SSSAJ Journal of Natural Resources and Life Sciences Education
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Published online 1 January 2007
Published in Soil Sci Soc Am J 71:64-74 (2007)
DOI: 10.2136/sssaj2006.0111
© 2007 Soil Science Society of America
677 S. Segoe Rd., Madison, WI 53711 USA
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (7)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Scheel, T.
Right arrow Articles by Kalbitz, K.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Scheel, T.
Right arrow Articles by Kalbitz, K.
Agricola
Right arrow Articles by Scheel, T.
Right arrow Articles by Kalbitz, K.
Related Collections
Right arrow Humic Substances
Right arrow Soil Organic Matter
Right arrow Carbon Sequestration

SOIL CHEMISTRY

Precipitation of Dissolved Organic Matter by Aluminum Stabilizes Carbon in Acidic Forest Soils

T. Scheel*, C. Dörfler and K. Kalbitz

Dep. of Soil Ecology, Bayreuth Center of Ecology and Environ., Research (BayCEER), Univ. of Bayreuth, D-95440 Bayreuth, Germany

* Corresponding author (thorsten.scheel{at}uni-bayreuth.de).


   ABSTRACT
TOP
 NOTES
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Dissolved organic matter (DOM) is often neglected as a factor in the formation of stable soil organic matter (OM). Precipitation of DOM by dissolved Al could contribute substantially to C retention in acidic forest soils; however, no information is available on the stability of precipitated OM against microbial decay. We investigated the stability of Al–OM precipitates against microbial decay as related to (i) DOM composition, (ii) Al speciation, and (iii) the dissolved Al/C ratio. We produced Al–OM precipitates by adding AlCl3 (molar Al/C ratios: 0.05–0.3) at pH values of 3.8 and 4.5 to DOM solutions derived from Oi and Oa horizons, from either beech (Fagus sylvatica L.) or spruce [Picea abies (L.) Karst.] litter. Between 13 and 84% of the C was precipitated, depending on pH, Al/C ratio, and the type of DOM. Precipitates were found to be enriched in aromatic C and mostly depleted in N when compared with DOM. Only 0.5 to 7.7% of precipitated C was mineralized during 7 wk of incubation. Mineralization of Al–OM precipitates was up to 28 times less than that of the respective DOM solutions. The extent of mineralization of Al–OM precipitates formed at pH 3.8 was reduced by 50 to 75% when compared with those formed at pH 4.5. The stability of precipitates against microbial decay increased with larger aromatic C content and larger C/N ratios. Our study clearly demonstrated that a large fraction of DOM can be precipitated and is thereby substantially stabilized against microbial decay.


   INTRODUCTION
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Sorption of DOM to mineral surfaces is an important and well-documented process for organic C stabilization in soils (Kaiser and Guggenberger, 2000; Kalbitz et al., 2005). Another process potentially involved in the stabilization of soil organic C against microbial decay is precipitation of DOM by polyvalent cations, such as Fe and Al (Baldock and Skjemstad, 2000). Nevertheless, little is known about stabilization of organic matter by formation of insoluble Al–OM complexes (Boudot, 1992), which could be an important pathway for the formation of stable soil OM.

The amount of precipitated OM may comprise up to 90% (Nierop et al., 2002) of the initial dissolved organic carbon (DOC). Plankey and Patterson (1987) observed that the formation of precipitates is a relatively rapid and temperature-independent process. The extent of DOM precipitation by Al depends strongly on the prevailing pH in the soil solution and the interrelated speciation of Al (Nierop et al., 2002). Larger pH values lead to an increase in precipitation of DOM by Al (Nierop et al., 2002). The formation of Al(OH)3 controls the solubility of Al at a pH >4.2 (Gustafsson et al., 2001), whereas at lower pH, the proportion of Al3+ in solution is dominant. Dissolved organic matter tends to coprecipitate in the presence of Al hydroxides rather than forming complex bindings (Boudot, 1992). The term coprecipitation will be used for the immediate adsorption of DOM on freshly formed Al(OH)3(s) as well as for bonding between DOM and Al(OH)3, leading to precipitation, because these processes are not separable. Also tridecameric Al (Al13), AlO4Al12(OH)24(OH2)127+, could be of importance for DOM precipitation. The Al13 forms at pH values >4.2 (Furrer et al., 1992; Hu et al., 2006) with Al concentrations >10–5 mol L–1. Yamaguchi et al. (2004) observed the depletion of Al13 in solutions after adding humic acid and interpreted this as an aggregation–precipitation reaction of Al13–humic acid complexes.

The solubility of the DOM itself is also affected by pH (Kalbitz et al., 2000). It increases with pH because of deprotonation of carboxylic groups. This and the changes in Al speciation with pH let us assume that the form and strength of bondings between Al and DOM depend on pH.

Furthermore, the extent of precipitate formation depends on the Al/C ratio in solution. At Al/C ratios <0.03 the major fraction of Al–OM complexes is soluble (Jansen et al., 2003). Extensive precipitation of DOM occurs at Al/C ratios exceeding 0.03. With increasing Al/C ratios also DOM precipitation increases (Nierop et al., 2002); however, only a certain fraction of DOM can be precipitated. Thus, after precipitation of this fraction, a further increase in Al/C ratios will not increase DOM precipitation. Soil solutions in acidic forest soils can have Al/C ratios of up to 0.5 (Schwesig et al., 2003; Lumsdon et al., 2005). Thus, the formation of precipitates in acidic forest soils is very likely.

