African Journal of Agricultural Research Vol. 5(22), pp.

3088-3095, 18 November 2010

Available online at http://www.academicjournals.org/AJAR
ISSN 1991-637X ©2010 Academic Journals

Full Length Research Paper

Loss on ignition: Measuring soil organic carbon in soils

of the Sahel, West Africa
1 2* 3
H. Konare , R. S. Yost , M. Doumbia1, G. W. McCarty , A. Jarju4 and R. Kablan2
1
Labo SEP, Institut d’Economie Rurale, BP 262 Bamako-Mali.
2
University of Hawai`i, 3190 Maile Way, St. John 102, Honolulu, Hawaii 96822, USA.
3
USDA-ARS, Hydrology and Remote Sensing Laboratory, Beltsville, Maryland, USA.
4
National Agricultural Research Institute, Gambia.
Accepted 27 September, 2010

Traditional methods of measuring soil organic carbon (SOC) are not adequate for accurate evaluation of
C sequestration nor to meet environmental safety requirements. Methods that permit accurate
assessment of SOC, which are environmentally healthy, and yet feasible for use in developing countries
are needed. Initial and ignition temperatures were adjusted so that loss on ignition (LOI) measures of
SOC matched measurements by combustion methods. A comparison of combustion, LOI, and Walkley-
-1
Black methods was carried out with soils of Mali, West Africa, ranging from 2 - 25 g kg SOC. The LOI
° °
method, when using initial (preparation) temperatures of 105 C and ignition temperatures of 350 C,
2
related closely with combustion measures (Adj. R of 0.89). The SOC measurements by combustion,
however, differed from those of Walkley-Black procedure. The LOI calibration equation was %C
2
(combustion) = 0.03 + 0.36 x %weight loss by LOI, adj. R =0.89. The Mali calibration of LOI with
combustion SOC was tested on soils of Gambia, however, a systematic over-prediction suggested that
re-calibration may be needed to ensure LOI method accuracy with soils different from the calibration
set. This need for re-calibration is similar to the Type 2 error, where the sample belongs to a different
calibration set.

Key words: Loss on ignition, Sahel, Mali, Africa, carbon sequestration, combustion, Walkley and Black.

INTRODUCTION

Soil organic matter plays a critical role in the conservation sustainability improves with increases in organic matter in
of fertility, especially in the extremely coarse textured soils of the Sahel so that farmers can continue to grow
soils of West Africa (Pieri, 1985). In these soils SOC is food and at the same time to maintain a robust and
both is a source of nutrients and mechanism for nutrient healthy environment (Piéri, 1985). Soil organic carbon,
retention, it affects water infiltration as well as increases derived from the decay of plants and animals in various
retention of plant available water, and it provides forms, is a key component of organic matter playing a
favorable conditions for soil biota. Several studies have great role in increasing soil fertility. Today, a better
characterized soils of the Sahel in West Africa as poorly management of soil organic carbon is needed to enhance
buffered, acidic (pH 4.9), sandy and low in soil organic soil fertility. As has been emphasized elsewhere, farmers
-1
carbon (2 - 5 gkg ) with a dominance of kaolinite and and producers of tomorrow may need to not only farm
sesquioxides of Fe and Al (Piéri, 1985, Bationo and soil judiciously but to also “farm carbon” (Lal, 2008).
Mokwunye, 1991; Kablan et al., 2009). Agricultural Accurate, rapid and cost-effective soil carbon
determination is also important to the development of a
soil carbon accounting system (Antle and Uehara, 2002).
Unfortunately, soil organic C measurement in developing
*Corresponding author. E-mail: rsyost@hawaii.edu. Tel: 808- countries is difficult because accurate methods are either
956-7066. not environmentally acceptable or they are too expensive,
 Konare et al. 3089

