SSSAJ Grow Your Career with SSSA
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Published online 2 June 2005
Published in Soil Sci Soc Am J 69:1110-1117 (2005)
DOI: 10.2136/sssaj2004.0353
© 2005 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 (1)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by López-Ulloa, M.
Right arrow Articles by de Koning, G. H. J.
Right arrow Search for Related Content
PubMed
Right arrow Articles by López-Ulloa, M.
Right arrow Articles by de Koning, G. H. J.
GeoRef
Right arrow GeoRef Citation
Agricola
Right arrow Articles by López-Ulloa, M.
Right arrow Articles by de Koning, G. H. J.
Related Collections
Right arrow Soil Conservation
Right arrow Tropical Soil Management
Right arrow Soil Mineralogy
Right arrow Tropical Soils
Right arrow Soil Organic Matter

Soil & Water Management & Conservation

Soil Carbon Stabilization in Converted Tropical Pastures and Forests Depends on Soil Type

M. López-Ulloaa,b,*, E. Veldkampa and G. H. J. de Koninga,b

a Georg-August Univ., Institute of Soil Science and Forest Nutrition, Buesgenweg 2, 37077 Goettingen, Germany
b German Technical Cooperation (GTZ), Ecuador Casilla 17-07-8721, Quito, Ecuador

* Corresponding author (bioteam1{at}uio.telconet.net)


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
The influence of soil C stabilization mechanisms is normally not considered in studies on the effects of land use changes. Instead, observed changes are typically explained by differences in litter input. As a result, it is not well known if and how quickly newly incorporated C is stabilized in soils. Our goals were to find out how much soil C was stabilized in two different soil orders (Andisols and Inceptisols) and which are the responsible mechanisms of C stabilization. Furthermore, we looked for evidence that newly incorporated soil C was stabilized in these contrasting soil orders. We selected 25 sites in northwestern Ecuador with two paired plots per site: one plot where pasture was converted to secondary forest and one plot where forest was converted to pasture. In all the plots, soil C content, stocks, and stable isotope ({delta}13C) signal were measured in the surface soil. The {delta}13C values were used to estimate the stocks of soil C derived from forest (Cdf) and from pasture (Cdp) in all plots. We calculated correlations between these stocks and soil and environmental characteristics to identify mechanisms of soil C stabilization. Our results show that long-term stabilization in Andisols was through formation of metal–humus complexes and allophane, while in Inceptisols long-term stabilization was through sorption to clay minerals. We found evidence that recently incorporated C was not stabilized in Andisols, while in Inceptisols, poorly crystalline (hydr-) oxides seemed to have stabilized part of this soil C. We conclude that unless soil C stabilizing mechanisms are explicitly considered, we will not be able to predict the direction and magnitude of changes in soil C stocks following land use changes in the tropics.

Abbreviations: {delta}13C, stable isotope • Alo, aluminium extracted by acid oxalate • Alp, aluminium extracted by pyrophosphate • Cdf, carbon derived from forest • Cdp, carbon derived from pasture • Cp, carbon extracted by pyrophosphate • Feo, iron extracted by acid oxalate • Fep, iron extracted by pyrophosphate • LF, light fraction • masl, meters above sea level • Sio, silica extracted by acid oxalate


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
GLOBALLY, SOILS CONTAIN approximately 1500 Gt of soil organic C in the upper meter, of which about 44% is estimated to be located in soils in the tropics (Jobbagy and Jackson, 2000). At the same time, soil C residence times decrease with increasing mean annual temperature and mean annual precipitation (Amundson, 2001). As a result, the humid tropics is the area where the shortest residence times of soil C can be found. Because of the short residence times of soil C, the soil C pool responds much more rapidly to environmental changes (e.g., land use changes) in the tropics than in temperate areas. This is one of the reasons why deforestation is presently estimated to contribute about 23% to the human-induced CO2 emissions. While the majority (about 75%) of these emissions originate from the aboveground biomass, the remaining 25% is attributed to the decomposition of soil C (Detwiler, 1986; Melillo et al., 1996).

