Y

PRmPRg

ARc

ARd

SSXr

Ben56

SSP/GSSX Ben28Ct

0 2 3 4 50

Mt Thomas, r = 0.43

Kairanga, r = 0.98

Wakanui, r = 0.99

Figure 2 Relations between total soil S and soil C

as influenced by varied management. The total S

at theWakanui site is the sum of all the S fractions

recovered by the sequential fractionation.

Treatment codes are given in Table 1.



Mt Thomas (high country Yellow Brown Earth) has a larger

anion exchange capacity (% phosphate retention, see Metson,

1979) than the Kairanga (Recent Gley) or Wakanui silt loam

(Yellow Grey Earth), and so more SO42� can be retained on the

greater hydrous oxide surfaces.

Improved permanent pasture. In fertilized pasture soils (Mt

Thomas), S in all the fractions increased with the increase in

total S in soil from increasing amounts of fertilizer-S (Figure 6).

The rate of increase in 0.1M NaOH-S was significantly greater

(P< 0.001) than that in residual-S, and both rates of increase

were significantly (P< 0.001) greater than increases in resin-S

and 1M NaOH-S (Figure 6). When applied at the same rate, the

different forms of fertilizer had no significant effect on any

particular fraction of S, except resin-S (Figure 3). For example,

treatment Ben56 increased resin-S significantly in excess of that

obtained from SSX treatment (Figure 3) although these two

sources supplied equal amounts of S. When applied at the

different rates (cf. 56 against 112 kgSha�1, Figure 3), Ben56

increased 0.1M NaOH-S significantly more than that obtained

from SSP–G and Ct treatments. However, treatment SSXr

resulted in significantly larger resin-S and 0.1M NaOH-S than

did the other treatments. The amount of S in the residual

fraction was significantly greater in the SSXr than in the Ct

treatment (Figure 3). This may be attributed to the unoxidized

elemental S0 that remained from fresh top dressings. The 1M

NaOH-S was not influenced significantly by the various fertil-

izer treatments (Figure 3).

Cropping and pasture renovation phases. At the Kairanga

sites, cultivation (M4Y, M11Y or B30Y) caused 0.1M NaOH-

S, 1M NaOH-S and residual-S in soils to decrease. However,

resin-S increased in theM4YandM11Y treatments, anddecreased

in the B30Y treatment and also when the M11Y site was restored

back to pasture for 10 years (M11YþP10Y) (Figure 4a). In

terms of proportional change, 0.1M NaOH-S was 50% smaller

at theM11YorB30Y sites than at the PP site. Restoration of the

M11Y site with pasture (M11YþP10Y), however, markedly

increased 0.1M NaOH-S to within 12% of that at the original

PP site. Similarly, residual-S in the M11Y and the B30Y was

33% and 50%, respectively, of that in the PP site. When the

cultivated soil (M11Y) was returned to pasture (M11YþP10Y)

the increases in the residual-S fraction were smaller (i.e.

increased from 33% to 47% of that in the PP site) relative to

changes in the 0.1MNaOH-S fraction (Figure 4a). Trends in the

change of 1M NaOH-S, with the change in duration of cultiva-

tion, were irregular and changed little when the cultivation site

(M11Y) was restored to pasture (M11YþP10Y). The rates of

change in both 0.1MNaOH-S and residual-S with the change in

total S were similar and were significantly greater (P< 0.01)

than the rates of change in resin-S and 1M NaOH-S (data not

shown).

All C fractions were smaller in the cultivated (M4Y, M11Y

or B30Y) soils than in pasture soils (PP). Short-term

cultivation (M4Y) was associated with marked reductions in

0.1M NaOH-C and residual-C fractions and small reductions

in resin-C and 1M NaOH-C (Figure 4b). In the medium- to

long-term cultivated treatments (M11Y or B30Y), however, all

C fractions were about one-third less than in the PP phase. The

pasture restoration phase (M11YþP10Y) contained twice as

much 0.1M NaOH-C than the medium-term cultivated site

(M11Y), but smaller differences existed in resin-C, 1M

NaOH-C and residual-C (Figure 4b). The rates of change in

both 0.1M NaOH-C and residual-C with the change in total

soil C were similar and were significantly (P< 0.001) greater

than the rates of change in resin-C and 1M NaOH-C (data not

shown).

