HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Figures Only Full Text (PDF) Free Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in Web of Science Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Web of Science (5) Citing Articles via Google Scholar Google Scholar Articles by Fontes, M. P. F. Articles by Carvalho, I. A. Search for Related Content PubMed Articles by Fontes, M. P. F. Articles by Carvalho, I. A., Jr. GeoRef GeoRef Citation Agricola Articles by Fontes, M. P. F. Articles by Carvalho, I. A. Related Collections Tropical Soils Pedology Soil Classification and Mapping Soil Mineralogy

Soil Mineralogy

Color Attributes and Mineralogical Characteristics, Evaluated by Radiometry, of Highly Weathered Tropical Soils

^a Dep. de Solos, Univ. Federal de Viçosa, 36570-000, Viçosa, MG, Brazil
^b Faculdade Politécnica de Uberlândia, 38400-436, Uberlândia, MG, Brazil

^* Corresponding author (mpfontes{at}ufv.br)

	ABSTRACT

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS AND DISCUSSION CONCLUSIONS REFERENCES

 
Highly weathered soils are important around the world. Determination of their observed colors and acquisition of spectral curves were used to study their mineralogy. Color was determined through visual comparison with Munsell Soil Color Charts, and soils spectral characterization was made with an infrared intelligent spectroradiometer (IRIS) covering a spectral window from 300 to 3000 nm. The variables of color behaved differently among soils. The Munsell hue was related to hematite content and the hematite/goethite ratio (Hm/Gt), and the Munsell value to organic content of goethitic soils. The Munsell chroma was related to the presence of Fe oxides, giving an idea of relative amounts of either hematite or goethite. Regression equations relating color indices showed that the redness factor (RF) gave the highest determination coefficients to estimate hematite content and Hm/Gt, and reconfirmed that redness saturation occurs with increasing hematite amounts. The typical minerals of these soils showed characteristic absorption bands in their spectral curves. Absorption bands of 450 and 900 nm for Fe oxides, 2200 nm for kaolinite and 2280 for gibbsite, were distinctive features of the main minerals present. Organic matter decreased the intensity of reflectance for most soils and, under its influence, the characteristic bands of goethite practically disappeared while the ones for hematite, kaolinite, and gibbsite were attenuated. High amounts of hematite decreased the reflectance factor strongly as compared with the soil organic matter (SOM). Qualitative mineralogy of highly weathered soils was assessed by radiometry using measurements in the visible and near infrared regions of the electromagnetic spectrum.

Abbreviations: BRF, bidirectional reflectance factor • HH, Humic Hapludox • Hm/Gt, hematite/goethite ratio • IRIS, infrared intelligent spectroradiometer • PT, Plinthaquox • RA, Rhodic Acrudox • RE, Rhodic Eutrudox • RF, redness factor • RH, Rhodic Hapludox • RR, redness rating • SOM, soil organic matter • TH, Typic Hapludox • XH, Xanthic Hapludox

	INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS AND DISCUSSION CONCLUSIONS REFERENCES

 
IN DELINEATING DIFFERENCES between soils and in describing the characteristics of a soil profile, color is one of the most obvious and useful attributes for documenting these differences (Stoner et al., 1980a). Color is readily determined, and because it reflects differences in mineralogical, organic, and textural composition, it is very important in soil identification, particularly with the highly weathered soils. The organic matter and Fe oxides are the most important pigmenting agents for the color of highly weathered soils. These pigmenting agents act, in general, on a white background of the clay minerals, specially kaolinite and gibbsite. The Fe oxides appear mainly as goethite, which is responsible for the yellowish and brownish colors, and hematite, which imparts the red color. Good relationships between the color of Brazilian soils and the nature of the Fe oxides have been reported (Bigham et al., 1978; Kämpf & Schwertmann 1983a, 1983b; Fontes, 1988). As hematite presents a high pigmenting power, it is very effective in masking the yellow color of goethite. According to Scheffer et al. (1958), only 1.7% hematite could impart red color to a soil and merely 1% hematite would be enough to turn a yellow soil red (Resende, 1976).

The common method for determining this important soil property is a visual comparison between a given soil sample and the various color chips in an array of artificially produced Munsell colors (Munsell Color Company, 1975) arranged according to hue, value, and chroma (Stoner et al., 1980a). The use of the Munsell Soil Color Chart is practically mandatory in soil surveys, and the complete color profile should be presented for every soil examined and described in the field (Soil Survey Staff, 1960). The color attribute is widely used in the Brazilian Soil Classification System, where some classes of soils, particularly Latosols and Argisols, (approximately equivalent to Oxisols and Ultisols–Alfisols, respectively, in U.S. soil taxonomy) are defined and separated based on the field color they present (Empresa Brasileira de Pesquisa Agropecuária, 1999).