Knowledge about mechanisms of soil C stabilization has improved considerably in the past decades. Mechanisms of soil C stabilization can be divided into three groups: recalcitrance, interactions, and accessibility (Christensen, 1996; Six et al., 2002). Recalcitrance refers to molecular level characteristics of organic substances that influence their degradation by microbes and enzymes. As microbes selectively degrade the less-recalcitrant compounds, they gradually increase the average recalcitrance of the residual soil C (Oades, 1988). Interaction comprises intermolecular interactions between organic and inorganic substances that alter the rate of degradation. Examples are sorption and complexation. Clays provide the vast majority of surface area for sorption of organic groups. Complexation of ions such as Fe3+ and Al3+ by organic substrates is a clear example of an interaction that increases stability. This is supported by the observation that soil C accumulates in very large amounts in soils that lack even amorphous aluminosilicate clays such as allophane (Mizota and van Reeuwijk, 1989). Accessibility comprises the location of organic substances as it influences their access by microbes and enzymes (Sollins et al., 1996). Aggregation can decrease accessibility of substrate to microbes.

This increased knowledge of soil C stabilization, however, has barely increased our ability to predict the direction and magnitude of changes in soil C stocks following land use changes in the tropics. This is mainly because soil C stabilization mechanisms are not considered to be critical when studying land use changes. In most cases, the productivity of the different land uses is considered to be the key (e.g., Trumbore et al., 1995). As a result, only a few of the known stabilizing mechanisms (recalcitrance and protection by clay) are explicitly considered in models that describe soil C dynamics. Therefore, it is not very well known if and how quickly newly incorporated C is stabilized by one of the above mentioned processes.

In the present study, we wanted to highlight these problems by studying the soil C dynamics following land use changes in two soil groups in the pacific coastal plain of Ecuador with a contrasting genesis (Andisols and Inceptisols). We wanted to answer the following questions: How much soil C remains stable after land use change and what are the responsible mechanisms of C stabilization in Andisols and Inceptisols?

To answer these questions we compared a selection of pasture and secondary forest sites on Andisols and Inceptisols. We measured the soil C content, indicators of potential factors that stabilize soil C, and we used the {delta}13C signals to determine the size of labile and passive C pools (Balesdent et al., 1987). Although several studies have used {delta}13C in deforestation studies (e.g., Veldkamp, 1994; Neill et al., 1996), few have analyzed the effects of secondary forest regrowth on sites formerly occupied by pastures.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Study Area, Site Selection, and Sampling
The study area is located in tropical northwestern Ecuador, within the geographical co-ordinates of 80°05' W, 1°30' N and 78°40' W, 0°05' S. Elevation of the study sites varies between sea level and 1600 m above sea level (masl). Yearly annual precipitation varies from 1000 mm to just over 5000 mm.

For the present study, we used a selection of 25 sites, which were part of a separate study to quantify the soil C sequestration potential of secondary forests (de Koning et al., 2003). The volcanic soils used by de Koning et al. (2003) were identical to the Andisols sites (n = 12) used in this study. The Andisols are relatively young soils developed on volcanic ashes with a mineralogy characterized by the presence of allophane. Generally, they are acid or slightly acid, have high water retention, low bulk density, sandy or loamy texture, and a base saturation under 35 cmolc/100 g. Andisol sites receive between 3000 and 5000 mm rain per year. As the sedimentary soils used by de Koning et al. (2003) had quite a diverse soil genesis (and therefore probably different mechanisms of soil C stabilization), we excluded soils which were heavily weathered (Oxisols, Ultisols) and soils developed in recent alluvial and marine deposits (Entisols). The remaining sites (n = 13) can be characterized as being dominantly Inceptisols developed on sedimentary rocks (mainly siltstone) of Tertiary origin. These soils, which we call Inceptisols in this study, are more developed than Andisols, with a loamy to clayey texture, and they receive <2500 mm rainfall per year.