0

100

200

300

400

500

600

Ct SSP–G Ben28 SSX Ben56 SSXr

S fertilizer treatments

S e

xtra

cted

/mg

S k

g–1

so

il

Resin 0.1 M NaOH 1 M NaOH Residual

b

a

d

c

ab ab ab ab

a

a

a aa a

a

ac c b

bbccd bc

56 kg S ha–1 112 kg S ha–1

236 kg S ha–1

c

Figure 3 Effect of S fertilizer application to

pasture on different S fractions in the 0–7.5 cm

depth of Mt Thomas soil. Treatment codes are

given in Table 1. Error bars: one standard error

(SE) of treatment means. Comparing across

fertilizer treatments, fractions in the figure

followed by the same letter (a, b, c, d) are not

significantly different at P< 0.05.



Comparison of the amounts of soil S and C recovered in the

0.1M NaOH and residual fractions indicates that in both these

fractions, losses and gains of S and C with the change in

cultivation practices (Figure 7) were significantly correlated

(P< 0.05).

At the Wakanui site, the short-term annual and perennial

pasture cropping caused insignificant changes in the size of 1M

NaOH-S and resin-S (Figure 5). Within either annual (ARc,

ARd) or perennial (PRm, PRg) pasture treatments, the size of

0.1M NaOH-S fractions were not significantly different. This

contrasts with residual-S (Figure 5). Between annual and

perennial treatments there were differences in S fractions.

The 0.1M NaOH-S fraction was significantly smaller in the

ARc treatment than in the PRm treatment only, whereas

residual-S in the ARc treatment was significantly smaller

than in the other treatments (ARd, PRm, PRg). The largest

amounts of 0.1M NaOH-S and residual-S were in the PRm

and PRg treatments, respectively. The rate of change in 0.1M

NaOH-S with the change in total soil S was significantly

(P< 0.01) greater than that in residual-S and both rates of

change were significantly greater (P< 0.001) than changes in

resin-S and 1M NaOH-S (data not shown).

The above-mentioned results from the three sites indicate that

the 0.1M NaOH and residual fractions changed more than

other fractions as pastoral or cropping management changed

in the short or long term. They show that the residual fraction,

which represents the soil organic matter not extracted with

alkali (humin fraction), also contains a relatively labile organic

S and C. McLaren & Swift (1977) also found that residual

humin fraction, which is often regarded as being fairly inert,

cannot in the long term be neglected as a potential source of

mineralizable S. Using ‘bomb’ 14C enrichment, Goh et al. (1976)

also showed that the residual humin fraction contained both old

(i.e. stable) and young (i.e. labile) forms of soil organic matter.

Short- to long-term cultivation

M4Y M11Y B30Y M11Y+P10Y

PP

PP

M4Y M11Y B30Y M11Y+P10Y

(a)

(b)

0

100

200

300

400

500

600

700

0

10

20

30

40

50

60

70

S e

xtra

cted

/mg

S k

g–1

so

ilC

ext

ract

ed /g

C k

g–1

so

il


Figure 4 Effect of short- to long-term cultivation

on (a) different S and (b) different C fractions

in the 0–10 cm depth of Kairanga soil. Treatment

codes are given in Table 1. Error bars: 1 SE of

sample means.



The 0.1M NaOH fraction had a narrower C:S ratio (50–75:1)

than the wider C:S ratio (96–141:1) of the residual fraction

(Figure 7). This is because the 0.1M NaOH recovered a greater

proportion of total S than did the residual fraction (Figure 7),

thus suggesting that 0.1M NaOH could have hydrolysed a

greater proportion of S in protein, and therefore resulted in a

0

100

200

300

400

ARc ARd PRm PRg

Short-term pasture cropping

a a a a

a ababb

aa

a a

ab

bc

S e

xtra

cted

/mg

S k

g–1

so

il


Figure 5 Effect of short-term pasture cropping

on different S fractions in the 0–5 cm depth of

Wakanui soil. Treatment codes are given in

Table 1. Error bars: 1 SE of treatment means.

Comparing across pasture cropping treatments,

fractions in the figure followed by the same

letter (a, b, c) are not significantly different at

P< 0.05.