On the basis of the Munsell notation, color indices were developed and are being used. Hurst (1977) introduced the concept of converting the Munsell Color notation to a single number by defining a variable, H*, to substitute for the Munsell hue and multiplying it by the ratio of value divided by chroma (H* x chroma/value). On that basis, Torrent et al. (1980) proposed the redness rating (RR) as a numerical index that could measure quantitatively the relationship between the redness and the hematite content of the soil. To make up this index, Munsell color was converted to RR according to the expression RR = H x C/V, where C and V are the numerical values of Munsell chroma and value, respectively, and H is 12.5 for hue 7.5R; 10 for 10R; 7.5 for 2.5YR; 5 for 5YR; 2.5 for 7.5YR and 0 for 10YR. Later, Torrent et al. (1983) used a slight variation in how to calculate the index using the formula RR = (10 – H) x C/V, where C and V are the same as defined before and H is the figure preceding YR in the Munsell hue. Kämpf & Schwertmann (1983a) and Fontes & Weed (1996) applied this index to highly weathered Brazilian soils, and they found significant correlations between RR and the Hm/Gt. Later, in an attempt to improve this index quantification, Santana (1984) proposed the use of the redness factor (RF). In this case, the index was defined as RF = (10 – H) + C/V using the same variables as defined before. Santana claims that this slight change in replacing the operator x by + improved the correlations between the index and either hematite content or the Hm/Gt.

Soil color is related to numerous other soil properties and precise quantitative reflectance measurements in the visible and near infrared portion of the electromagnetic spectrum often reveal textural, structural, mineralogical and/or other significant differences which may not be detectable by standard color observations (Stoner et al., 1980a). Quantitative measurements of visible as well infrared reflectance spectra of soils are possible using spectroradiometric techniques developed to simulate the geometry of remotely sensed data (Stoner and Baumgardner, 1981). Interest in remote sensing of earth resources emphasizes the need to accurately define soil spectral reflectance to aid in the interpretation of color aerial photography and multispectral imagery which, in turn, can be used with automatic data processing for the identification and subsequent mapping of soils (Matthews et al., 1973). To develop techniques whereby satellite sensors may be used most effectively in the preparation of field soil maps, land use capability maps, and soil productivity ratings as they relate to food production, a better understanding of the relationships existing between the spectral reflectance of soils and the soil parameters important in differentiating soils is required (Coleman et al., 1991).

Radiometric data on soil reflectance has been increasingly used for tropical soils in the last decade for various purposes, mainly as a basis for soil mapping and land use (Epiphanio et al., 1992; Valeriano et al., 1995; Formaggio et al., 1996; Demattê et al., 1998; Demattê & Garcia, 1999; Pizarro, 1999; Clemente et al., 2000). Demattê & Garcia (1999) and Clemente et al. (2000) showed that variations in the degree of weathering promoted alterations in soil properties which were detected by the reflectance intensity attributed mainly to the differences in mineralogy, OM content, texture, weatherable material, and differences in Fe forms. The organic matter content of soil has an important effect in the spectral characteristics of the soils (Clemente et al., 2000), and it has the capacity to decrease the reflectance of the soils along the whole spectral curve, particularly in the range of 400 to 2500 nm (Hoffer & Johannsen, 1969). It can mask features, peaks, and depressions provoked by the other components.

A highly weathered soil mineralogy is dominated by Fe oxides such as hematite, and goethite, kaolinite, and gibbsite (Fontes & Weed, 1991), each of which have a definite influence on the soil's reflectance behavior. The Fe oxides have a large influence on both form and intensity of the spectral curve of the soils. Goethite influences by producing bands in 435, 480, 650, and 917 nm and hematite imparts bands in 445, 530, 650, and 885 nm, making it difficult to differentiate between these two minerals (Beck et al., 1976). For other Fe oxides, such as magnetite and ilmenite, the reflectance is low in the whole visible and near infrared spectrum, and they are considered opaque minerals (Strens and Wood, 1979; Epiphanio et al., 1992).

Kaolinite presents moderate absorption bands at 1400 and 1900 nm, and strong absorption at 2200 nm (Matthews et al., 1973). Clark (1999) almost disregarded the 1900-nm band for kaolinite, saying that a mineral whose spectrum has a 1900-nm absorption band contains water, such as hectorite and halloysite, but a spectrum that contains a 1400-nm band without the 1900-nm band indicates that only hydroxyl is present, as in kaolinite. According to Madeira Netto (1996), the kaolinite spectrum has characteristically sharp features in the reflected infrared region, centered at 1400 and 2200 nm due to, respectively, the overtones of fundamental OH^– stretching and to combinations involving OH^– stretching and Al-OH bending modes. Gibbsite presents spectral features due to OH^– vibrations with the near and medium infrared spectra showing the stretching harmonic (1550 nm) and the combination of stretching and bending mode (2300 nm; Madeira Netto, 1996).