At each site, a paired pasture plot and secondary forest or forest plantation plot were selected, with the plot sizes at least 1 ha. Pastures and forests had different ages to allow for reconstruction of soil C changes across time. All secondary forests or plantation forests were established after abandonment or conversion of former pastures. These former pastures were established after cutting and or burning of the original forest. Also, all pastures in our sample were established after cutting and or burning of the original forest. Moreover, we selected two natural forests (one on an Inceptisol and one on an Andisol) used as reference forests. Before soil sampling, litter was collected of each plot for 13C analysis.

On each plot, eight sampling points were selected within a 50- by 50-m area using a stratified random sampling design (de Koning et al., 2003). At each sampling point, soil samples were taken with an auger at two fixed soil depths: 0 to 0.25 m and 0.25 to 0.5 m. Soil samples were air dried and then passed through a 2-mm sieve before laboratory analyses. At four sampling points, bulk density samples were taken and pH was measured for both depth intervals. Land use history, vegetation age, and actual land management was obtained through interviews with landowners. In the forest plots, tree biomass was estimated by means of nondestructive inventories (López et al., 2002). Biomass was not measured in pasture plots. Annual precipitation was estimated using a digital interpolated precipitation map based on 20 weather stations in the study area.

Laboratory Analysis of Soil Samples
Part of the individual samples was used to make composite samples for each layer per plot consisting of mixed material of the corresponding 8 samples. Carbon, N, and C isotopes (13C) were analyzed for all individual samples, and results are reported here as the mean value from eight samples, while analysis of the other soil characteristics was done for the composite samples only. Carbon and N content was determined by means of dry combustion using an automated C & N analyzer (Heraeus vario EL, Hanau, Germany). The light fraction (LF) organic matter was isolated by suspending the soil in a dense liquid and extracting the LF from the surface (Gregorich and Ellert, 1993). Texture of composite samples was determined with the pipette method, distinguishing the three fractions: clay (particle size < 0.002 mm), loam (particle size between 0.002 and 0.063 mm), and sand (particle size between 0.063 and 2 mm). Mineralogy of composite samples was examined through acid-oxalate extractions of aluminium (Alo), iron (Feo), and silica (Sio), and with pyrophosphate extractions of aluminium (Alp), iron (Fep), and carbon (Cp); see for details de Koning et al. (2003). Oxalate extractions of Al, Fe, and Si indicate all active components of Al, Fe, and Si, dissolving noncrystalline minerals such as allophane, imogolite, amorphous and poorly crystalline oxides like ferrihydrite, as well as organomineral Al– and Fe–humus complexes (Mizota and Van Reeuwijk, 1989). Pyrophosphate extractions of Al, Fe, and C indicate all Al, Fe, and C present in organomineral humus complexes (Shoji et al., 1993). The ratio of Alp/Alo is indicative of the contents of allophane vs. Al–humus complexes in volcanic soils. The Alp/Alo values near 0 suggest that allophane is dominant while Alp/Alo values near 1 indicate the predominance of Al–humus complexes (Mizota and van Reeuwijk, 1989). Similarly, Alo minus Alp is an indication for noncrystalline minerals, with high values indicating high contents of these components. The Feo minus Fep is an indication of the content of ferrihydrite.

Carbon isotope ratios were determined for soil C, LF, and litter samples. Tropical grasses are C4-type vegetation, while forest trees are predominantly C3-type vegetation. The mechanism of photosynthetic CO2 uptake of C3 plants is more discriminating against 13C than the CO2 uptake by C4 plants, resulting in a lower 13C/12C ratio in the C3 plants (Balesdent et al., 1988). In case of a conversion from C3-vegetation to C4-vegetation or vice versa, the C isotope ratio of soil organic matter can be used to determine which fraction of the soil organic matter originates from either vegetation. The C isotope ratios are expressed as {delta}13C{per thousand} (Balesdent et al., 1988). The amounts of soil C derived from forest and pastures in the pasture plots (representing a conversion from C3-forest to C4-pasture) were calculated using a simple mixing equation (Balesdent and Mariotti, 1996):