requiring costly instruments and highly trained carbon (SOC): LOI, dry combustion, and the Heanes
technicians. Traditionally soil organic C determination (1984) adaptation of the Walkley-Black (1934)
relies on two methods and each method has its own dichromate method.
limitations: (1) Oxidation of organic matter by K2Cr2O7
(Walkley and Black, 1934), and (2) Dry combustion at
1350°C (Nelson and Sommers, 1996). MATERIALS AND METHODS
The Walkley-Black method only estimates a fraction the Soils
soil C (Allison 1935). Moreover, the method is time-
consuming, it subjects the analyst to handling highly Mali: Twenty-five soils from Mali were selected and
dangerous concentrated sulfuric acid, and it produces sampled to provide clay contents ranging from 30 - 600 g
hazardous waste including the chromate ion (Swift, -1 -1
kg and organic C ranging from 2 to 25 g kg C (Table
1996). In addition, the incomplete oxidation of organic 1). Soils were collected in regions with representative
carbon often results in erroneous values (Heanes, 1984). agricultural systems of Mali, including potential C-
Dry combustion, on the other hand, is usually regarded sequestration sites: Cinzanna (millet); Konobougou
as the most accurate method (McCarty et al., 2002); (cotton, maize, millet); Oumarbougou (cotton, millet,
however, the method is not widely available in the Sub- maize); Niono (irrigated rice); Niessoumana (cotton,
Saharan tropics because of the high cost of combustion irrigated rice) Longorola (cotton, maize, cereal, irrigated
analyzers, developed infrastructure for maintenance and rice); Fansirakoro (millet, maize). Prior to analysis soil
reagent requirements and the need for highly skilled samples were dried, finely ground to pass a sieve of 60
operators. mesh (0.251 mm).
Loss on ignition (LOI) methods has long been used by The Gambia: A set of 144 samples from Gambia were
soil testing laboratories to measure soil organic matter included in this comparison as a test of the LOI
and soil organic carbon. Mitchell (1932) reported that calibration on soils from Mali. The Gambian soils were
temperatures of between 350 and 400°C were suitable. obtained from 72 locations in an experiment to test for
Davies (1974) illustrated that even a temperature of effects of Amenagement en Courbes de Niveau (ACN) on
430°C could be used in the presence of calcium C sequestration (Doumbia et al., 2009). The soils of
carbonate without errors. However, higher temperatures Gambia were prepared similar to those of Mali. Rainfall at
can drive off structural water of clays causing the Gambian site was substantially higher than that at the
decomposition of carbonates and hydrated salts Mali sites – a mean of approximately 1200 mm compared
increasing error associated with higher temperature use. to a range between 450 - 1000 mm.
Schulte and Hopkins (1996) developed an LOI method Objective 1. Determine the combination of selected
that has been successful in the North Central Region of initial (preparation) and ignition temperatures of LOI
the USA. Konen et al. (2002), using the method of methods that gives results most similar to the combustion
Schulte and Hopkins (1996), which included pretreatment method of measuring soil organic C.
drying at 105°C and 2 h ignition at 360°C, reported The LOI procedure described by Nelson and Sommers
regression relationships with SOC determined by (1996) was used as the basic method. Five gram of
2
combustion ranging from R of 0.94 - 0.98, but slopes of sample was placed in 30 ml crucibles that have been pre-
the regressions differed by nearly a factor of 2 between weighed and pre-heated at 105°C, then cooled. Crucibles
soil groups of different states (0.57 to 1.14). Konen et al., with soil samples were then handled as follows:
(2002) concludes that LOI methods are inexpensive,
rapid, and require a minimum of expensive equipment, (1) Heat in a drying oven at the initial (preparation)
but that unique relationships exist for soils from differing temperature over night,
geographic locations, possibly due to varying amounts of (2) Cool to room temperature in a desiccator over silica
soil clay, silt, and differing mineralogy. gel,
This study sought to evaluate a loss on ignition (LOI) (3) Weigh the crucibles + sample to the nearest mg.
method as an alternative to combustion methods and Reweigh at 105°C.
dichromate methods in soils of the Sahel of West Africa, (4) Place into a muffle furnace and heat at the ignition
but with the objective of duplicating combustion results. temperature for 16 h.
The published studies discussed above reveal that a (5) Remove from the furnace and cool to room
variety of initial (preparation) and ignition temperatures temperature in a desiccator.
have been used. The objectives of this study were: (6) Weigh the ignited samples to the nearest mg and
record weight.
(1) To compare various initial (preparation) and ignition
The mass of soil loss expressed on a dry weight basis by
temperatures of the LOI method in order to duplicate
the following equation:
combustion method results, which are generally
considered the most accurate. -1
LOI (g kg ) = (Weightinitial T – Weightignition T)/ Weightinitial T x
(2) To compare three methods of measuring soil organic 1000.
3090 Afr. J. Agric. Res.