The precipitation of DOM cannot be induced only by Al, but also by Fe(III) and, to a minor extent, by Fe(II) (Nierop et al., 2002). The extent of DOM precipitation by Fe is pH independent, whereas the extent of Al–OM precipitate formation depends strongly on pH (Nierop et al., 2002). The concentrations of DOC in soil solutions are often largest at low pH (Kalbitz et al., 2000; Weng et al., 2002), which is in contrast to the larger solubility of DOM at higher pH. An explanation for this could be the increase in DOM precipitation by Al with increasing pH. Further, considering that Al concentrations are generally significant in acidic forest soil solutions, the relevance of DOM precipitation by Al should be more important than that by Fe.

Dissolved organic matter compounds with a large number of functional groups (e.g., -OOH and -OH) and of high molecular weight, e.g., aromatic compounds, preferentially precipitate from DOM solutions (Julien et al., 1994; Römkens and Dolfing, 1998; Blaser et al., 1999). These compounds already have a low solubility and offer a wide range of binding sites for Al. The negative charge of the functional groups is compensated by the positive charge of the Al species [e.g., Al3+, AlOH2+, Al(OH)2+], further reducing solubility. Aromatic structures in DOM solution are highly stable, whereas carbohydrates can be considered as easily biodegradable (Kalbitz et al., 2003). Thus, we expected Al–OM precipitates to be more stable. So, precipitation of citric acid by Al led to greater stability against microbial decay (Boudot, 1992). This effect increased with increasing Al/C ratios. Although there is broad discussion on the potential contribution of Al–OM precipitates to podzolization (Gustafsson et al., 2001; Zysset and Berggren, 2001; Jansen et al., 2004), no information on the stability of such precipitates is available.

The objective of our study was to determine the stability of Al–OM precipitates as related to (i) the composition of the DOM solution, (ii) the Al speciation, affected by pH, and (iii) the influence of the Al/C ratio. Further goals were the identification of the intrinsic properties of the precipitates that govern their stability.


   MATERIALS AND METHODS
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Samples
We used soil samples from a Norway spruce site (Waldstein-Fichtelgebirge, Germany; Michalzik and Matzner, 1999) and a deciduous stand with European beech as the dominant tree species (Steinkreuz-Steigerwald, Germany; Solinger et al., 2001). Samples were taken at both sites from the litter horizon (Oi-spruce, Oi-beech; thickness: 1 cm) and the humified horizon at the spruce site (Oa-spruce; sampling depth: 5–9 cm) and a mixture of the fermented and humified horizon at the beech site (Oa-beech; sampling depth: 1–3 cm). Samples were homogenized and stored frozen (Kalbitz et al., 2003).

Preparation of Dissolved Organic Matter Solutions
Water extracts of the soil samples were prepared at 5°C with a soil to (ultrapure) water ratio of 1:10 and stirred three times manually. After 1 d, solutions were filtered through a preconditioned ceramic filter plate and filtered through a 0.2-µm membrane filter at 5°C (OE 66, Schleicher and Schuell, Dassel, Germany) to exclude microorganisms. The obtained solutions varied in DOC concentrations (Table 1). Therefore, all solutions were diluted with ultrapure water to 40 mg C L–1 to ensure comparability. We chose that specific concentration because preliminary tests showed that sufficient amounts of precipitates would be produced. The concentration was also similar to DOC concentrations at the study sites (Solinger et al., 2001; Kalbitz et al., 2004). The concentrations of Al and Fe in the DOM solutions were fairly small (Table 2), so an initial presence of precipitates could be excluded.


View this table:
[in this window]
[in a new window]

  		Table 1. Total organic carbon (TOC) content of the soils used and dissolved organic carbon (DOC) content and pH of the water extracts.

 

View this table:
[in this window]
[in a new window]

  		Table 2. Chemical compositions of the dissolved organic matter solutions used, after adjustment to 40 mg C L–1 (standard error of three replicates in parentheses).

 
Production of Precipitates
We added AlCl3·6H2O (analytical grade, Merck, Darmstadt, Germany) from a stock solution to 200 mL of the diluted DOM solutions to achieve molar Al/C ratios of 0.05, 0.1, 0.2, and 0.3. As our intention was to conduct the experiments under conditions similar to those in the field, we limited the maximum Al/C ratio to 0.3. Immediately after the addition of AlCl3, the pH was adjusted with NaOH or HCl to either 3.8 or 4.5. These two pH values were chosen to include different types of interaction of Al with OM. At pH 4.5, the importance of hydrous Al species for precipitation should be larger than at pH 3.8, where Al3+ is the dominant species (Gustafsson et al., 2001). Each treatment was performed in triplicate. All bottles were gently shaken for 24 h at 5°C to minimize mineralization before incubation. Then precipitates were separated by passing the solution through a prewashed quartz fiber filter (QF20, Schleicher and Schuell, Dassel, Germany). The properties of the filtrates were analyzed as described below.