The objectives of this study were to investigate the soil color attributes of highly weathered soils and to analyze the influence of the soil color main determining factors, such as mineralogy and organic matter, on soil spectroradiometric behavior.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS AND DISCUSSION CONCLUSIONS REFERENCES

 
Samples of the most representative highly weathered soils from Minas Gerais State, Brazil, were collected. Thirteen profiles covering a wide range of climatic conditions and parent materials from the southern part of the State were selected and A, B, and A/B horizons were sampled. Brazilian soil classification, approximately equivalent to soil taxonomy (Soil Survey Division Staff, 1993) and parent material of the samples can be seen in Table 1. Location of the sampling is shown in Fig. 1 .

View larger version (14K): [in this window] [in a new window]   Fig. 1. Brazilian map showing Minas Gerais State and the location of the profiles sampled.

Redness rating (Torrent et al., 1983) was determined as

and the RF (Santana, 1984) as

where C and V are the numerical values of the Munsell chroma and value, respectively, and H is the figure preceding YR in the Munsell hue, so that it is 10 for 10R; 7.5 for 2.5YR; 5 for 5YR; 2.5 for 7.5YR, and 0 for 10YR.

Hematite and goethite contents (Table 2) were calculated by allocation (Resende et al., 1987) based on the procedure described by Jones (1981), using data from sulfuric acid digestion, from where the Fe₂O₃ content (Table 2) was determined (Carvalho, 2000). The sulfuric acid digestion was performed on the fine earth by extraction with H₂SO₄/H₂O 1:1 by volume, heating until boiling under reflux, with subsequent cooling, dilution, and filtration. Fe₂O₃ was determined in the filtrate (Empresa Brasileira de Pesquisa Agropecuária, 1997). The Hm/Gt (Table 2) was calculated using the expression

Spectral Characterization
The spectral characterization of the soils was made using the sensor system Dual Field of View Mark IV IRIS (Epiphanio et al., 1992). The instrument covers a spectral window from 300 to 3000 nm, with spectral resolution of 2 nm between 300 to 1000 nm and 4 nm between 1000 and 3000 nm, in a continuous scan. The system is operated by a microcomputer programmed to measure and to store the data in memory, besides supplying output in diskettes.

The light source, a 600-W halogen lamp with noncollimated rays, was positioned over the reference plate and over the ground samples, previously spread in 10-cm-diam. Petri dishes, forming a 1-cm layer of soil. Detailed description of the instrument can be found elsewhere (Valeriano et al., 1995). The ratio of the flux reflected by an object under specified conditions of irradiation and viewing (soil sample) to that reflected by the ideal, completely reflecting, perfectly diffusing surface, identically irradiated and viewed (standard) was taken as the bidirectional reflectance factor (BRF) [Nicodemus et al. (1977), cited by Stoner et al., 1980b]. The restriction applied is that the measurements have to be made through negligibly small solid angles of illumination and viewing [Nicodemus et al. (1977), cited by Stoner et al., 1980b]. Then, the BRF was plotted against wavelength expressed in nanometers, which denotes the portion of the electromagnetic spectrum under consideration. Wavelengths used in this paper are in the visible (400–700 nm) and in the near-infrared (701–2500 nm) regions.

	RESULTS AND DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS AND DISCUSSION CONCLUSIONS REFERENCES

 
Munsell Hue, Value, and Chroma
The soil colors determined using the Munsell Soil Color Chart are presented in Table 2. The soils are arranged in such way that the first ones are the yellower soils with hues 5Y, 2.5Y, 10YR, and 7.5YR, and following them are the redder soils with hues 5YR, 2.5YR, and 10R (Table 2). The first six soils have no or low hematite content and predominance of goethite. This group of soils is also characterized by parent materials poorer in Fe content (Table 1), high SOM content (Table 2), and humid soil moisture regimes which favor the formation of goethite as compared with hematite (Kämpf & Schwertmann, 1983a; Schwertmann, 1985, 1988). The last seven soils are derived from mafic rocks, itabirite, and calcareous materials (Table 1). The soils derived from the mafic rocks and itabirite are characterized by high Fe₂O₃ contents, products of the weathering of rocks with high Fe release under tropical climates. These conditions favor the formation of ferrihydrite, a necessary precursor for hematite (Schwertmann, 1985, 1988), inducing the predominance of hematite as compared with goethite. The soil derived from the calcareous material contains a reasonable amount of Fe (Table 1) that, in conjunction with a high pH during the formation of this soil, favors the formation of hematite which predominates in terms of soil color.