350 400 450 500 550 650600 700

Total soil S /mg kg–1soil

0

50

100

150

200

250

300

350

400

S e

xtra

cted

/mg

S k

g–1

so

il

0

0.1 M NaOH-S, y = 0.42x + 88.31, R

2 = 0.81

Residual-S, y = 0.25x – 26.99, R

2 = 0.70

Resin-S, y = 0.10x + 0.88, R

2 = 0.72

1 M NaOH-S, y = 0.05x + 7.30, R

2 = 0.24

Increasing fertilizer S

Figure 6 Changes in amounts of S fractions, y,

as total soil S, x, increases from increasing

amounts of fertilizer application to the

permanent pasture soil (Mt Thomas) over

6 years.



narrower C:S ratio of the 0.1M NaOH fraction. The residual

fraction probably included remains of undecomposed or partly

decomposed roots and crop residues, and dead microbial tissues

rich in cellulose and lignin (Murata & Goh, 1997) that are not

soluble in alkali (Lowe, 1978). The residual fraction also

includes a mixture of complex organic forms (humic acid),

which could be strongly complexed with clays and hydrous

oxides (Schulten & Schnitzer, 1997), and were not extracted

even with strong alkali. However, Hatcher et al. (1985), using

solid-state 13C-NMR, found that the spectra of humin (the

alkali-insoluble fraction) differed dramatically from those of

their respective humic acids, suggesting that humin is not a

clay–humic acid complex. They also showed the presence of

polysaccharide, aromatic (lignin) and paraffinic C in humin.

Clearly, there is a need to examine the origin of C appearing

in the residual fraction.

The amounts of S and C in the 1M NaOH fraction changed

little with change in management and contribute only a very

small proportion of the total S and C in the soil. The 1M

NaOH fraction might represent relatively humified organic

matter. The contribution of this fraction to available S in the

soil might be small and therefore of little value in predicting

the S requirement of plants.

Conclusions

Pastures, renowned for increasing soil organic matter and

aggregate stability, increased both S and C content of the

soil when S-deficient permanent pasture was fertilized with S

or when previously tilled soils were returned to pasture.

Our study has shown that the P fractionation scheme

extracts from soils S and C fractions that vary in their response

to change. Under conditions of organic S and C accumulation

or depletion, both the 0.1M NaOH fraction (representing

49–69% soil S and 38–48% soil C) and the residual fraction

(representing 19–38% soil S and 46–53% soil C) showed sig-

nificant changes and can be used as indicators of the quantity

and quality of the soil organic matter.

The novelty of our study is that S in the soil has been

determined in fractions usually used to characterize soil P.

We could not, however, separate organic S and C into

fractions of decreasing lability by sequential extractions with

alkali of increasing concentration as Hedley et al. (1982) did

for P. The reason for this difference in the fractionation of S

and P is perhaps their presence in different compounds in the

organic complexes within the soil organic matter. As the

residual-S (or C) appeared to contain both labile and stable

fractions of organic matter, in subsequent research we shall

examine how C in recent plant root material is distributed

amongst the fractions of this sequential extraction procedure.

Acknowledgements

We thank the farmers who gave us access to their land,

T.G. Shepherd (Landcare Research, New Zealand) and

M. Craighead (Ravensdown Fertilizer Co-op. Ltd) for supplying

soil samples, R. Webster, Soren Holm and Anne West for

providing valuable comments and checking statistics on an

earlier version of the script, and the New Zealand Ministry

of Foreign Affairs and Trade for awarding a NZODA–PGS

fellowship to Bhupinderpal-Singh to study for a PhD.

0

50

100

150

200

250

300

350

400

0 5000 10 000 15 000 20 000 25 000 30 000

Amount of soil C in the fractions /mg kg–1

Am

ou

nt

of

soil

S in

th

e fr

acti

on

s /m

g k

g–1

0.1 M NaOH r = 0.92

Residual r = 0.95

M11Y+P10Y

M11Y+P10Y

M11Y

M11Y

B30Y

M4Y

M4Y

PP

PP

B30Y

Figure 7 Comparison of amounts of S and C

recovered in 0.1M NaOH and residual fractions

from 0–10 cm depth of Kairanga soil. Treatment

codes are given in Table 1.



References

Amato, M. 1983. Determination of carbon 12C and 14C in plant and

soil. Soil Biology and Biochemistry, 15, 611–612.

Anderson, D.W., Saggar, S., Bettany, J.R. & Stewart, J.W.B. 1981.

Particle size fractions and their use in studies of soil organic matter: I.

The nature and distribution of forms of carbon, nitrogen and sulfur.