Table 1. Selected soils and representative agricultural systems of Mali and their range in soil clay.

-1
Sample Sampling locations Depth (cm) Agricultural systems Clay, g kg
1 Cinzana Sol 1 0-20 Millet 52
2 -«- 20-60 -«- 124
3 Diou 0-20 110
4 -«- 20-60 -«- 40
5 Niessouma bas-fonds 0-20 Cotton, irrigated rice 454
6 -«- 20-60 -«- 520
7 Cinzana Sol 3 0-20 Millet 56
8 -«- 20-60 -«- 182
9 Cinzana Sol 5 0-20 Millet 476
10 -«- 20-60 -«- 538
11 Konobougou 0-20 Cotton, maize, millet 72
12 -«- 20-60 -«- 140
13 Longorola Glacis 0-20 Cotton, maize, cereal 88
14 -«- 20-60 -«- 342
15 Oumarbougou 0-20 Cotton, maize, millet 308
16 -«- 20-60 -«- 96
17 Cinzana Sol 4 0-20 -«- 70
18 -«- 20-60 Maize, Millet 166
19 Fansirakoro 0-20 -«- 104
20 -«- 20-60 Millet 72
21 Cinzana Sol 2 0-20 -«- 22
22 -«- 20-60 Irrigated rice 420
23 Longorola bas-fonds 0-20 Irrigated rice 102
24 Danga 0-20 Irrigated rice 426
25 Moursi 0-20 Irrigated rice 278

The loss in weight of the soil is taken as a raw measure 2) Modified Walkley and Black or dichromate oxidation
of the organic carbon content. (method: The analyses were performed at the Soil and
A set of initial temperatures, usually designed to Plant Laboratory in the Tropical Plant and Soil Science
eliminate of adsorbed water, was selected as was a set Department of University of Hawai`i using the modified
of ignition temperatures as follows: Two drying dichromate oxidation method (Heanes, 1984). According
temperatures: 105 and 120°C were employed reflecting to the procedure, K2Cr2O7 was added to the soil in an
differences by existing laboratories. Ignition tempera- acidic solution for quantitative oxidation of organic C. In
tures: several temperatures have been used ranging from an attempt to achieve complete oxidation of organic C,
350 to 600°C (Davies, 1974; Heanes 1984, White and the finely ground soil was digested in test tubes by
Roth, 1984). A range of ignition temperatures from 350 - heating externally in a drying oven at 150°C with 0.167 M
600°C in 50°C increments was tested. This combination K2Cr2O7 mixed with concentrated H2SO4. The excess
of initial, preparatory and ignition temperatures formed dichromate was quantified colorimetrically to assess how
the basis of the exploration to calibrate the LOI method much reacted with the soil C. The amount of reacted
with combustion measurements of soil C. The dichromate was roughly proportional to soil organic
comparison of temperatures was carried out using a carbon content.
Thermolyne (Use of a specific brand name does not (3) Dry combustion: Soils were ashed at 1350°C and the
imply a recommendation by the University of Hawaii or released CO2 was measured with a LECO CNS-2000
any institutions of other authors of this manuscript.) analyzer (Nelson and Sommers 1996).
muffle furnace, Model F-11730, 220v.
Objective 2. Comparison of methods of determining soil
organic carbon: The soil organic carbon in the samples Statistical analysis
was measured by three methods: The wet oxidation and the combustion procedures were
compared in a 1:1 plot on the base of the mean error, ME
(1) Loss on ignition: The LOI method, developed for = (Xpred.-Xobs.)/n, (n is the number of observations) and
Objective 1, was used for the comparison of three methods. the root mean squared error (RMSE). If the ME is zero
 Konare et al. 3091

25 25

2
A. Adj. R =0.894 2
B. Adj. R =0.736
20 20

SOC by combustion, g kg-1

-1
SOC by combustion, g kg
15 15

10 10

5 5

0 0

0 20 40 60 0 20 40 60

Weight loss by LOI at 105-350oC, g kg-1 Weight loss by LOI at 120-350oC, g kg-1

Figure 1. Calibration curve between soil organic carbon SOC (g kg-1) by combustion and the weight
loss (g kg-1) by LOI at the two temperatures (A): 105 - 350°C, (B): 120 - 350°C.