The use of Munsell Soil Color Charts in field observations of Brazilian soils has shown that hues 5Y, 2.5Y, and 10YR indicate that the soil has goethite but does not have hematite; hues 7.5YR, 5YR, 2.5YR, and 10R indicate that the soil contains a mixture of goethite and hematite; and hues 2.5YR and 10R indicate that hematite predominates (Kämpf and Schwertmann, 1983b). This is in line with the reasoning of Torrent et al. (1980), who proposed the use of a variable H taken from the hue (Table 2). According to these authors, this variable, directly related to the Munsell hue, was chosen because it provides a good discrimination for pure goethite (10YR) and pure hematite (7.5R) determined colors. In fact, there was a strong relationship between H and the Hm/Gt ratio (r = 0.933, P = 0.01, N = 28) and also between H and hematite content (r = 0.837, P = 0.01, N = 28), which seems to support their affirmative.

The data of the Munsell values for each soil (Table 2) show that there are no large variations among horizons of the same soil when they are richer in hematite and, in most cases, the values are the same (Fig. 2 , hematitic soils). On the other hand, for the goethitic soils there is an increment of the Munsell value from the A to the B horizon, particularly for the soils with a higher organic matter content, as in the Typic Hapludox 2 (TH2) and Humic Hapludox (HH), where the increment in Munsell value was two units (Fig. 2, goethitic soils). The Munsell value, which relates directly to the darkest tonality of the color, is reduced in the soils by the presence of the organic matter (Scheinost & Schwertmann, 1999). Thus, the value can reflect the largest organic matter content of the surface horizons and because it has a weaker coloring power, goethite suffers a greater masking influence of the organic matter content in the surface horizon that translates into larger darkening. Hematite, due to its higher coloring power and darker color, does not suffer this interference as much. Therefore, in the hematitic soils, there is no clear differentiation between the A and B horizons reflected in the Munsell values. These data demonstrate why a misconception persisted for a long time that the tropical soils, in general, are low in organic matter. According to Greenland et al. (1992), the myth used to be stated as: "There is negligible organic matter in tropical soils..." The authors still say that the origins of the myth lie in the explorers' tales of the red color, showing little indication of the black or brown colors of the better soils of the temperate regions. In other words, the reason for this old misconception seems to be the fact that the hematitic soils, even with high organic matter content, do not show a clear color differentiation among A and B horizons of their actual content of organic matter.

View larger version (18K): [in this window] [in a new window]   Fig. 2. Munsell values of A and B horizons of some goethitic and hematitic soil samples.

Color Indices
Table 2 presents the values for the RR and the RF of the soils. The values ranged from 0 to 20, with smaller values for the more yellowish soils and higher values for the more reddish soils. The Rhodic Acrudox 2 showed maximum values in both indices.

Because of the high coloring power of hematite, numerical color indices such as the RR have been shown to correlate well with hematite content in soils (Childs et al., 1979; Torrent et al., 1980, 1983), therefore, the relationship between RR and RF and hematite was tested. Figure 3 shows the regression lines between RR, RF, and hematite content for all samples. The linear regressions yielded determination coefficients of 0.691 and 0.688, respectively, which indicate that the linear relationship of the data was not very strong. The relationship between these variables improved the R² substantially (0.845 and 0.930) when a quadratic polynomial was tried, demonstrating a large improvement in the data fit. This behavior clearly demonstrates that a redness saturation occurred as hematite content increased, as described by Torrent et al. (1983), causing a lack of linearity between the variables, due to the pigmenting power of hematite.

View larger version (14K): [in this window] [in a new window]   Fig. 3. Regression equations for the relationship between hematite content and (a) the redness rating (RR) and (b) the redness factor (RF).

View larger version (14K): [in this window] [in a new window]   Fig. 4. Regression equations for the relationship between hematite content and (a) redness rating (RR) and (b) redness factor (RF), excluding Rhodic Acrudox 2.

View larger version (11K): [in this window] [in a new window]   Fig. 5. Regression equations for the relationship between hematite/goethite ratio and (a) the redness rating (RR) and (b) the redness factor (RF), excluding Rhodic Acrudox 2.