Inoculation
Air-dried samples of Oa-spruce and Oa-beech material were used to obtain a mixed inoculum, which contained microorganisms from both sites and ensured the presence of a broad diversity of degraders. Before extraction, the soils were rewetted to a water capacity of 60% and incubated for 2 wk at 20°C to reactivate the microorganisms. The soils were shaken separately for 30 min with a 4 mM CaCl2 solution (1:2 soil/solution ratio) and filtered through 5-µm filters (SMWP 4700, Millipore, Bedford, MA). We counted the total cell number for each solution by DAPI (4'-6-diamidino-2-phenylindole) staining (Zarda et al., 1997). The solutions were combined so that the inoculum contained a similar number of cells from each solution.

Incubation
We placed the quartz fiber filter with the Al–OM precipitates into incubation flasks (120 mL) and added 60 mL of incubation solution, made up from ultrapure water, NH4NO3, and K2HPO4 (analytical grade, Merck, Darmstadt, Germany). The precipitates were not removed from the filters because we considered them as an additional surface for microorganisms. All incubation solutions were adjusted to pH 4.5 by adding HCl. The C/N and C/P ratios of the incubation solution were set to 20 (mean load of 2 mg precipitated C per filter). We added 0.5 to 1 mL of inoculum per flask, giving a total cell count of 5 x 106 per flask. Further, we incubated quartz fiber filters without precipitates under identical conditions as blank values, to measure mineralization of the added inoculum. Samples were shaken manually each day and incubated at 20°C for 7 wk, as preliminary tests had shown that this time period was sufficient to cover the largest part of C mineralization.

Analytical Methods
The CO2 concentration in the headspace of each incubation flask was measured at increasing intervals ranging from 2 d at the beginning to 9 d at the end of the experiment (gas chromatograph HP 6890, Hewlett-Packard, thermal conductivity detector, Böblingen, Germany). To ensure proper sampling of the headspace, we applied an overpressure of about 30 kPa to each flask. Ambient air pressure and the pressure in the incubation flasks were measured before each CO2 analysis. We calculated the amount of CO2 in the flasks by using the general gas equation, and added the amount of CO2 dissolved in the liquid phase, estimated from the measured pH and the respective solubility constant. The amount of C mineralized from the Al–OM precipitates was considered as the difference between the measured CO2 of the sample and the blank without precipitates.

The DOM solutions and filtrates were analyzed for total organic C and total N (High TOC, Elementar, Hanau, Germany), NO3 (ion chromatography, Dionex DX 100, Idstein, Germany), NH4 (flow injection, photometric mLE– FIA LAB, Dresden, Germany), total Al (inductively coupled plasma–optical emission spectrometer [ICP–OES, GCP Electronics, Melbourne, Australia), pH (pH 323, WTW, Weilheim, Germany), and ultraviolet (UV) absorbance at 280 nm (UVIKON 930, Bio-Tek Instruments, Bad Friedrichshall, Germany) as an estimate of the aromaticity of DOM (Chin et al., 1994; McKnight et al., 1997).

The DOM solutions were further analyzed for PO4, SO4 (ion chromatography, Dionex DX 100), and Ca, Cu, Fe, K, Mn, Na, S, Si (ICP–OES, GCP Electronics). Also 1H-NMR and 13C-nuclear magnetic resonance (NMR) spectra of freeze-dried samples of DOM (150 mg) dissolved in 3 mL 0.5 M NaOD solution were recorded on a DRX 500 NMR spectrometer (11.7 T, Bruker Analytische Messtechnik GmbH, Rheinstetten, Germany) at a temperature of 290 K under the following conditions: 13C NMR–10-mm sample tubes, spectrometer frequency–125 MHz, inverse-gated decoupling, acquisition time–0.16 s, delay time–1.84 s, line-broadening factor–100 Hz; 1H NMR: 5-mm sample tubes, spectrometer frequency–500 MHz, homonuclear presaturation for solvent suppression, acquisition time–1.16 s, delay time–1 s, line-broadening factor–2 Hz. Intensities of signals were determined by electronic integration.

Incubation solutions were analyzed at the end of the experiment for Al3+ (capillary electrophoresis, HP 3D CE, Waldbronn, Germany).

Calculation of Aluminum, Carbon, Aromatic Carbon, and Nitrogen Content of Precipitates
The Al, C, and N contents of the precipitates were calculated from the differences in Al, organic C, and organic N concentrations in solution before and after precipitation. The specific UV absorption of the DOM solutions correlated very well with the aromatic C content (r2 = 0.98) as determined by 13C NMR (Fig. 1 ). The regression was used to calculate the aromatic C content in the filtrates. The aromatic C content of the precipitates was estimated by difference, taking into account the amount of DOM precipitated.


Figure 1
View larger version (26K):
[in this window]
[in a new window]

  		Fig. 1. Specific ultraviolet (UV) absorption at 2 0 nm of the extracted dissolved organic matter solutions from the oa and oi horizons of a beech and a spruce forest in relation to their aromatic c content, determined by­ solution 1 c nuclear magnetic resonance.