[1]

[2]
where {delta}13Cps = stable carbon isotope value of sample from pasture soil; {delta}13Cfs = stable carbon isotope value of sample from forest reference soil; {delta}13Cpl = stable carbon isotope value of pasture residues; and {delta}13Cfl = stable carbon isotope value of forest residues. The advantage of this mixing equation compared with others is that it can be applied even if the isotope enrichments during C decay are high (Balesdent and Mariotti, 1996). The amounts of soil C derived from forest and pastures in the secondary forest plots (representing a conversion from C4-pasture to C3-forest) were calculated in a similar way.

For the determination of the 13C value, soil samples were ground to powder using a ball mill. The 13C values were measured with an elemental C&N analyzer coupled with an isotope ratio mass spectrometer. Forest and pasture litter samples were dried at 65°C to a constant weight, finely ground for homogenization, and analyzed in the same way. The average soil {delta}13C and bulk density of each plot for the 0- to 0.25-m layer and {delta}13C for pasture vegetation and forest vegetation were calculated and used for the calculation of fractions of Cdp and Cdf in the pasture and forest plots. Average {delta}13C values measured for litter were –14.18{per thousand} for pastures and –30.43{per thousand} for forests.

Statistical Analyses
We used Spearman rank correlation coefficients to explore the relationships among the contents of Cdp, Cdf, and the soil and environmental variables. Analyses were done separately for Andisols and for Incepsols, as both soil groups have a different soil genesis.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 CONCLUSIONS
 REFERENCES
 
{delta}13C Signal Following Land Use Changes
Average soil {delta}13C values in each plot show that, in most sites, soil C originated from a mixture of forest and pasture C (Table 1). Compared with the pasture soils, {delta}13C values in the 0- to 0.25-m layer of the forest soils were, on average, 2.9{per thousand} lighter for Inceptisols and, on average, 2.0{per thousand} lighter for Andisols. On Inceptisols, forests contained, on average, 12.5 Mg C ha–1 more soil C than pastures. On Andisols, forests contained 8.0 Mg C ha–1 more soil C than pastures (Table 1). These results can be explained by the pasture age that on average is higher on Inceptisols (29 yr) than on Andisols (17 yr). The {delta}13C of the 0- to 0.25-m soil layer of each land use type was strongly related to age (Fig. 1a, 1b) , with highly significant correlation coefficients (P ≤ 0.01) for both pastures (r = 0.85; P = 0.01) and forests (r = –0.83; P = 0.01) on Andisols and for pastures (r = 0.58; P = 0.01) on Inceptisols, but not for forests on Inceptisols (r = 0.17; P = 0.44). For both soil types, {delta}13C values decreased with age in forest soils (Fig. 1a). However, in ≥15-yr-old secondary forests on Andisols, values were more negative than in ≥15-yr-old secondary forests on Inceptisols. In pasture soils, {delta}13C values increased with age, reflecting a gradual replacement of C3 carbon with C4 carbon (Fig. 1b). In pasture soils, no clear differences between the two soil groups could be determined. Similar patterns with time could be observed in the {delta}13C signal of the LF (Fig. 1c and 1d), but, in general, the {delta}13C values of the LF in forest soils and pasture soils were lower than the values of total soil C. Furthermore, the variation in {delta}13C signal of the LF was larger than that of the soil C.


View this table:
[in this window]
[in a new window]
  		Table 1. Mean of soil {delta}13C values and soil C stocks in forest and pasture soils, at a soil depth of 0 to 0.25 m in each soil type.

 


View larger version (22K):
[in this window]
[in a new window]
  		Fig. 1. (a) Plot of {delta}13C values of soil C in secondary forest soils against forest age. (b) Plot of {delta}13C values in pasture soils against pasture age. (c) Plot of {delta}13C values of light fraction in forest soils against forest age. (d) Plot of {delta}13C values of light fraction in pasture soils against pasture age (circles corresponds to Inceptisols and squares to Andisols).

