Soil Carbon Fractions Based on their Degree of Oxidation, and the ...

Aust. J. Agric. Res., 1995, 46, 1459-66
<strong>Soil</strong> <strong>Carbon</strong> <strong>Fractions</strong> <strong>Based</strong> on their Degree of Oxidation,
and the Development of a <strong>Carbon</strong> Management Index
for Agricultural Systems
Graeme J. Blair, Rod D. B. Lefroy and Leanne Lisle
Department of Agronomy and <strong>Soil</strong> Science, University of New England, Armidale, NSW 2351.
Abstract
Increasing population pressure is increasing the demand on agricultural systems in many parts
of the world and this has often led to the degradation of the soil resource. <strong>Soil</strong> carbon (C) is
a major determinant of sustainability of agricultural systems and changes can occur in both
total and active, or labile, C pools.
A procedure is presented to determine the degree of lability of soil C. By treating a ground
sample of soil with 333 mM potassium permanganate (KMn04) to oxidize a proportion of
the carbon and by determining the total carbon by combustion, two fractions of C can be
measured. These fractions represent carbon of different lability, with fraction I representing
the Labile C (CL), which is oxidized by 333 mM KMn04, and fraction I1 representing the
non-labile C (CNL), which is not oxidized by 333 mM KMn04. On the basis of changes in
total carbon (CT), a <strong>Carbon</strong> Pool Index (CPI) is calculated and, on the basis of changes
in the proportion of labile C in the soil between a reference site and those subjected to
agricultural practice or research treatments, a Lability Index (LI) is determined. These two
indices are used to calculate a <strong>Carbon</strong> Management Index (CMI), with CMI = C Pool Index
(CPI) xLability Index (LI) x 100.
Analyses of paired samples (cropped and uncropped) from three sites in northern and
central New South Wales, Australia, have shown a decline in CPI, a greater decline in LI and
hence a decline in the CMI with cropping. Introduction of a legume into a wheat cropping
system restored the CMI from 22 to 37 at the Warialda site. Analyses of paired samples from
a sugarcane area in north Queensland have shown a decline in CMI in systems dominated by
trash burning, but an increase in CMI in systems dominated by green cane trash management.
Similar data from Brazil showed no increase in CT with mulching but a 48% increase in CMI
due to an increase in the lability of C in the soil. The fractionation procedure and CMI
outlined can be used to determine the state and rate of change in soil C of agricultural and
natural systems.
Keywords: soil organic matter, labile carbon, KMn04 oxidation, sustainability.
Introduction
Throughout the world, agricultural activity has generally involved exploitation
of soil organic matter reserves (SOM) as a source of nutrients (Salter and Green
1933). <strong>Soil</strong> organic matter content (pool size) is a balance between addition
and decomposition rates (turnover rates) and, as such, changes in agricultural
practices can result in marked changes in both the pool size and turnover rate
of SOM, carbon and therefore, nutrients.

Graeme J. Blair et al.
The key to sustained productivity of agricultural systems is the maintenance
of SOM levels and nutrient cycling. Both are closely related through the
microbially driven mobilization/immobilization processes (Duxbury et al. 1989).
Some researchers argue that measurement of the size of the soil microbial biomass
(SMB) is the key to understanding the turnover rate of the SOM (Ocio et al.
1991), whilst others consider SMB is a poor indicator of these changes because
factors such as the particular species of organisms comprizing the SMB and the
soil moisture can markedly affect the size of the SMB (Mazzarino et al. 1987).
Sanchez and Logan (1992) calculated from data of Greenland and Nye (1959)
that the generally higher SOM decomposition rates found in the humid tropics
are balanced by higher litter inputs, such that mean SOM contents vary little
between tropical and temperate soils (Sanchez et al. 1982). However, Buol et al.
(1990) found that at the same annual mean temperature, soils from tropical
areas, with isothermic regimes, had higher organic carbon (OC) contents than
those from temperate areas with non-isothermic regimes. The difference between
these two groups diminished as mean temperature increased.
Small changes in total SOM or C are difficult to detect because of the generally
high background levels and natural soil variability. For this reason many attempts
have been made to use subpools of SOM or C as more sensitive indicators
of changes in pool size. Jenkinson and Rayner (1977) identified five pools in
their organic matter cycling model ranging from a decomposable pool, with a
radiocarbon age of less than 1 year, through a biomass pool at 25.9 years to a
chemically stabilized pool with a radiocarbon age of 2565 years. Variations on
these pools have been used in other organic matter cycling models (Hunt 1977;
Paul and van Veen 1978; Smith 1979; Coughenour et al. 1980; McCaskill 1987;
Parton et al. 1989).