 
Statistics and Modeling
We used a general linear model with repeated measurements to test for differences in the extent of mineralization between precipitates obtained (i) from different DOM solutions, (ii) at different Al/C ratios, and (iii) at different pH values. The analysis was performed with the statistical package SPSS (SPSS, 2003). This analysis required a normal distribution of the data. To meet this criterion, our data set was transformed by extracting the root of the amount of C mineralized. The variances of the transformed values were homogenous. To test the significance of the differences, Fisher's LSD posthoc test was applied. A three-factorial variance analysis was performed with the residuals of the multiple linear regression between the amount of C mineralized and the aromatic C content and the C/N ratio of the precipitates. The residuals were tested on differences originating from different (i) DOM solutions, (ii) Al/C ratios, and (iii) pH values. The significance of the differences was again tested with Fisher's LSD posthoc test.

The mineralization data were best described using an exponential model (Table 3). Fitting was done using SigmaPlot (SPSS, 2002), with the following equation:

Formula
where k is the mineralization rate constant, A is the mineralizable C (% of initial C), t is time (d), and c is a correction factor (axis intercept).


View this table:
[in this window]
[in a new window]

  		Table 3. Fitting parameters of an exponential model (Cmineralized [%] = A[1 – exp(–kt)] + c, where k is the mineralization rate constant, t is time [d], A is mineralizable dissolved organic C [%], and c is a correction factor) for the mineralization of Al–organic matter precipitates.

 
The correction factor was implemented to include axis intercepts below zero at the beginning of the incubation, which originated from the subtraction of the blank values. Fitting a double exponential model to the data was successful only in few cases.

For calculation of saturation indices of Al(OH)3, the program Visual Minteq (Gustafsson, 2005) was used.


   RESULTS AND DISCUSSION
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
CONCLUSIONS
 REFERENCES
 
Formation and Properties of Aluminum–Organic Matter Precipitates
Dissolved organic matter precipitated within minutes after adding Al to the solution, as already reported by Plankey and Patterson (1987). We found that increasing Al/C ratios led mostly to increased amounts of precipitated OM (Table 4). Aluminum/C ratios beyond ~0.1 led to only slightly increasing DOM precipitation. We consistently observed a larger fraction of DOM precipitating at pH 4.5 than at pH 3.8 (Table 4), which confirms the results of Nierop et al. (2002). The maximum proportions of DOM that precipitated from the four solutions increased in the order of Oi-beech < Oi-spruce < Oa-beech < Oa-spruce and ranged from 29 to 84%. The content of aromatic C in solution (Table 5) determined the maximum portion of DOM that could be precipitated (Fig. 2 ). A linear regression between these two factors resulted in a coefficient of determination of r2 = 0.98 (p < 0.01, n = 4). Generally, DOM solutions from Oa horizons contain larger proportions of aromatic compounds (Kalbitz et al., 2003) and more carboxylic groups (Guggenberger et al., 1994) than the Oi solutions. The larger number of possible binding sites in Oa DOM solutions resulted in larger amounts of precipitates.


View this table:
[in this window]
[in a new window]

  		Table 4. Percentage of dissolved organic carbon (DOC) precipitated at different Al/C ratios and two pH values (standard error of three replicates in parentheses).

 

View this table:
[in this window]
[in a new window]

  		Table 5. Carbon and H moieties (as a percentage of the total) of the dissolved organic matter solutions, as determined by solution 13C and 1H nuclear magnetic resonance.

 

Figure 2
View larger version (19K):
[in this window]
[in a new window]

  		Fig. 2. The maximum percentage of dissolved organic C precipitated from the dissolved organic matter solutions from the Oa and Oi horizons of a beech and a spruce forest (at pH 4.5 and Al/C ratio of 0.3, except for the Oi-spruce solution, which had an Al/C ratio of 0.1) after Al addition in relation to the content of aromatic C and aromatic H.

 
One reason for precipitation of DOM is charge neutralization by Al, resulting in a reduced solubility of Al–OM complexes. At pH 4.5, the DOM is less protonated than at pH 3.8, allowing more Al to bind and, thus, more precipitates to form. The binding of cations reduces the solubility of DOM strongly, as the electrical charge of DOM becomes more neutral (Weng et al., 2002), which leads to extensive coagulation; however, Ares and Ziechmann (1988) pointed out that the effect cannot be fully explained by charge neutralization. Aluminum may also cause formation of n-dentate complexes with organic groups, probably inducing various structural changes, such as folding or formation of macrocycles, that enhance precipitation (Ares and Ziechmann, 1988).

Coprecipitation of organic macromolecules after adsorption onto colloidal Al (hydr)oxides is a further mechanism of DOM precipitation (Boudot, 1992; Dolfing et al., 1999), which explains the increased precipitation at pH 4.5.

At an Al/C ratio of 0.05, less DOM precipitated than at larger Al/C ratios. This agrees well with the results of Jansen et al. (2002), who found that the fraction of Al bound in soluble Al–OM complexes strongly decreased with increasing Al/C ratios. The Al/C ratios >0.1 led to only minor changes in DOM precipitation. Therefore a fraction of DOM must exist that does not precipitate with Al. This has been observed before by Schwesig et al. (2003).