Table 2. Comparison among LOI initial and ignition temperatures and dry combustion soil C.

LOI temperature combinations: 1 1 2

Intercept Slope Adjusted R Residual mean square error (RMSE)
Initial – Ignition (°C)
120 - 350 0.1733 0.3344 0.736 5.93
120 - 600 0.2239 0.1276 0.693 6.91
105 - 350 0.1958 0.3602 0.894 2.37
105 - 400 0.4641 0.2659 0.736 5.95
105 - 450 0.5553 0.1668 0.684 7.09
105 - 500 0.3969 0.1362 0.702 6.83
105 - 550 0.3351 0.1337 0.731 6.05
105 - 600 0.3828 0.1318 0.744 5.76
1
Carbon by combustion = intercept + slope x LOI weight loss.

-1
then there is no bias indicating a high accuracy, on the range 30 to 40 g kg , which may be due to two reasons:
average, for the calibration model, but if the ME is differences in organic matter composition among soils or
positive then there is a systematic over-prediction, if it is the presence of volatile inorganic components that lose
negative then there is systematic underprediction weight during prolonged heating at 350°C (Figures 1 and
2
(Loague and Green, 1991). Data were analyzed using 2) (University of Georgia, 2003).The Adjusted R of 0.894
Minitab 14. The accuracy of the prediction was with the 105 to 350°C combination, which was the
2
determined based on the AdjR and the root mean highest, suggests a minimum of errors associated with
square error (RMSE). those temperatures.
The ignition temperatures may explain much of the
variation in the measured organic C content. The weight
RESULTS AND DISCUSSION loss increased when the ignition temperature was higher
(Table 2). Ben Dor and Banin (1989) reported that the
Effect of varying initial, preparatory temperatures on loss of mass from the soil might be subdivided into
LOI estimates of soil C several stages:

Figure 1 shows that some soils depart from the (i) Hygroscopic water loss (50 to 100°C),
regression line, particularly when the initial preparatory (ii) Organic matter loss (100 to 400°C),
temperature was 120°C. In this range weight loss did not (iii) Thermal reactions such as dehydroxylation of
correspond with differences in combustion C. The phyllosilicates (200 to 700°C),
disparity seemed particularly large on soils with LOI in the (iv) Decarboxylation of carbonates (700 to 1000°c).
3092 Afr. J. Agric. Res.

25 25

A. Adj. R2=0.684 B. Adj. R2=0.744

20 20

SOC by combustion, g kg-1

SOC by combustion, g kg-1

15 15

10 10

5 5

0 0
0 20 40 60 80 100 120 140 0 20 40 60 80 100 120 140
o -1
o
Weight loss by LOI at 105-450 C, g kg -1 Weight loss by LOI at 105-600 C, g kg

Figure 2. Calibration curve between soil organic carbon (SOC) (g kg-1) by combustion and the weight loss (g kg-
1
) by loss on ignition (LOI) at the two temperatures (A): 105 - 450°C, (B): 105 - 600°C.