In addition to these major pools, SOM fractionation has been carried out on
the basis of extraction of humic substances (Schnitzer 1982), dissolved organic C
(Cook and Allan 1992), particle size (Christensen 1986), natural abundance
(Balesdent et al. 1990; Lefroy et al. 1993), microbial biomass C (Sparling 1992)
and ease of oxidation of C (Loginow et al. 1987). Although this latter method
does not provide quantitative data on C fractions, it does provide qualitative
characterization of soil C.
Changes in the lability of soil carbon have been proposed by Lefroy et al.
(1993) as a measure of sustainability. This procedure relies on the ease of
oxidation of the soil organic C by potassium permanganate. This paper reports
on the standardization of the KMn04 oxidation procedure of Loginow et al.
(1987) and on the development of a carbon management index based on changes
in the total C in the soil and its lability as determined by KMn04 oxidation.
Materials and Methods
The carbon fractionation method is based on that described by Loginow et al. (1987).
Potassium permanganate is a powerful oxidizing agent under neutral conditions, but it is
relatively unstable and thus cannot be used as a primary standard. This creates some problems
in obtaining repeatable and quantitative results. If the precautions outlined in this paper are
taken, however, this technique can be used as a quantitative measure of soil carbon fractions.
The original method of Loginow et al. (1987) relied on using three different concentrations
of the oxidizing agent to oxidize increasing proportions of the soil C within a fixed time
interval. Preliminary studies by Lefroy et al. (1993) found that the use of a single strength of

<strong>Soil</strong> <strong>Carbon</strong> <strong>Fractions</strong> and a <strong>Carbon</strong> Management Index
KMn04 provided sufficient characterization of the labile C to define the state of soil systems.
The total soil carbon, measured by combustion, and the amount of oxidizing agent consumed
by the KMn04 are used to calculate two fractions of organic carbon; one which is oxidized
by KMn04 and a second fraction which is not oxidized by the KMn04.
<strong>Soil</strong> samples were sieved to < 2 mm, subsamples were taken and ground to < 500 pm
and the total C measured in an automatic carbon and nitrogen analyser mass spectrometer
system (ANCA-MS), consisting of a Dumas-type dynamic flash catalytic combustion sample
preparation system (Carlo Erba NA1500), with the evolved gases separated and analysed
by mass spectrometry (Europa Scientific Tracermass Stable Isotope Analyser). Where
significant amounts of carbon are present as carbonates, they are removed by adding sufficient
orthophosphoric acid (2.5% v/v, 85% orthophosphoric acid) to the sample prior to combustion
so as to evolve all the carbonates with gentle heating (approximately 60°C). Repeated
measurement of reasonably uniform samples of soil gives total carbon measurements with CVs
of less that 5%.
Samples of soil containing 15 mg C were weighed into 30 mL plastic screw top centrifuge
tubes and 25 mL of 333 mM KMn04 were added to each vial. Blank samples, containing no
soil, and samples of a standard soil were analysed in each run. The centrifuge tubes were
tightly sealed and tumbled for 1 h, at 12 rpm, on a tumbler with a radius of 15 cm. The
tubes were centrifuged for 5 min at 2000 rpm (RCF=815 g) and the supernatants diluted
1:250 with deionized water. The absorbances of the diluted samples and standards were read
on a split beam spectrophotometer at 565 nm. The range for the standards was chosen to
adequately cover the sample range, normally 300 to 333 mM.
The change in the concentration of KMn04 is used to estimate the amount of carbon
oxidized, assuming that I mM MnO4 is consumed (MnVII + MnII) in the oxidation of
0.75 mM, or 9 mg, of carbon. The results are expressed as mg C g-l soil. The two fractions
are Labile C (CL) = the C oxidized by 333 mM KMn04 and non-labile C (CNL) = the C not
oxidized by 333 mM KMn04.
The presence of MnOz and light causes decomposition of the KMn04. To avoid the change
of KMn04 concentration of the extraction solutions and the standards, solutions have to be
carefully prepared and stored, and the concentrations of all solutions need to be regularly
measured by titration against a primary standard, such as arsenous oxide. When the stock
solutions are prepared, they must be filtered through a non-carbon filter, such as glass wool,
to remove any trace of MnOa, and stored in the dark in glassware which has been thoroughly
cleaned to remove any oxidizable material which would result in MnOa. All glassware must
be thoroughly clean. If appropriate procedures are followed, the standard solutions can be
used for more than 1 month.