Changes in Aluminum/Carbon Ratio from Solution to Precipitate
Generally, the Al/C ratios of the precipitates reflected the initial Al/C ratios in the solutions. The precipitates produced at pH 3.8 had, on average, 55 to 75% smaller Al/C ratios than the Al/C ratios in solution (Fig. 3 ). Also, precipitates produced at pH 4.5 and Al/C ratios in solution of 0.1 or larger had, on average, 10 to 40% smaller Al/C ratios than the Al/C ratios in solution. In contrast, precipitation at pH 4.5 and Al/C ratios in solution smaller than 0.1 led to an increase of the Al/C ratios of up to 40% in comparison to the Al/C ratios in solution. In all cases, the largest decrease of the Al/C ratios from solution to the precipitates was found for large Al/C ratios (0.3).


Figure 3
View larger version (17K):
[in this window]
[in a new window]

  		Fig. 3. The Al/C ratios in precipitates (solid phase) of the dissolved organic matter (DOM) solutions from the Oa and Oi horizons of a beech and a spruce forest in relation to the initially adjusted Al/C ratios in the DOM solutions. Precipitation was initiated at two different pH values: pH 3.8 (left) and pH 4.5 (right). Mean values and standard error of three replicates.

 
At pH 4.5, the formation of Al(OH)3 (Lofts et al., 2001) and possibly also Al13 (Furrer et al., 1992; Hu et al., 2006) is likely, leading to subsequent coprecipitation of DOM with Al hydroxides. Because hydrous oxide Al forms have a less positive charge than Al3+, more Al was needed to achieve the same extent of charge neutralization. This could explain the larger Al/C ratios in the Al–OM precipitates at pH 4.5 than at pH 3.8. Another reason for smaller Al/C ratios at pH 3.8 may be that the solubility of DOM at pH 3.8 was less than at pH 4.5 (Kalbitz et al., 2000) because of increased protonation. Thus, less Al was needed to achieve the same extent of charge neutralization, consequently resulting in more efficient precipitation of DOM by Al at lower pH values.

We observed that precipitation of equal amounts of DOM required two to 13 times more Al at pH 4.5 than at pH 3.8, depending on whether the Al/C ratio was large or small, respectively (Fig. 4 ). Thus, larger precipitation of DOM at pH 4.5, as observed in our experiments (Table 4), can be expected only if Al is present in excess. At limiting Al concentrations, precipitation of DOM should be considerably greater at lower pH values.


Figure 4
View larger version (22K):
[in this window]
[in a new window]

  		Fig. 4. Ratio of Al needed to precipitate similar amounts of dissolved organic C at pH 4.5 and 3.8 (dissolved organic matter solutions from the Oa and Oi horizons of a beech and a spruce forest, different Al/C ratios). Ratios >1 mean that more Al was necessary at pH 4.5 to precipitate the same amount of C as at pH 3.8.

 
Changes in Organic Matter Composition by Precipitation
The specific UV absorption at 280 nm decreased in the DOM solutions by 7 to 77% after precipitation (Fig. 5 ). Because specific UV absorption and aromatic C content correlated well (Fig. 1), this indicates that aromatic compounds precipitated preferentially. The decrease in specific UV absorption by precipitation became more pronounced with increasing Al/C ratios. Precipitation at pH 4.5 reduced the specific UV absorption to a larger extent (on average, 24–54%) than at pH 3.8 (on average, 9–30%), reflecting also the larger portion of precipitates formed at pH 4.5.


Figure 5
View larger version (30K):
[in this window]
[in a new window]

  		Fig. 5. Specific ultraviolet (UV) absorption (280 nm) of initial dissolved organic matter solutions from the Oa and Oi horizons of a beech and a spruce forest (Al/C ratio 0) and after removal of precipitates by filtration. Precipitation was initiated at two pH values and four different Al/C ratios. Mean values and standard error of three replicates.

 
The calculated content of aromatic C of the precipitates increased in comparison with the original DOM solution by 13 to 270% (Fig. 6 ). In most cases, the aromatic C content was largest in precipitates formed at Al/C ratios of 0.05. It decreased with increasing solution Al/C ratios. Furthermore, precipitates formed at pH 3.8 contained more aromatic C than those formed at pH 4.5. The pH effect was most pronounced for the Oa solutions.


Figure 6
View larger version (27K):
[in this window]
[in a new window]

  		Fig. 6. Aromatic C content of initial dissolved organic matter solutions from the Oa and Oi horizons of a beech and a spruce forest (Al/C ratio 0) and of precipitates for two pH values and four Al/C ratios. Mean values and standard error of three replicates.

 
The enrichment of aromatic compounds in all precipitates (Fig. 6) supported the idea that aromatic compounds are preferentially removed from solution, as already observed for DOM adsorption on soils (Kalbitz et al., 2005). Aromatic C was most enriched in precipitates formed at small Al/C ratios. This indicated that aromatic compounds must be the first components precipitating after Al binding. We observed less aromatic C in precipitates formed at pH 4.5 than at pH 3.8. This indicated a less selective precipitation of aromatic compounds by Al hydroxides. Also, it should be considered that the DOM solutions contained limited amounts of aromatic compounds. Therefore, increased precipitation, as observed for large Al/C ratios and at higher pH, may have resulted in less aromatic C in the precipitates.