subsequent inaccurate estimate of organic matter content

SOC by Walkley & Black, g kg-1
SOC by Walkley and Black, g kg-1

Mean error = - 1.08 (Heanes, 1984; White and Roth, 1984). The uncertainty
20
is more pronounced with high clay soils that contain
gibbsite with small amounts of soil carbon, particularly in
15 subsoils (Heanes, 1984; White and Roth, 1984). These
authors point out that gibbsite tends to lose much water
at 300°C. Thus, the discrepancy could be due to the fact
10
that hydrated aluminosilicates, carbonate minerals, and
some hydrated salts are decomposed at the high
5 temperatures (White and Roth, 1984; Ben Dor and Banin,
1989). The careful control of heating temperatures and
the use of an analytical balance with mg accuracy is
0
0 5 10 15 20 needed to improve method precision.
-1 -1 Table 2 illustrates the results of the performance of
SOCbybycombustion,
SOC Combustion, g kg
g kg combinations of initial LOI temperatures with increasing
2
combustion temperatures. The Adjusted R for the 105 -
Figure 3. Comparison (1:1 line) between soil organic carbon
(SOC) measured by the combustion method and the Walkley
350°C combination provided the closest approximation of
– Black method (dichromate) in 25 soils of Mali, West Africa. SOC measured by the combustion method. Our results
A mean error of -1.08 indicates that averaged for all samples corroborate those obtained by previous researchers, who
dry combustion analysis values were 1.08 g kg-1 larger than also found a strong correlation between combustion or
Walkley-Black values. wet oxidation and loss-on-ignition (LOI105-350°C) (Ben Dor
and Banin, 1989; Davies, 1974). Given the regression Y
= 0.36X + 0.03, where X is the weight loss resulting from
LOI105-350°C and Y is the soil carbon measured by the
Effect of varying the ignition temperature combustion method, the 95% confidence limits would
cover all data points (data not shown).
When the LOI was carried out at the highest temperature
(600ºC) it resulted in a considerably greater loss of
weight than when lower temperatures were used (Figure
2
2). The Adjusted R of the comparison of LOI at 600°C Comparison between the Walkley and Black and
with the combustion estimate of C was much lower than combustion methods
the one at 350°C (Table 2). The higher uncertainty of LOI
estimates at the ignition temperature 600°C was probably There was a moderate agreement between the
due to dehydroxylation and decomposition of inorganic determination of SOC by the Walkley and Black (W and
constituents resulting in increased weight losses and B) and the dry combustion procedures (Figure 3). The
 Konare et al. 3093

25

Niessouma bas-fonds

20
g kgg-1kg-1
Longorola bas-fonds
Carbon,

15
carbon,
Organic

10
organic
SoilSoil

0
0 200 400 600
-1
Soil clay, g kg
Figure 4. Relationship between C (g Kg-1) by combustion and clay for 25 soils from Mali (SOC =
1.854 + 0.019 Soil Clay, Adj. R2=0.45.

-1
mean error (ME) of –1.08 g kg indicated a systematic organic matter may differ between kaolinite and smectite
underprediction by the W and B method in comparison (Ben-Dor and Banin, 1989; Nelson and Sommers, 1996).
with the combustion method. The Root Mean Square It can be seen in the graph that low organic carbon is
Error (RMSE) of the relationship between the combustion associated with low clay. Specifically, as clay contents
and W and B = 0.054. increased a linear increase occurred in organic C
2
concentrations. Only a weak regression (R = 0.45)
existed for soil clay versus organic C concentrations.
Organic C and soil clay Several previous studies have revealed strong
correlations between soil C content and clay content
The amount of soil clay was determined using the pipette (Nichols, 1984), but the correlation varied among soils
sampling method (Gee and Bauder, 1986). Clay minerals with different iron and aluminum oxide content. Kaolinite
play an important role in storage and long-term stability of and gibbsite are the predominant types of clay existing in
organic matter in soils (Stevenson, 1982). Clay minerals Mali soils (PIRT, 83; Doumbia, 1994).
have a high specific surface area and carry a charge,
binding and thereby chemically stabilizing organic matter.
2
The results from Figure 4 with an R of 0.45 illustrate a Soil carbon prediction of unknown samples
poor relationship between soil clay and organic C content
-1 -1
up to 400 - 500 g kg clay. Above 300 to 400 g kg of In the study the best calibration model was chosen on the
2
clay the relationship between soil clay and the C content basis of highest regression prediction (R = 0.89) and the
was even less apparent. It may be noteworthy that two lowest values of RMSE (Table 2). The accuracy of C
soils that exhibited particularly high amounts of soil C prediction in unknown samples depends on the accuracy
than most soils were from poorly drained soils (bas of the constructed model, which must also capture the
fonds) that may have led to greater amounts of 2:1 clay variability in soil C of the sites represented in the original
minerals (Figure 4). Factors such as type of clay mineral calibration. However, it is clear that the soils differ in
have been reported to affect LOI weight loss in soils degree of weathering, mineral composition and structural
(Ben-Dor and Banin, 1989). It may be that the water water all of which affect LOI results, and which can result
contained in the interlayer caused an increase of the in significant error. In order to test the transferability of
weight loss (phyllosilicate dehydroxylation) (Ben-Dor and the LOI calibration developed with Mali to soils in other
Banin, 1989). In addition, the amount of clay-bound countries of the Sahel, soil samples from The Gambia
3094 Afr. J. Agric. Res.