Group 1 is formed by the PT_B, TH3_B, and Xanthic Hapludox (XH_B) (Fig. 6) . In spite of having a large variation for the reflectance intensities among the samples, the shapes of the spectral curves were very similar. There was an increase in the reflectance intensity up to 800 nm, and several reflectance absorption bands appeared along the curve at 1400, 1500, 1900, 2200, and 2280 nm, which are indicative of kaolinite and gibbsite (Matthews et al., 1973; Madeira Netto, 1996; Clark, 1999). The reflectance factor reached the value of 0.4 for the PT_B sample, and it exceeded 0.6 for the TH_B sample. The PT_B sample has very low goethite content and no hematite. The TH_B sample possesses low Fe oxide content (3.7% goethite), an intermediate amount of kaolinite, and higher amount of gibbsite (Table 2). The increase in the intensity of the reflectance, together with the clear change of slope of the spectral curve around 650 nm, seem to be linked to the presence of the goethite, while the sharpening of the bands at 1400, 1500, 2200, and 2280 nm are related to the presence of kaolinite and gibbsite in that sample. The band at 2280 nm, indicative of gibbsite, showed that the PT_B sample is richer in that mineral. The XH_B sample presents kaolinite, gibbsite, and goethite in the clay fraction (Table 2). Its spectral curve reflected that mineralogy, with very subtle concavities at 450 and 900 nm, which indicated the presence of small amounts of Fe oxides, and the slope of the line suggested that the Fe oxide is goethite. The 1400 and 2200 nm bands showed the presence of kaolinite, whereas the band at 2280 nm revealed the presence of gibbsite.

View larger version (19K): [in this window] [in a new window]   Fig. 6. Spectroradiometric curves of the B horizons of the Typic Hapludox 3 (TH3B), Plinthaquox (PTB), and Xantic Hapludox (XHB). BRF = bidirectional reflectance factor.

View larger version (17K): [in this window] [in a new window]   Fig. 7. Spectroradiometric curves of the B horizons of the Typic Hapludox 1 and 2 (TH1B and TH2B), and Humic Hapludox (HHB). BRF = bidirectional reflectance factor.

View larger version (18K): [in this window] [in a new window]   Fig. 8. Spectroradiometric curves of the B horizons of the Rhodic Hapludox 1 and 2 (RH1B and RH2B), Rhodic Eutrudox 1 (RE1B). BRF = bidirectional reflectance factor.

View larger version (18K): [in this window] [in a new window]   Fig. 9. Spectroradiometric curves of the B horizons of the Rhodic Acrudox 1 and 2 (RA1B and RA2B) and Rhodic Eutrudox 2 (RH3B). BRF = bidirectional reflectance factor.

Organic Matter Influence
The influence of organic matter on the radiometric behavior of these soils was studied by comparing the A and B horizons of each sample analyzed. The comparison is possible because the Latosols are deep soils with virtually no weatherable primary minerals and very little differentiation among horizons (Resende et al., 1995). Therefore, the content and species of Fe oxides would be very similar for the A and B horizons, making a direct comparison less problematic as it could be for most soils.

The SOM has a common interference on the spectral curves of the minerals kaolinite and gibbsite, showing that, in general, SOM attenuated the absorption bands characteristic for these minerals. For the Fe oxides, important pigmenting agents in soils, the influence of the organic matter was different according to the type of mineral involved.

The soil sample from the TH3 has no hematite, a small amount of goethite, and a reasonable amount of SOM (Table 1). It has the highest reflectance factor in its B horizon (Fig. 10) , and in the A horizon the organic matter decreased the reflectance factor in almost all the entire spectrum. The organic matter attenuated the absorption bands for kaolinite and gibbsite at 1400, 2200, and 2280 nm.

View larger version (16K): [in this window] [in a new window]   Fig. 10. Spectroradiometric curves of the A and B horizons of the Typic Hapludox 3 (TH3A and TH3B). BRF = bidirectional reflectance factor.

View larger version (18K): [in this window] [in a new window]   Fig. 11. Spectroradiometric curves of the A, AB, and B horizons of the Humic Hapludox (HHA, HHA/B, and HHB). BRF = bidirectional reflectance factor.

View larger version (16K): [in this window] [in a new window]   Fig. 12. Spectroradiometric curves of A and B horizons of the Typic Hapludox 2 (TH2A and TH2B). BRF = bidirectional reflectance factor.

View larger version (16K): [in this window] [in a new window]   Fig. 13. Spectroradiometric curves of the A and B horizons of the Rhodic Hapludox 2 (RH2A and RH2B). BRF = bidirectional reflectance factor.

View larger version (16K): [in this window] [in a new window]   Fig. 14. Spectroradiometric curves of the A and B horizons of the Rhodic Acrudox 1 (RA1A and RA1B). BRF = bidirectional reflectance factor.

View larger version (16K): [in this window] [in a new window]   Fig. 15. Spectroradiometric curves of the A and B horizons of the Rhodic Acrudox 2 (RA1A and RA2B). BRF = bidirectional reflectance factor.