In the initial DOM solutions, a large proportion of total N was bound organically (portion of dissolved organic nitrogen [DON] to total N: Oa-beech, 37%; Oa-spruce, 57%; Oi-spruce, 69%; Oi-beech, 89%). Precipitation resulted in most cases in larger organic C/organic N ratios of the precipitates than in the original DOM solutions (Fig. 7 ). In some cases, we found an enrichment of N in precipitates, mostly for those from the Oi-beech solution. Precipitates produced at pH 3.8 had larger C/N ratios than precipitates formed at pH 4.5, except for precipitates from the Oa-spruce solution.


Figure 7
View larger version (28K):
[in this window]
[in a new window]

  		Fig. 7. Organic C/organic N ratios of initial dissolved organic matter solutions from the Oa and Oi horizons of a beech and a spruce forest (Al/C ratio 0) and of precipitates for two pH values and four Al/C ratios. Mean values and standard error of three replicates.

 
The observed depletion of organic N in the Al–OM precipitates indicated that a large proportion of DON did not precipitate with Al. Kaiser et al. (2000) and Kaiser and Zech (2000) observed a preferential adsorption of DOM with low contents of organic N to the mineral soil and attributed this to diminished adsorption of hydrophilic compounds, enriched in DON. Therefore, the depletion of organic N in the precipitates can be explained by preferential precipitation of aromatic and hydrophobic compounds, possibly derived from lignin or tannin and therefore being poor in N. Only peptides, proteins, or amino acids bound to high-molecular-weight or aromatic structures tend to precipitate (Yu et al., 2002). In precipitates of the Oi-beech solution, enrichment of organic N was observed. Thus, a larger portion of DON must have been bound in organic structures that formed insoluble complexes with Al than in the other DOM solutions. We cannot present a convincing explanation why only a major part of the organic N in the Oi-beech solution was able to form insoluble complexes with Al.

Mineralization of Aluminum–Organic Matter Precipitates
The extent of C mineralization was significantly (p < 0.01) less for precipitates from all DOM solutions, at all Al/C ratios and pH values, than for the corresponding untreated DOM solutions. During 7 wk of incubation, only 0.5 to 7.7% of the precipitated C was mineralized (Fig. 8 ), whereas the corresponding DOM solutions showed a much larger mineralization of 5% (Oa-spruce), 24% (Oa-beech), 38% (Oi-beech), and 49% (Oi-spruce), respectively.


Figure 8
View larger version (31K):
[in this window]
[in a new window]

  		Fig. 8. Dynamics of C mineralization of Al–organic matter (OM) precipitates (as a percentage of initial C) during 7 wk of incubation at pH 4.5 and 20°C. The Al–OM precipitates were produced from dissolved organic matter solutions from the Oa and Oi horizons of a beech and a spruce forest at two pH values and four Al/C ratios. Mean values and standard error of three replicates.

 
Three-factorial variance analysis indicated that the different DOM solutions, the pH at precipitate formation, and the Al/C ratio as well as all cross-interactions of the factors had a highly significant (p < 0.001) effect on the mineralizable C of the precipitates. The percentage of C mineralized was significantly (p < 0.01) smaller for precipitates from Oa solutions (spruce: 0.5–1.2%; beech: 1.0–3.3%) than for those from Oi solutions (spruce: 1.8–4.8%; beech 3.7–7.7%). We further found that the precipitates from beech DOM showed a significantly (p < 0.01) larger mineralization than those from spruce DOM. These findings reflect the mineralizability of the original DOM solutions.

Mineralization of precipitates formed at pH 4.5 was in nearly all cases increased by approximately 50 to 75% when compared with the corresponding precipitates formed at pH 3.8 (Fig. 8, 9 ). Surprisingly, we found a significant (p < 0.01) increase in mineralization with increasing Al/C ratios up to an Al/C ratio of 0.2. The influence of the Al/C ratio on mineralization of Al–OM precipitates was more pronounced at pH 4.5 than at pH 3.8. From Al/C ratios of 0.05 to 0.2, the mineralization increased on average by 33% for precipitates formed at pH 4.5 and by 19% for precipitates formed at pH 3.8.


Figure 9
View larger version (22K):
[in this window]
[in a new window]

  		Fig. 9. Mineralization of Al–organic matter precipitates (Al/C ratio 0.1) produced from dissolved organic matter solutions from the Oa and Oi horizons of a beech and a spruce forest at pH 3.8 and 4.5 (incubation at pH 4.5 and 20°C for 7 wk). Mean values and standard error of three replicates.

 
Reduced mineralization of precipitates compared with the original DOM solutions suggested C stabilization. It might have been the result of spatial rearrangements and inaccessibility of organic matter for microorganisms and enzymes due to complexation and precipitation by Al. Furthermore, preferential precipitation of inherently stable compounds can contribute to the stabilization. Aromatic and complex compounds are preferentially removed from the soil solution by sorption (Kalbitz et al., 2005) and comprise stable fractions of DOM (Kalbitz et al., 2003). An analogous effect can be expected for precipitation by Al. The different DOM used varied in chemical composition, especially in aromatic C content, and showed large differences in mineralization. Therefore, it was not surprising that the precipitates from the four solutions differed in stability also. The most likely reason for the greater stability of precipitates from Oa than from Oi solutions was the larger aromatic C content of Oa DOM solutions (Table 5). Spruce DOM solutions contained more aromatic C than the respective beech DOM solutions; the precipitates reflected this.