The amount of carbon in the sample and the time the soil was in contact with the extracting
solution were found to have significant effects on the amount of oxidation. Standardizing
on 15 mg C and establishing a strict protocol for the various steps in the extraction were
found to significantly reduce errors. The volume (or weight) of the extracting solution and
the dilution of the supernatant after extraction must be precise, as small errors in these steps
are compounded. All steps in the extraction procedure were carried out at 25' C. If the
tumbling facilities outlined here are not available, any alternative can be used, as long as the
conditions are standardized. This is possible because in calculating the <strong>Carbon</strong> Management
Index the data are compared with a reference soil measured at the same time.
To demonstrate the technique and the value of CMI, paired soil samples were collected
from cropped and uncropped sites in three areas of northern and central New South Wales,
Australia. The uncropped site adjacent to the cropped one is used as a reference soil on the
basis that it has been subjected to minimal disturbance and is likely to have more stable C
pools than a disturbed site. In one area, Warialda, a sample was also collected from a site
which had been cropped to wheat, but had subsequently been returned to lucerne pasture.
A second set of paired samples was collected from sugarcane growing areas near Mackay,
Queensland. A further set, from the sugarcane experiment in Brazil of Ball-Coelho et al.
(1993), was kindly made available by the authors for analysis. These samples were taken from
experimental plots at the time of return of sugarcane mulch and 12 months after the mulch
return. Where samples are derived from experiments, the control plot or a sample taken at
the start of the experiment is used as the reference.

Graeme J. Blair et al.
Results and Discussion
The results of the C fractionation of three cropped and uncropped soils from
northern and central New South Wales, Australia, have been presented by Lefroy
et al. (1993) using three strengths of KMn04 and thus four carbon fractions. The
results showed that reduction in all three fractions was of a similar magnitude
and that the decline in each fraction was greater than in the unoxidized fraction.
Consequently, the two lower concentrations of KMn04, 33 and 167 mM, are no
longer used.
In all three soils, the reduction in CL due to cropping was proportionally
greater than the decline in CNL or CT (Table 1). Averaged over the three soils
the values were -63.3%, -39.3% and -44.9% for CL, CNL and CT respectively.
Incorporation of 2 years lucerne into the cropping rotation at Warialda restored
soil C, with the increases being +58.8, +15.7 and +21.6% for CL, CNL and CT
respectively.
Table 1. Labile and non-labile C and <strong>Carbon</strong> Management Indices for some cropped and
uncropped soils in New South Wales, Australia
Cropping Labile Non-labile Total <strong>Carbon</strong> Lability Lability <strong>Carbon</strong>
or grazing carbon carbon carbon pool of C index management
history index index
(CL) (CNL)~ (CT) (CPI) (L) (LI) (CMI)*
(mg g- )
Nyngan (Solonized Brown Earth, Palexeralfl
Uncropped 4.50~ 13.53 18.03 - 0.332 - -
4 yrs cropping 2.21 10.38 12.59 0-70 0.213 0.64 45
Gunnedah (Black Earth, Pellustert)
Stock route grazing 3.62 16.83 20.45 - 0.256 - -
7 yrs cropping 1.55 9-18 10.73 0.51 0.169 0.66 33
Warialda (Red Earth, Paleustalf)
Lightly grazed 3.99 12.78 16.77 - 0.312 - -
18 yrs cropping 1.02 6.49 7.51 0.45 0.157 0.50 23
16 yrs cropping 1.62 7.50 9.13 0.54 0.215 0-69 38
and 2 yrs lucerne
CT sample
A CMI = CPIxLIx loowhere CPI = , LI
C
=
and L = L.
CT reference L reference CNL
CL values are the means of duplicate samples. The mean CV of these seven values = 2.93%.
These data demonstrate that CL declines faster and is restored faster than
CNL or CT, and hence is a more sensitive indicator of the C dynamics of the
system.
Data from the two soils from Mackay, Queensland (Table 2), provide a contrast.