 
Soil Carbon Dynamics Following Land Use Changes
We plotted the calculated soil C stocks (Cdp, Cdf, and total C) against age of forest or pasture and used logarithmic curves to fit through the observed C stocks (Fig. 2) . In Andisols, about 50 Mg C ha–1 of the original forest remained in the soil after 30 yr of pasture (Fig. 2a). In the same period, Cdp increased to about 20 Mg C ha–1. Although we cannot reconstruct how much soil C was present in the original forest, total soil C stock clearly decreased with time. Part of the observed variation in total soil C was probably caused by the variation of the original amount of soil C. In Inceptisols (Fig. 2b), clearly less Cdf remained after 30 yr of pasture ({approx}35 Mg C ha–1). The increase in Cdp stocks was, however, more or less the same as in the Andisols ({approx}20 Mg C ha–1 in 30 yr; Fig. 2b). In the Inceptisols, total soil C stocks had a slight tendency to increase with time, but this increase was not significant and had a low r2 (0.02), which was mainly caused by the slope that was very close to 0.



View larger version (36K):
[in this window]
[in a new window]
  		Fig. 2. (a) Stocks of Soil C derived from pasture, forest and total C on pastures Andisols soil in the 0- to 0.25-m layer; (b) stocks of soil C derived from pasture, forest, and total C on pastures Inceptisols in 0-0.25m layer, (c) stocks of soil C derived from pasture, forest, and total C on secondary forest Andisols in the 0- to 0.25-m layer; (d) stocks of soil C derived from pasture, forest, and total C on secondary forest Inceptisols in the 0- to 0.25-m layer.

 
In the secondary forest soils, changes in total soil C stocks, Cdf, and Cdp were opposite to the changes observed in the pastures (Fig. 2c, 2d). Curves start at the forest age of 7 yr, as no younger forests were included in the sample. In the Andisols, the total soil C stock significantly increases with time (r2 = 0.42). This is caused by a strong increase in Cdf (which now consists of the C remainder from the original forest and the C from the new, secondary forest that was established after pasture abandonment). At the same time, we observed a strong decrease in the amount of Cdp in Andisols. This resulted in a virtually complete disappearance of Cdp on secondary forest sites older than 15 yr (Fig. 2c). In the forests on Inceptisols, we observed no significant trend of total soil C or Cdf, although both had a weak tendency to increase with time. This lack of trend was caused by the large variation in observations combined with the slope that was close to zero. In contrast to Andisols, there was no decrease in Cdp with time (Fig. 2d). Most secondary forest sites ≥ 15 yr still had a Cdp stock of 10 Mg or higher.

Correlations between Pasture- and Forest-Derived Carbon and Environmental Factors
To explain the direction and magnitude of the changes in Cdp and Cdf in pasture and secondary forest, we calculated the Spearman correlation coefficients of Cdp and Cdf (Mg C ha–1) with the following soil characteristics and environmental variables: clay (%), sand (%), silt (%), clay + silt (%), Alo (%), Feo (%), Sio (%), Alp (%), Fep (%), Alp/Alo, Alo – Alp (%), elevation (masl), slope (degrees), precipitation (mm yr–1), soil C/N ratio, and aboveground biomass (Mg C ha–1). Correlations were calculated for Andisols and Inceptisols independently (Table 2). The Cdp in Andisol pastures did not have significant correlations with any of the environmental variables or soil characteristics, although the correlations with pasture age and elevation were close to significant. As expected, Cdf in Andisol pastures significantly decreased with pasture age, but Cdf also correlated with Alo, Sio, Alp, the Alp/Alo ratio, and the difference Alo – Alp. In Inceptisol pastures, Cdp had a positive correlation with pasture age and a negative correlation with slope. The Cdf did not have the correlations that were shown in the Andisols; instead, the only positive correlation was with clay + silt content.