Soil Science Society of America Journal, 45, 767–772.

Bettany, J.R., Stewart, J.W.B. & Saggar, S. 1979. The nature and

forms of sulphur in organic matter fractions of soils selected along

an environmental gradient. Soil Science Society of America Journal,

43, 981–985.

Craighead, M.D., Burgess, W.B., Clark, S.A. & Duffy, R.G. 1990.

Development of sunny-facing high country using different forms of

sulphur fertilizer. Proceedings of the New Zealand Grassland

Association, 52, 203–206.

Dean, G.A. 1966. A simple colorimetric finish for the Johnson–Nishita

micro-distillation of sulphur. Analyst, 91, 530–532.

Eriksen, J., Lefroy, R.D.B. & Blair, G.J. 1995. Physical protection of

soil organic S studied using acetylacetone extraction at various

intensities of ultrasonic dispersion. Soil Biology and Biochemistry,

27, 1005–1010.

Francis, G.S., Tabley, F.J. & White, K.M. 1999. Restorative crops for

the amelioration of degraded soil conditions. Australian Journal of

Soil Research, 37, 1017–1034.

Freney, J.R., Melville, G.E. & Williams, C.H. 1969. Extraction,

chemical nature, and properties of soil organic sulphur. Journal of

the Science of Food and Agriculture, 20, 440–445.

Ghani, A., McLaren, R.G. & Swift, R.S. 1992. Sulphur mineralisa-

tion and transformations in soils as influenced by additions of

carbon, nitrogen and sulphur. Soil Biology and Biochemistry, 24,

331–341.

Giddens, K.M., Saggar, S., Craighead, M. & Searle, P.L. 1995.

Evaluation of three recently developed soil sulphur tests in relation

to pasture response. In: Proceedings of the Workshop on Fertiliser

Requirements of Grazed Pastures and Field Crops: Macro- and

Micro-nutrients (eds L.D. Curie & P. Loganathan), pp. 182–189.

Massey University, Palmerston North.

Goh, K.M., Rafter, T.A., Stout, J.D. & Walker, T.W. 1976. The

accumulation of soil organic matter and its carbon isotope content

in a chronosequence of soils developed on aeolian sand in New

Zealand. Journal of Soil Science, 27, 89–100.

Gregorich, E.G. 1996. Soil quality: a Canadian perspective. In: Soil

Quality Indicators for Sustainable Agriculture in New Zealand:

Proceedings of a Workshop (eds K.C. Cameron, I.S. Cornforth,

R.G. McLaren, M.H. Beare, L.R. Basher, A.K. Metherell &

L.E. Kerr), pp. 40–52. Lincoln Soil Quality Research Centre,

Lincoln University, New Zealand.

Hatcher, P.G., Berger, I.A., Maciel, G.E. & Szeverenyi, N.M. 1985.

Geochemistry of humin. In: Humic Substances in Soil, Sediment and

Water: Geochemistry, Isolation and Characterization (eds G.R. Aiken,

D.M. McKnight, R.L. Wershaw & P. MacCarthy), pp. 275–302.

Wiley-Interscience, New York.

Haynes, R.J. 2000. Labile organic matter as an indicator of organic

matter quality in arable and pastoral soils in New Zealand. Soil

Biology and Biochemistry, 32, 211–219.

Hedley, M.J., Stewart, J.W.B. & Chauhan, B.S. 1982. Changes in

inorganic and organic soil phosphorus fractions induced by

cultivation practices and by laboratory incubations. Soil Science

Society of America Journal, 46, 970–976.

Hedley, M.J., Kirk, G.J.D. & Santos, M.B. 1994. Phosphorus

efficiency and the forms of soil phosphorus utilised by upland rice

cultivars. Plant and Soil, 158, 53–62.

Hewitt, A.E. 1992. New Zealand Soil Classification. Landcare Research

Science Series No 1, Manaaki-Whenua Press, Lincoln, New Zealand.

Huffman, S.A., Cole, C.V. & Scott, N.A. 1996. Soil texture and

residue addition effects on soil phosphorus transformations. Soil

Science Society of America Journal, 60, 1095–1101.

Lilienfein, J., Wilcke, W., Ayarza, M.A., Vilela, L., Lima, S.D.C. &

Zech, W. 2000. Chemical fractionation of phosphorus, sulphur, and

molybdenum in Brazilian savannah Oxisols under different land

use. Geoderma, 96, 31–46.