	CONCLUSIONS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS AND DISCUSSION CONCLUSIONS REFERENCES

 
The variables that comprise conventional soil color notation behaved differently in their relationship with soil characteristics. The Munsell hue had a significant relationship with hematite content and the Hm/Gt ratio. The Munsell value was related directly to the amount of organic matter in the more goethitic soils, but not in the hematitic ones. The Munsell chroma was related to the amount of Fe oxides, and in conjunction with the hue, chroma was able to give an idea of the relative amounts of hematite or goethite present in the soil.

Numerical indices based on the Munsell notation showed a good relationship with soil characteristics. The RR can be a good predictor of hematite contents, whereas the RF can be useful to predict the Hm/Gt in highly weathered soils, in both cases, for soils containing less than 100 g kg^–1 of hematite. This is because of the redness saturation phenomenon that appears as hematite content increases to higher levels.

The qualitative mineralogy of highly weathered soils can be evaluated by radiometry using measurements in the visible and near-infrared regions of the electromagnetic spectrum. Hematite and goethite, the main forms of Fe oxide, have characteristic concavities at about 450 and 900 nm, and hematite decreases whereas goethite increases soil reflectance, allowing for a distinction between them. Kaolinite has strong absorption bands at 1400 and 2200 nm and can be distinguished from gibbsite by the absorption bands at 1500 and 2280 nm.

Organic matter decreased the reflectance intensity for most soils, attenuating the typical absorption bands of the minerals kaolinite and gibbsite. For the Fe oxides, organic matter's influence depended on the type of mineral. For goethite, organic matter decreased the reflectance factor and attenuated the absorption bands for all spectrum more strongly than it did for hematite. High amounts of hematite even reversed this behavior, showing the high coloring power of this Fe oxide.

Spectral curves can be a good auxiliary tool to study the mineralogy of highly weathered soils, typical of the tropical climates.

	ACKNOWLEDGMENTS

 
This research, a part of the project "Carbon dynamics in Latosols of Minas Gerais State," was supported by an award of FAPEMIG–Fundação de Amparo à Pesquisa do Estado de Minas Gerais, through the fellowship CAG-2260/96. Thanks are extended to the associate editor and the anonymous reviewers who helped to improve the manuscript.

Received for publication December 2, 2003.

	REFERENCES

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS AND DISCUSSION CONCLUSIONS REFERENCES

 