The soil from Marian had been cropped for 90 'years and has a markedly lower
CL, CNL and CT than the reference (uncropped) soil. By contrast, the Victoria
Plains soil which has been cropped for 15 years has a higher CL, CNL and
CT concentration than the adjacent non-cropped reference soil. Green trash
management has been practiced in this area for some time, replacing the practice
of trash burning. As such, although both sites have had similar periods of green

<strong>Soil</strong> <strong>Carbon</strong> <strong>Fractions</strong> and a <strong>Carbon</strong> Management Index 1463
trash management, the proportion of green trash to burnt trash management is
much lower for the Marian site.
Table 2.
Labile and non-labile C and <strong>Carbon</strong> Management Index *om sugar cane cropped
and adjacent non-cropped areas of Mackay, Queensland
Cropping Labile Non-labile Total <strong>Carbon</strong> Lability Lability <strong>Carbon</strong>
history (years) carbon carbon carbon pool of C index management
index
index
(CL) (CNL)~ (CT) (CPI) (L) PI) (CW
(mg g-
Marian (Yellow Podzolic, Haplustalf)
0 4.08~ 10.91 14.99 - 0.374 - -
90 1.54 7.04 8.55 0.57 0.219 0.59 34
Victoria Plains (Black Earth, Pelloxerept)
A CL values are the means of duplicate samples. The mean CV of these four values = 3.47%.
In the data from Brazil (Table 3), similar trends in the C fractions are
evident, with an increase of +39.7%, +2.4% and +8.5% in CL, CNL and CT
respectively, over the 12 month period after mulch return. Ball-Coelho et al.
(1993) reported no significant increase in CT in the mulched treatment reported
here. As for the other sites, the CL data indicate a marked change in the
amount of labile or active C in the soil. The CL values in Table 3 are the means
of duplicate analyses of six soil samples taken from each of three replicates at
each time. This set of 18 samples, with 36 analyses, provides an opportunity to
assess the spatial and analytical variability of CL. Analysis of variance of the
36 values shows no significant difference between duplicate analyses, a significant
(P = 0.04) subsampling effect and a highly significant, replicate, or block effect
(P < 0 e00). There was a highly significant block effect with CT values.
Table 3. <strong>Carbon</strong> data and <strong>Carbon</strong> Management Index for soil samples from a mulch return
treatment of a sugarcane experiment conducted on a Red Yellow Lalosolic Podzol (Oxic
Haplustult) soil by Ball-Coelho et al. (1993) in Brazil
Time after Labile Non-labile Total <strong>Carbon</strong> Lability Lability <strong>Carbon</strong>
mulch return carbon carbon carbon pool of C index management
(months) index index
(CL) (CNL)~ (CT) (CPI) (L) (LI) (CMI)
(mg g-
Derivation of the <strong>Carbon</strong> Management Index
Since the continuity of C supply depends on both the total pool size and
the lability (an estimate of turnover rate), both must be taken into account in
deriving a carbon management index. This can be achieved as follows.
(a) Change in total C pool size

1464 Graenle J. Blair et al.
The loss of C from a soil with a large carbon pool is of less consequence than
the loss of the same amount of C from a soil already depleted of C or which
started with a smaller total C pool. Similarly, the more a soil has been depleted
of carbon the more difficult it is to rehabilitate. To account for this a C Pool
size Index is calculated as:
sample total C (mg g-l)
CT sample
C Pool Index (CPI) = -
reference total C (mg g-l) CT reference'
(b) The loss of labile C is of greater consequence than the loss of nonlabile
C. To account for this a carbon Lability Index is calculated as:
Lability of C (L) =
C in fraction oxidized by KMn04
C remaining unoxidized by KMn04
- &.
CNL
'
Lability of C in sample soil
Lability Index (LI) = Lability of C in reference soil '
(c) The <strong>Carbon</strong> Management Index (CMI) can then be calculated as:
CMI = C Pool Index x Lability Index = CPI x LI x 100
Measurements Required to Establish the <strong>Carbon</strong> Management Index
Calculations of the CMI require samples of the soil of interest and a sample
collected from a reference area. The choice of the reference area depends on
the circumstances. In experiments the control treatment represents a useful
reference point. In studying the impact of agricultural practices, an undisturbed
or relatively undisturbed site on the same soil type and near to the study unit
should be selected. Such a site represents an area where change in soil C
dynamics is likely to be slow relative to the disturbed area, and this can serve
as a reference sample.
The following measurements are required on both soils to be able to calculate
the CMI:
(a) total C (CT) by direct measurement not by Walkley-Black (1934) which
underestimates total C in those soils with a high proportion of recalcitrant C.