Lowe, M. 1978. Carbohydrates in soil. In: Soil Organic Matter (eds

M. Schnitzer & S.U. Khan), pp. 65–93. Elsevier, Amsterdam.

McLaren, R.G. & Swift, R.S. 1977. Changes in soil organic sulphur

fractions due to the long term cultivation of soils. Journal of Soil

Science, 28, 445–453.

Metson, A.J. 1979. Sulphur in New Zealand soils. 3. Sulphur distribu-

tion in soils derived from parent materials of predominantly sediment-

ary origin. Report 41, New Zealand Soil Bureau, Wellington.

Murata, T. & Goh, K.M. 1997. Effects of cropping systems on soil

organic matter in a pair of conventional and biodynamic mixed

cropping farms in Canterbury, New Zealand. Biology and Fertility

of Soils, 25, 372–381.

Murata, T., Nguyen, M.L. & Goh, K.M. 1995. The effects of long-

term superphosphate application on soil organic matter content and

composition from an intensively managed New Zealand pasture.

European Journal of Soil Science, 46, 257–264.

Nguyen, M.L. & Goh, K.M. 1990. Accumulation of soil sulphur

fractions in grazed pastures receiving long-term superphosphate

applications. New Zealand Journal of Agricultural Research, 33,

111–128.

Nguyen, M.L. & Goh, K.M. 1992. Status and distribution of soil

sulphur fractions, total nitrogen and organic carbon in camp and

non-camp soils of grazed pastures supplied with long-term

superphosphate. Biology and Fertility of Soils, 14, 181–190.

Quinn, J.G. & Saloman, M. 1964. Chloride interference in the dichro-

mate oxidation of soil hydrolysates. Soil Science Society of America

Proceedings, 28, 456.

Saggar, S., Hedley, M.J. & White, R.E. 1990. A simplified resin

membrane technique for extracting phosphorus from soils. Fertilizer

Research, 24, 173–180.

Saggar, S., Hedley, M.J., White, R.E., Perott, K.W. & Cornforth, I.S.

1992. Development and evaluation of an improved soil test for

phosphorus, 2. Comparison of Olsen and mixed cation–anion

exchange resin test for predicting the yield of ryegrass grown in

pots. Fertilizer Research, 33, 135–144.

Saggar, S., Yeates, G.W. & Shepherd, T.G. 2001. Cultivation effects

on soil biological properties, microfauna and organic matter

dynamics in Eutric Gleysol and Gleyic Luvisol soils in New

Zealand. Soil and Tillage Research, 58, 55–68.

Sakadevan, K., Mackay, A.D. & Hedley, M.J. 1993. Sulphur cycling in

New Zealand hill country pastures. II. The fate of fertilizer sulphur.

Journal of Soil Science, 44, 615–624.

Schulten, H.-R. & Schnitzer, M. 1997. Chemical model structures for

soil organic matter and soils. Soil Science, 162, 115–130.



Shepherd, T.G., Saggar, S., Newman, R.H., Ross, C.W. & Dando, J.L.

2001. Tillage induced changes to soil structure and organic carbon

fractions in New Zealand soils. Australian Journal of Soil Research,

39, 465–489.

Sparling, G.P., Hart, P.B.S., Feltham, C.W., August, J.A. &

Searle, P.L. 1991. Simple methods to produce dual labelled (14C

and 15N) ryegrass and to estimate 14C in soils, plants and

microbial biomass. Land Resources Technical Record No 77,

Department of Scientific and Industrial Research, Lower Hutt,

New Zealand.

Steinbergs, A., Lismaa, O., Freney, J.R. & Barrow, N.J. 1962. Deter-

mination of total sulphur in soil and plant material. Analytica

Chimica Acta, 27, 158–164.

Tabatabai, M.A. & Bremner, J.M. 1970. An alkaline oxidation

method for determination of total sulfur in soils. Soil Science

Society of America Proceedings, 34, 62–65.

Tiessen, H., Stewart, J.W.B. & Moir, J.O. 1983. Changes in organic

and inorganic phosphorus composition of two grassland soils and

their particle size fractions during 60–90 years of cultivation.

Journal of Soil Science, 34, 815–823.



chemical fractionation to characterize changes in sulphur and carbon in soil caused by management

Documents