Toxic effects of Al on microorganisms during incubation were unlikely, as no appreciable redissolution of Al from Al–OM precipitates occurred. Incubation solutions had average Al3+ concentrations of 0.2 mg L–1 (maximum 0.5 mg L–1) at the end of the experiment. According to the literature, such concentrations are at least 10 times less than those reported to have toxic effects (De Wit et al., 2001; Illmer et al., 2003).

The observed differences in the stability of precipitates formed at the two pH values clearly indicated that the binding form is crucial for stabilization of DOM. Aluminum hydroxides preferentially form mono- or bidentate bonds, in contrast to Al3+, which can form n-dentate bonds (Ares and Ziechmann, 1988), resulting in more bonds between Al and DOM, and probably more complex spatial structures at lower pH. The fraction of Al present in solution as Al3+ is much larger at low pH (Jansen et al., 2002), which could cause stronger binding to OM in the precipitates and consequently less mineralization of precipitates. Differences in mineralization of precipitates formed at the two pH values studied were much smaller at small Al/C ratios. This could be attributed to the equally reduced formation of Al hydroxide species at small Al/C ratios at both pH values.

The reduced mineralization of precipitates produced at smaller Al/C ratios is in contrast to the findings of Boudot et al. (1989) and Boudot (1992). They observed less mineralization with increasing Al/C ratios, testing citric acid and fungal melanins, which are by far not as complex as DOM from soil. One explanation for the decreased mineralization at smaller Al/C ratios could be preferential precipitation of aromatic (Fig. 6) and high-molecular-weight structures, leading to small but stable precipitates. With increasing amounts of Al added, also less complex and aromatic compounds precipitate, resulting in precipitates more prone to mineralization. A further explanation could be that first bidentate bonds were formed, which are more stable (Simonsson, 2000).

The influence of the Al/C ratio on mineralization was more pronounced at pH 4.5 than at pH 3.8. With increasing Al concentrations, more hydrous Al forms were present in solution, allowing for formation of coprecipitates. Thus, at higher pH and larger Al/C ratios, the fraction of Al–OM precipitates formed by coprecipitation was larger, which was reflected in the larger mineralization of these precipitates. The assumption that more Al(OH)3 was present at higher pH values and larger Al/C ratios was tested by using the speciation program Visual Minteq (Gustafsson, 2005). Input parameters were the chemical characterizations of the initial DOM solutions (Table 2) and the added amount of Al at the respective pH. By including the Stockholm humic model, organic complexation was taken into account. Positive saturation indices (SI) indicate oversaturation, negative SI undersaturation of a phase. The results of the modeling (Table 6) indicated oversaturation for Al(OH)3 only at pH 4.5. Further, it can be seen that the SI increased with increasing Al/C ratios, supporting the idea that at small Al/C ratios, formation of Al hydroxide was less important, and thus, differences in mineralization of precipitates between the pH values were less at small Al/C ratios.


View this table:
[in this window]
[in a new window]

  		Table 6. Saturation indices [SI = log(IAP/KS), where IAP is the ion activity product and Ks is the saturated hydraulic conductivity] for Al(OH)3 in the dissolved organic matter solutions at 20°C and different pH and Al/C ratios, calculated with Visual Minteq (SI > 0 = oversaturation, SI < 0 = undersaturation).

 
The Al13 species might also influence the formation and stability of Al–OM precipitates. At pH values <4.2, no formation of Al13 can be expected, whereas the prevailing Al concentrations (0.17–1 mM) at pH 4.5 favor the formation of Al13. On the other hand, organic ligands can depolymerize Al13 (Masion et al., 1994). Disappearance of Al13 from DOM solutions has been observed by Yamaguchi et al. (2004) and Thomas et al. (1993), and attributed by them to the formation of Al13–organic complexes. This effect is of increasing influence with increasing Al/C ratios (Furrer et al., 1992; Thomas et al., 1993). Here, the influence of Al13 species on precipitation and stabilization could not be separated from the effect of Al hydroxides, but may be important.

Factors Governing the Stability of Aluminum–Organic Matter Precipitates
Although we found preferential precipitation of aromatic compounds, which should be the most stable DOM components, the correlation between aromatic C content in the precipitates and mineralizable C was poor (r2 = 0.21, p < 0.001, n = 94). The C/N ratio of the precipitates and mineralizable C were equally poorly correlated (r2 = 0.30, p < 0.001, n = 94). Multiple linear regression with both variables explains 76% of the variability (p < 0.001; n = 94) in mineralizable C (Fig. 10 ). A collinearity analysis confirmed (variance inflation factor = 1.1, condition index = 8.8) that the two parameters C/N ratio and aromatic C were independent. Therefore, the aromatic C content and the C/N ratio of the precipitates were the two main factors controlling the stability of Al–OM precipitates. Further, we tested whether the type of DOM, pH, and Al/C ratio influenced only these two factors or directly affected C mineralization. A three-factorial variance analysis was performed with the residuals from the multiple linear regression. The type of DOM did not significantly (p > 0.1) influence C mineralization after removing the effect on aromatic C and the C/N ratio; however, the pH (p < 0.05) and the Al/C ratio (p < 0.01) had an additional effect on C mineralization, besides their effect on aromatic C content and C/N ratio of the precipitates. This means that differences in mineralization between precipitates from the four solutions could be satisfactorily explained by changes in aromatic C content and by the C/N ratio of the precipitates; however, these two factors did not fully explain the differences caused by pH variation and by variations in the amount of Al added. This supported the idea that Al speciation and the functional groups involved, as discussed above, influenced the degree of stabilization.