(b) determination of the proportion of C oxidized by 333 mM Kh!ho4 (CL) by
the procedure outlined above,
(c) determination of the proportion of C not oxidized by 333 mM KMn04 (CNL)
by difference.
The above calculations have been made on the soils reported above (Tables 1, 2
and 3).
At each of the sites examined in Table 1, cropping has resulted in a marked
decline in both labile and non-labile C and hence the CMI has declined. By
taking both pools of C into account, as is done in calculating the CMI, a more
definitive picture of the soil C is obtained than when only total C or labile C is
measured.

<strong>Soil</strong> <strong>Carbon</strong> <strong>Fractions</strong> and a <strong>Carbon</strong> Management Index
In the sugarcane soil from Marian, which had been cropped for 90 years,
the CMI was 34, whereas at Victoria Plains the CMI was 110 as a result of a
shorter cropping period and a higher proportion of conservation oriented residue
management.
Where cane tops and trash mulch had been returned at the Brazilian site
(Table 3) there was only a small increase in CPI but a large increase in LI,
resulting in a 48% increase in CMI.
There is no 'ideal' value of CMI. The index provides a sensitive measure of the
rate of change in soil C dynamics of systems relative to a more stable reference
soil. When monitored over time or when a new practice is introduced, as in the
case of the data from Warialda in Table 1, the CMI indicates if the system is
in decline or being rehabilitated. In experimental situations, CMI can be used
to monitor differences in soil C dynamics between treatments and over time.
In the studies reported here, total C has been measured by ANCA-MS
combustion analysis. The changes in the lability of organic C that result
from cropping indicate that the constant multiplier factor of 1.3 used in the
Walkley-Black (1934) method will result in an incorrect estimate of total C in
many soils. In soils with a small proportion of labile C, such as the Warialda
cropped soil (Table I), it will underestimate total C, whereas in the Warialda
uncropped soil it will overestimate total C.
The labile C component of total C determined by KMn04 represents a larger
pool than that commonly measured as soil microbial biomass. This relatively
large pool of easily oxidizable C has been measured in soil samples deep in the
profile, which indicates that it is mobile. As such it could represent the C source
required to drive denitrification and methanogenesis processes.
Acknowledgements
The project was supported by ACIAR PN9102, the Grains Research and
Development Corporation, the Cotton Research and Development Corporation
and the Cooperative Research Centre (CRC) for Sustainable Cotton Production.
Mr L. Chapman, Bureau of Sugar Experiment Stations, Mackay, and Dr B.
Ball-Coelho, Agriculture Canada, kindly provided soil sampIes.
The technical input of Vanessa Hunter and Judith Kenny is greatly appreciated.
References
Balesdent, J., Mariotti, A., and Boisgontier, D. (1990). Effect of tillage on soil organic carbon
mineralization estimated from 13c abundance in maize fields. Journal of <strong>Soil</strong> Science 41,
587-96.
Ball-Coelho, B., Tiessen, H., Stewart, J. W. B., Salcedo, I. H., and Sampio, E. S. B. (1993).
Residue management effects on sugarcane yield and soil properties in northeastern Brazil.
Agronomy Journal 85, 1004-8.
Buol, S. W., Sanchez, P. A., Kimble, J. M., and Weed, S. B. (1990). Predicted impact of
climate warming on soil properties and use. In 'Impact of <strong>Carbon</strong> Dioxide Trace Gases,
and Climate Change in Global Agriculture7. ASA Special Publication No. 53. (Eds B. A.
Kimball et al.) pp. 71-82. (ASA, CSSA, SSSA: Madison, Wisconsin, USA.)
Christensen, B. T. (1986). Straw incorporation and soil organic matter in macro-aggregates
and particle size separates. Journal of <strong>Soil</strong> Science 37, 125-35.
Cook, B. D., and Allan, D. L. (1992). Dissolved organic carbon in old field soils: compositional
changes during the biodegradation of soil organic matter. <strong>Soil</strong> Biology and Biochemistry
24, 595-600.

Graenle J. Blair et al.
Coughenour, M. B., Parton, W. J., Lauenroth, W. K., Dodd, J. L., and Woodmansee, R. G.
(1980). Simulation of a grassland sulfur-cycle. Ecological Modelling 9, 179-213.