• Beck, R.H., B.F. Robinson, W.W. McFee, and J.B. Peterson. 1976. Spectral characteristics of soils related to the interaction of moisture, organic carbon, and clay content. LARS Information Note 081176, Purdue Univ., West Lafayette, IN.
• Bigham, J.M., D.C. Golden, S.W. Buol, S.B. Weed, and L.H. Bowen. 1978. Iron mineralogy of well-drained Ultisols and Oxisols: II. Influence on color, surface area, and phosphate retention. Soil Sci. Soc. Am. J. 42:825–830.[Abstract/Free Full Text]
• Carvalho, I.A., Jr. 2000. Influence of mineralogy and organic matter on aggregation, color and radiometry of highly weathered soils from Minas Gerais State. (In Portuguese, with English summary.) D.S. thesis. Univ. Federal de Viçosa, Viçosa, MG, Brazil.
• Childs, C.W., B.A. Goodman, and G.J. Churchman. 1979. Application of Mössbauer spectroscopy to the study of iron oxides in some red and yellow/brown soil samples from New Zealand. p. 555–565. In M.M. Mortland and V.C. Farmer (ed.) 6th Int. Clay Conf. 1978. Elsevier, Amsterdam.
• Clark, R.N. 1999. Spectroscopy of rocks and minerals, and principles of spectroscopy. p. 3–58. In N. Rencz (ed.) Manual of remote sensing. Vol. 3. Remote sensing for the earth sciences. John Wiley & Sons, New York.
• Clemente, C.A., J.A.M. Demattê, A.L. Mafra, and S.R.P. Bentivenha. 2000. Spectral reflectance and mineralogy of materials developed from diábase. (In Portuguese with English summary.) Sci. Agric. 57:159–168.
• Coffin, D.E. 1963. A method for the determination of free iron in soil and clays. Can. J. Soil Sci. 43:7–17.
• Coleman, T.L., P.A. Agbu, O.L. Montgomery, T. Gao, and S. Prasad. 1991. Spectral band selection for quantifying selected properties in highly weathered soils. Soil Sci. 151:355–361.
• Demattê, J.A.M., and G.J. Garcia. 1999. Alteration of soil properties through a weathering sequence as evaluated by spectral reflectance. Soil Sci. Soc. Am. J. 63:327–342.[Abstract/Free Full Text]
• Demattê, J.A.M., G.J. Garcia, and L.I. Prochnow. 1998. Influence of induced chemical properties of three soils from Paraná State on the spectral reflectance. (In Portuguese with English summary.) Rev. Bras. Cienc. Solo 22:479–490.
• Empresa Brasileira de Pesquisa Agropecuária. 1997. Manual of methods of soil analyses. 2nd ed. (In Portuguese.) Centro Nacional de Pesquisa de Solos, Rio de Janeiro, Brazil.
• Empresa Brasileira de Pesquisa Agropecuária. 1999. Brazilian System of Soil Classification. (In Portuguese.) Centro Nacional de Pesquisa de Solos, Rio de Janeiro, Brazil.
• Epiphanio, J.C.N., A.R. Formaggio, M.M. Valeriano, and J.B. Oliveira. 1992. Spectral behavior of soils from São Paulo State. (In Portuguese.) São José dos Campos, Brazil.
• Fernandez, R.N., and D.G. Schulze. 1992. Munsell colors of soils simulated by mixtures of goethite and hematite with kaolinite. Z. Pflanzenernähr. Bodenkd. 155:473–478.
• Fontes, M.P.F. 1988. Iron oxide mineralogy in some Brazilian Oxisols. Ph.D. thesis. North Carolina State Univ., Raleigh.
• Fontes, M.P.F., and S.B. Weed. 1991. Iron oxides in selected Brazilian Oxisols. I. Mineralogy. Soil Sci. Soc. Am. J. 55:1143–1149.[Abstract/Free Full Text]
• Fontes, M.P.F., and S.B. Weed. 1996. Phosphate adsorption by clays from Brazilian Oxisols: Relationships with specific surface area and mineralogy. Geoderma 72:37–51.[CrossRef][Web of Science]
• Fontes, M.P.F., T.S. Oliveira, L.M. Costa, and A.A.G. Campos. 2000. Magnetic separation and evaluation of magnetization of Brazilian soils from different parent materials. Geoderma 69:81–99.[CrossRef]
• Formaggio, A.R., J.C.N. Epiphanio, M.M. Valeriano, and J.B. Oliveira. 1996. Spectral behavior (450–2450 nm) of tropical soils from São Paulo State, Brazil. (In Portuguese with English summary.) Rev. Bras. Cienc. Solo 20:467–474.
• Greenland, D.J., A. Wild, and D. Adams. 1992. Organic matter dynamics in soils of the Tropics—From myth to complex reality. p. 17–33. In R. Lal & P.A. Sanchez (ed.) Myths and science of soils of the Tropics. SSSA Special Publ. 29. SSSA and ASA, Madison, WI.
• Hoffer, R.M., and C.J. Johannsen. 1969. Ecological potentials in spectral signature analysis. In P.L. Johnson (ed.) Remote sensing in ecology. p. 1–16. Univ. Georgia Press, Athens.
• Hurst, V.J. 1977. Visual estimation of iron in saprolite. Geol. Soc. Am. Bull. 88:174–176.[Abstract/Free Full Text]
• Jones, R.C. 1981. X-Ray diffraction line profile analysis versus phosphorus sorption by 11 Puerto Rican soils. Soil Sci. Soc. Am. J. 45:818–825.[Abstract/Free Full Text]
• Kämpf, N., and U. Schwertmann. 1983a. Goethite and hematite in a climosequence in Southern Brazil and their application in classification of kaolinitic soils. Geoderma 29:27–39.
• Kämpf, N., and U. Schwertmann. 1983b. Relations between iron oxides and soil colour in kaolinitic soils of Southern Brazil. (In Portuguese with English summary.) Rev. Bras. Cienc. Solo 7:27–31.