Figure 10
View larger version (54K):
[in this window]
[in a new window]

  		Fig. 10. Mineralization of Al–organic matter precipitates (from dissolved organic matter solutions from the Oa and Oi horizons of a beech and a spruce forest) in dependence on the aromatic C content and the organic C/organic N ratio of the precipitates (multiple linear regression: R2 = 0.76; dissolved organic carbon [DOC] mineralized [as a percentage of initial DOC] = 9.992 – 0.091(C/N ratio) – 0.161(aromatic C) [%]).

 
Implications of Aluminum–Organic Matter Precipitation on Acidic Forest Soils and Surface Waters
The Al concentrations and pH values used in this study were typical of acidic forest soils. Therefore, the observed stability of precipitates should be important for C stabilization in the field. Soil solutions from the organic horizon that percolate downward into the mineral soil are gradually enriched in dissolved Al (Schwesig et al., 2003), similar to the artificial Al addition in our experiments, resulting in subsequent precipitation. In mineral soils, decreasing DOC concentrations have been observed with depth (Schwesig et al., 2003). Besides adsorption on Al and Fe oxides, this decrease may be explained by the formation of Al–OM precipitates. Differentiation in the field between sorptive preservation and stabilization by precipitation is difficult because the mechanisms would occur at the same time and the observable effects are rather similar. Nevertheless, precipitation of DOM by Al should result in a substantial accumulation of C and Al in the soil with time, as in podzols, because mineralization of precipitates proved to be very small.

Our results may also help to understand other recent observations, such as the reported increase in DOC concentrations of surface waters (Freeman et al., 2001; Worrall et al., 2004; Evans et al., 2005). It is assumed that soils are the main source of the additional C. A huge fraction of Al was leached from the upper soil until the mid-1980s, resulting in a smaller Al reservoir in the soil (Mulder et al., 1989; Berggren et al., 1998). Reduction of acid deposition in recent years led to increasing pH values and decreasing Al concentrations in soil solutions (Matzner et al., 2004). Therefore, Al might be the limiting factor for the formation of Al–OM precipitates in the field. We have shown that, at larger pH values, more Al was needed to precipitate the same amount of DOM (Fig. 4), especially at small Al concentrations. If we assume that larger pH values and smaller Al concentrations lead to reduced precipitate formation, this would not only result in less C stabilization in the soil, but also in larger amounts of DOM being exported to surface waters. The assumed reduction of precipitation should further lead to a change in the chemical composition of the transported DOM because precipitation removes preferentially stable compounds from the soil solution. Thus, the additional DOM in many surface waters should be mainly comprised of microbiologically stable compounds.


   CONCLUSIONS
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
REFERENCES
 
The precipitation of DOM by Al resulted in significant C stabilization against microbial decay. We observed preferential precipitation of compounds with large contents of aromatic C and little N. This was one reason for the large stability of Al–OM precipitates against microbial decay when compared with the original DOM. Precipitates formed at pH 3.8 and small dissolved Al/C ratios (0.05) had a greater stability against microbial decay than those formed at pH 4.5 and large dissolved Al/C ratios (>0.2). Thus, also the amount of Al bound as Al hydroxides in the precipitates had an influence on the stability of Al–OM precipitates against microbial decay.

The amounts of DOM precipitated by Al and the subsequent increase in stability against microbial decay were very sensitive to changes in environmental parameters, such as pH or the concentration of dissolved Al. Therefore, knowledge about the dynamics of this mechanism significantly improves our understanding of environmental processes such as podzolisation or the recently observed increase of DOC in surface waters. Thus, the stabilization of OM by the formation of Al–OM precipitates from DOM seems to be fundamental and should be incorporated into existing models and theories of C cycling.


   ACKNOWLEDGMENTS
 
This study was funded by the Deutsche Forschungsgemeinschaft (DFG) as part of the priority programme SPP1090 "Soils as source and sink of CO2". We thank the members of the Central Analytical Department of BayCEER for support, K. Kaiser, L. Haumaier, and K. Nierop for helpful comments, and L. Haumaier for recording the NMR spectra.


   NOTES
TOP
 NOTES
ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Abbreviations: DOC, dissolved organic carbon; DOM, dissolved organic matter; OM, organic matter; UV, ultraviolet.

Received for publication March 10, 2006.


   REFERENCES
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 




This article has been cited by other articles:


Home page
 Soil Sci.Home page
J.-H. Huang, K. Kalbitz, and E. Matzner
Mobility of Trimethyllead and Total Lead in the Forest Floor
Soil Sci. Soc. Am. J., May 29, 2008; 72(4): 978 - 984.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (7)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Scheel, T.
Right arrow Articles by Kalbitz, K.
Right arrow Search for Related Content