Duxbury, J. M., Smith, M. S., and Doran, J. W. (1989). <strong>Soil</strong> organic matter as a source
and sink of plant nutrients. In 'Dynamics of <strong>Soil</strong> Organic Matter in Tropical Ecosystems'.
(Eds D. C. Coleman, .J. M. Oades and G. Uehara.) pp. 33-67. (University of Hawaii Press:
Honolulu, Hawaii.)
Greenland, D. J., and Nye, P. H. (1959). Increases in carbon and nitrogen contents of tropical
soils under natural fallows. Journal of <strong>Soil</strong> Science 9, 284-99.
Hunt, H. W. (1977). A model simulation for decomposition in grasslands. Ecology 58, 469-84.
Jenkinson, D. S., and Rayner, .J. H. (1977). The turnover of soil organic matter in some of
the Rothamsted classical experiments. <strong>Soil</strong> Science 123, 298-305.
Lefroy, R. D. B., Blair, G. J., and Strong, W. M. (1993). Changes in soil organic matter with
cropping as measured by organic carbon fractions and natural isotope abundance.
Plant and <strong>Soil</strong> 1551156, 399-402.
Loginow, W., Wisniewski, W., Gonet, S. S., and Ciescinska, B. (1987). Fractionation of organic
carbon based on susceptibility to oxidation. Polish Journal of <strong>Soil</strong> Science 20, 47-52.
Mazzarino, M. J., Oliva, L., Abil, A., and Acosta, M. (1987). Factors affecting nitrogen
dynamics in a semiarid woodland (Dry Chaco, Argentina). Plant and <strong>Soil</strong> 138, 85-98.
McCaskill, M. R. (1987). Modelling S, P and N cycling. Ph.D. Thesis, The University of New
England, Armidale, Australia.
Ocio, J. A., Brooks, P. C., and Jenkinson, D. C. (1991). Field incorporation of straw and its
effect on microbial biomass and soil organic N. <strong>Soil</strong> Biology and Biochemistry 23, 171-6.
Parton, W. J., Sanford, R. L., Sanchez, P. A., and Stewart, K. W. B. (1989). Modeling soil
organic matter dynamics in tropical soils. In 'Dynamics of <strong>Soil</strong> Organic Matter in Tropical
Ecosystems'. (Eds D. C. Coleman, J. C. Oades and G. Uehara.) pp. 153-71. (University
of Hawaii Press: Honolulu.)
Paul, E. A., and van Veen, J. A. (1978). The use of tracers to determine the dynamic nature of
organic matter. Transactions of the 11th International Congress of <strong>Soil</strong> Science, Edmonton,
Canada, .June 1978. pp. 61-102.
Salter, R. M., and Green, T. C. (1933). Factors affecting the accumulation and loss of nitrogen
and organic carbon in cropped soils. Journal of American Society of Agronomy 25, 622-30.
Sanchez, P. A., and Logan, T. J. (1992). Myths and science about the chemistry and fertility of
soils in the tropics. In 'Myths and Science of <strong>Soil</strong>s in the Tropics'. SSSA Special Publication.
(Eds R. La1 and P. A. Sanchez.) pp. 35-46. (SSSA and ASA: Madison, Wisconsin, USA.)
Sanchez, P. A., Gichuru, M. P., and Katz, L. B. (1982). Organic matter in major soils of the
tropical and temperate region. 12th International Congress of <strong>Soil</strong> Science 1. New Dehli,
1982. pp. 99-114.
Schnitzer, M. (1982). Organic matter characterization. In 'Methods of <strong>Soil</strong> Analysis Part 2,
Chemical and Microbiological Properties'. Agronomy No. 9. 2nd Edn. (Eds A. L. Page, R.
H. Miller and D. R. Keeney.) pp. 581-94. (ASA, SSSA.: Madison, Wisconsin, USA.)
Smith, 0 . L. (1979). An analytical model of the decomposition of soil organic matter. <strong>Soil</strong>
Biology and Biochemistry 11, 585-606.
Sparling, G. P. (1992). Ratio of microbial biomass carbon to soil organic carbon as a sensitive
indicator of changes in soil organic matter. Australzan Journal of <strong>Soil</strong> Research 30, 195-207.
Walkley, A., and Black, I. A. (1934). An examination of the Degtareff method for determining
soil organic matter and a proposed modification of the chromic acid titration method. <strong>Soil</strong>
Sczence 37, 29-38.
Manuscript received 15 March 1995, accepted 13 June 1995