• Madeira Netto, J.S. 1996. Spectral reflectance properties of soils. Photo Interp. 2:59–70.
• Matthews, H.L., R.L. Cunningham, and W. Petersen. 1973. Spectral reflectance of selected Pennsylvania soils. Soil Sci. Soc. Am. J. 37:421–424.
• Munsell Color Company. 1975. Munsell soil color charts. Macbeth Division of Kollmorgen Corp., Baltimore.
• Nelson, O.W., and L.E. Sommers. 1982. Total carbon, organic carbon, and organic matter. p. 539–577. In A.L. Page et al. (ed.) Methods of soil analysis: Part 2—Chemical and microbiological properties. 2nd ed. Agron. Monogr. 9. ASA and SSSA, Madison, WI.
• Nicodemus, F.E., J.C. Richmond, J.J. Hsia, I.W. Insberg, and T. Limperis. 1977. Geometrical considerations and nomenclature for reflectance. National Bureau of Standards Monogr. 160. U.S. Gov. Print. Office, Washington, DC.
• Pizarro, M.A. 1999. Remote sensing hiper-spectral for identification and characterization of minerals in tropical soils. (In Portuguese with English summary.) MS thesis. Inst. de Pesquisas Espaciais, São José dos Campos, Brazil.
• Resende, M. 1976. Mineralogy, chemistry, morphology and geomorphology of some soils of Central Plateau of Brazil. Ph.D. thesis. Purdue Univ., West Lafayette, IN.
• Resende, M., A.F.C. Bahia Filho, and J.M. Braga. 1987. Clay mineralogy of Latosols estimated by allocation from the total content of oxides from sulfuric attack. (In Portuguese with English summary.) Rev. Bras. Cienc. Solo 11:17–23.
• Resende, M., N. Curi, S.B. Rezende, and G.F. Corrêa. 1995. Pedology: Basis for environmental distinction. (In Portuguese.) NEPUT, Viçosa, Brazil.
• Santana, D.P. 1984. Soil formation in a toposequence of oxisols from Patos de Minas region, Minas Gerais State, Brazil. Ph.D. thesis. Purdue Univ., West Lafayette, IN.
• Scheffer, F., E. Welte, and F. Ludwieg. 1958. Zur Frage der Eisenoxidhydrate im Boden. (In German.) Chem. Erde 19:51–64.
• Scheinost, A.C., and U. Schwertmann. 1999. Color identification of iron oxides and hydroxysulfates: Use and limitations. Soil Sci. Soc. Am. J. 63:1463–1471.[Abstract/Free Full Text]
• Schwertmann, U. 1985. The effect of pedogenic environments on iron oxide minerals. Adv. Soil Sci. 1:171–200.
• Schwertmann, U. 1988. Occurrence and formation of iron oxides in various pedoenvironments. In J.W. Stucki et al. (ed.) Iron in soils and clay minerals. p. 267–308. NATO Advanced Studies Institute Ser. C. Vol. 217. Reidel Publ., Boston, MA.
• Soil Survey Division Staff. 1993. Soil survey manual. USDA Handbook no. 18. U.S. Gov. Print. Office, Washington, DC.
• Soil Survey Staff. 1960. Soil classification. A comprehensive system, 7th approximation. U.S. Gov. Print. Office, Washington, DC.
• Stoner, E.R., and M.F. Baumgardner. 1981. Characteristic variation in reflectance of surface soils. Soil Sci. Soc. Am. J. 45:1161–1165.[Abstract/Free Full Text]
• Stoner, E.R., M.F. Baumgardner, L.L. Biehl, and B.F. Robinson. 1980a. Atlas of soil reflectance properties. Res. Bull. 962. Agric. Exp. Stn., Purdue Univ., West Lafayette, IN.
• Stoner, E.R., M.F. Baumgardner, R.A. Weismiller, L.L. Biehl, and B.F. Robinson. 1980b. Extension of laboratory measured soil spectra to field conditions. Soil Sci. Soc. Am. J. 44:572–574.[Abstract/Free Full Text]
• Strens, R.G.J., and B.J. Wood. 1979. Diffuse reflectance spectra and optical properties of some iron and titanium oxides and oxyhydroxides. Mineral. Mag. 43:347–354.
• Torrent, J., U. Schwertmann, H. Fechter, and F. Alferez. 1983. Quantitative relationships between soil color and hematite content. Soil Sci. 136:354–358.
• Torrent, J., U. Schwertmann, and D.G. Schulze. 1980. Iron oxide mineralogy of some soils of two river terrace sequences in Spain. Geoderma 23:191–208.
• Valeriano, M.M., J.C.N. Ephifanio, A.R. Formaggio, and J.B. Oliveira. 1995. Bi-directional reflectance factor of 14 soil classes from Brazil. Int. J. Rem. Sens. 16:113–128.

This article has been cited by other articles:

				N. Richter, T. Jarmer, S. Chabrillat, C. Oyonarte, P. Hostert, and H. Kaufmann Free Iron Oxide Determination in Mediterranean Soils using Diffuse Reflectance Spectroscopy Soil Sci. Soc. Am. J., January 21, 2009; 73(1): 72 - 81. [Abstract] [Full Text] [PDF]

This Article Abstract Figures Only Full Text (PDF) Free Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in Web of Science Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Web of Science (5) Citing Articles via Google Scholar Google Scholar Articles by Fontes, M. P. F. Articles by Carvalho, I. A. Search for Related Content PubMed Articles by Fontes, M. P. F. Articles by Carvalho, I. A., Jr. GeoRef GeoRef Citation Agricola Articles by Fontes, M. P. F. Articles by Carvalho, I. A. Related Collections Tropical Soils Pedology Soil Classification and Mapping Soil Mineralogy