Measured SOC by combustion, g kg-1

5

0
0 1 2 3 4 5 6
-1
Predicted SOC using Mali calibration, g kg -1
Predicted SOC using Mali Calibration, g kg
Figure 5. Predicted vs. observed soil organic C, g kg-1 for 144 samples from Gambia, using the
Mali calibration. Calibration for Gambia samples: Combustion SOC = -2.548 + 1.186*LOI value,
Adj. R2 = 0.73.

were analyzed and LOI C predictions were made. population of soils with respect to LOI estimation of soil
As illustrated in Figure 5 with the 1:1 line, all the organic carbon.
sample results for combustion analyses were below the One further concern for both the combustion and LOI
1:1 line indicating a systematic over-prediction by LOI methods in Sahelian soils is that neither method
based on the Mali model. The Mali model was not able to distinguishes between charcoal, coal fragments, and soil
accurately predict combustion carbon in the soils of The organic matter. It is obvious that C derived from the
Gambia. This type of error could result from an under- fragments of coal does not function in the same manner
estimate of loss of structural water in soil from The as organic matter in the soil.
Gambia. The greater presence of minerals with high
water contents, such as amorphous oxides, may result in
the over-prediction. It is interesting to note, however, that Conclusion
the relationship between combustion and LOI predictions
of soil C for the soils of Gambia continued to linear, but The soil carbon in soils of Mali measured by the
-1
shifted downwards, consistent with the possibility of combustion method in g kg can be estimated by an LOI
greater amounts of weight loss in the Gambia samples method using a simple calibrated regression equation Y =
for the same amount of soil organic C as in the Malian 0.36X + 0.03, where X is the weight loss in resulting from
case. ignition of the sample from - 105 - 350°C. The 95%
-1
The cause of the over-prediction of soil C by the LOI confidence limits, ± 2.5 g kg would cover all data points
method in Gambian soil subset is not known, but it is a (data not shown). The LOI method appears to be useable
clear warning that accurate prediction of soil organic C by for the range of soils in this set collected in Mali, but not
the LOI method requires that the unknown soils be for soil samples collected in The Gambia. The method is
among those in the calibration set, else substantial errors easier and more convenient than the expensive dry
can occur. The over-prediction indicates that some basic combustion method, especially for developing countries.
properties of Mali soils differ from those of the Gambian The LOI method, thus calibrated, presents a practical and
subset. These results indicate that re-calibration is relatively rapid alternative for determination of soil
needed to apply the LOI method to regional soils of The organic carbon in soils of the Sahel of Africa, although re-
Gambia. Thus the Gambian soils belong to a differing calibration will be needed if the unknown sample is from
 Konare et al. 3095

a different calibration set. Based on these results it Konen M, Jacobs PM, Burras CL, Talaga BJ, Mason JA (2002).
Equations for Predicting Soil Organic Carbon Using Loss-on-Ignition
appears necessary to check the calibration of the LOI
for North Central U.S. Soils. Soil Sci. Soc. Am. J., 66: 1878-1881.
method unless one can be certain that the unknown soils Lal R (2008). Soils and Sustainable Agriculture – A review. Agronomy
belong to the same population of soils as the calibration Sustain. Dev., 28” 57-64.
set. Thus cautious testing of a new set of soils needs to Loague K, Green RE (1991). Statistical and graphical methods for
evaluating solute transport models: Overview and application. J.
be considered when using this method.
Contaminant Hydrol., 4: 139-161.
Mitchell J (1932). The origin, nature, and importance of soil organic
constituents having base exchange properties. J. Am. Soc. Agron.,
ACKNOWLEDGMENTS 24: 256-275.
McCarty G, Reeves JB, III Reeves VB, Follet RF, Kimble JM (2002).
Mid-Infrared and Near-Infrared Diffuse Reflectance Spectroscopy for
This research was made possible through support Soil Carbon Measurement. Soil Sci. Soc. Am. J., 66:640-646.
provided by the Offices of Agriculture and Natural Nelson DW, Sommers LE (1996). Total Carbon, Organic Carbon, and
Resources Management, Bureau for Economic Growth, Organic Matter. In Methods of Soil Analysis, Part 3. Chemical
Methods,
Agriculture and Trade, U.S. Agency for International
961-1009. Madison WI:Book ser.Nº5, SSSA.
Development, under the terms of Award No. LAG-G-00- Nichols, JD 1984. Relation of organic-carbon to soil properties and
97-00002-00 to the University of Hawai`i at Manoa. The climate in the Southern Great Plains. Soil Sci. Soc. Am. J., 48: 1382-
opinions expressed herein are those of the authors and 1384.
Piéri C (1995). Long-Term Soil Management Experiments in Semiarid
do not necessarily reflect the views of the U.S. Agency
Francophone Africa, In R. Lal and B.A.Stewart, eds. Advances in Soil
for International Development. Science: Soil Management: Experimental basis for sustainability and
environmental quality.
PIRT (Projet d’Inventaire des Ressources Terrestres), 1983. Mali Land
REFERENCES and Water Ressources. Vol. III. Bamako. PIRT/TAMS.
Schulte EE, Hopkins BG (1996). Estimation of organic matter by weight
Allison LE (1935). Organic soil carbon by reduction of CrO3. Soil Sci., loss-on-ignition. In F.R. Magdoff et al. (eds.) Soil Organic Matter:
40: 311-320 Analysis and Interpretation. SSSA Spec. Publ., p. 46. SSSA,
Antle J, Uehara G (2002). “Creating Incentives for Sustainable Madison, WI, pp. 21-31
Agriculture: Defining, Estimating Potential and Verifying Compliance Stevenson FJ (1982). Humus Chemistry: Genesis, Composition,
with Carbon Contracts for Soil Carbon Projects in Developing Reactions. John Wiley AND Sons Inc., pp. 374-390.
Countries.” In A Soil Carbon Accounting and Management System for Swift SR (1996). Organic Matter Characterization. In Methods of Soil
Emissions Trading. Special Publication SM CRSP 2002-04, Soil Analysis Part 3, Chemical Methods, Madison WI: Book ser. Nº5.
Management Collaborative Research Support Program, University of SSSA, pp. 1011-1069.
Hawaii, Honolulu, Hawaii, pp. 1-11. Traoré CS, Abdou B (2009). Sequestration of organic carbon in West
Ben-Dor E, Banin A (1989). Determination of organic matter content in African soils by Aménagement en Courbes de Niveau. Agronomy
arid zone soils using a simple “loss-on-ignition” method. Sustain. Dev., 29:267-275.
Commununications in Soil Science and Plant Analysis 20:1675-1695. University of Georgia (2003). Carbon from Communities, a Satellite
Davies BE (1974). Loss-`on-ignition as an estimate of soil organic view. NAG 13-2003.
matter. Soil Science Society of America. Proceedings, 38: 150 - 151. Walkley AJ, Black IA (1934). An estimation of the Degtjareft method for
Doumbia M, Ansumana J, Modou S, Kalifa T, Russell Y, Richard K, determining of soil organic matter and a proposed modification of
Kevin B, Abou B, Charles Y, Antonio Q, Pierre C, Traoré S, Abdou B chromic acid titration method. Soil Sci., 37: 29-38.
(2009).Sequestration of organic carbon in West African soils by White JL, Roth CB (1984). Infrared Spectrometry. In Methods of Soil
Aménagement en Courbes de Niveau. Agronomy for Sustainable Analysis, Part 3. Chemical Methods. Madison, WI: Book Ser.
Development, 29: 267-275. Nº5.SSSA, pp. 291-326.
Gee GW, Bauder JW (1986). Particle-size analysis. In Methods of soil
analysis, Part 3. Chemical methods. Book Series No. 5. Madison,
Wisconsin: SSSA, pp. 383-409,
Heanes DL (1984). Determination of Total Organic C in Soils by
Improved Chromic Acid Digestion and Spectrophotometric
Procedure. Commununications in Soil Science and Plant Analysis,
15(10): 1191-1213