View this table:
[in this window]
[in a new window]
  		Table 2. Spearman rank correlation between carbon from pasture or from forest (Cdp or Cdf, respectively) in Mg C ha–1 soil, with soil characteristics and environmental variables [clay (%), sand (%), silt (%), (clay + silt) %, Alo (%), Feo (%), Sio (%), Alp (%), Fep (%), Cp (%), Alp/Alo ratio, Alo – Alp (%), elevation (m above sea level), slope (degrees), precipitation (mm yr–1), aboveground biomass (Mg C ha–1), soil C/N ratio, pasture and forest age (yr)] of pasture and forest soils at a soil depth of 0 to 0.25 m.

 
In secondary forest soils, variables that correlate with Cdp in Andisols were forest age, soil C/N ratio, and clay content, while Cdf was positively correlated with elevation, forest age, soil C/N ratio, and aboveground biomass (Table 2). In Inceptisol forest soils, Cdp was positively correlated with Alo and negatively with slope, while Cdf was positively correlated with clay, clay + silt, Cp, and soil C/N ratio.


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
CONCLUSIONS
 REFERENCES
 
Use of Space-for-Time Substitution to Reconstruct Carbon Dynamics in Chronosequences
In the study by de Koning et al. (2003), from which we selected 25 sites, we showed that pasture and forest plots within sites were comparable and that differences measured were caused by land use and not by inherent site variability. However, the sedimentary soils studied by de Koning et al. (2003) included a large variation of soils over all sites. By excluding sites which had a clearly different soil genesis (see Materials and Methods), we are confident that within the two groups of soils studied, soil-forming factors were relatively similar, following the recommendation by Powers and Veldkamp (2005). Furthermore, the number of sites within a soil group was relatively large (n = 12 for Andisols and n = 13 for Inceptisols), reducing the chance that trends observed in a chronosequence were based on outliers. We therefore think that within the two studied soil groups, sites with different times since forest or pasture conversion can be compared in a chronosequence. Information from this comparison should, however, be interpreted with care, as part of the variation will be caused by variability which was not caused by the land use considered.

Use of {delta}13C Signals as a Tracer in Forest-to-Pasture to Secondary Forest Conversions
In studies of soil C dynamics following land use changes where C3 vegetation (like forest) is replaced by C4 vegetation (like tropical grasses) and vice versa, {delta}13C can be used as a tracer to track back the origin of the soil C (e.g., Veldkamp, 1994; Neill et al., 1996; Bashkin and Binkley, 1998; Rhoades et al., 2000). However, several assumptions are made using {delta}13C, which may obstruct interpretation. Tropical pastures typically contain some C3 weeds, which contribution can be considerable (Trumbore et al., 1995). In a recent study in Costa Rica, Powers and Veldkamp (2005)(unpublished data) concluded that this uncertain contribution of C3 prevented the use of {delta}13C as a precise tracer. However, they also showed that the assumption of constant (or pure) C4 input in the pastures led to more consistent results than using measured {delta}13C signals of randomly harvested pasture plants at the time of sampling. We therefore used the assumption that input of pastures was purely C4, while acknowledging that this led to an underestimation of pasture-derived C in the pasture and a corresponding overestimation of forest-derived C in the pastures.

In the present study, we selected only one reference forest for each soil type, and we did not select reference forest plots for each site, as these primary forests were not available. This means that we do not know how much soil C was present in the original forest (before clearing for pasture), nor do we know how much soil C was present in the pastures before they were abandoned and a secondary forest grew back. This limits our possibilities to interpret the information from the {delta}13C analyses. In the pastures we cannot quantify how much C3 carbon has disappeared since forest clearing, but we can calculate which part of the C is still from the original forest in each soil type and which mechanisms of soil C stabilization are responsible. We can also see whether there are systematic changes with time of the total soil C stocks. In the secondary forests, we cannot distinguish between C3 carbon from the original forest and C3 carbon from the secondary forest. However, we can interpret C4 dynamics as this soil C can only originate from the pasture. Also, here we can interpret systematic changes with time of the total soil C stocks.

Mechanisms of Soil Carbon Stabilization in Pastures and Forests
What is stabilizing the soil C derived from forests in the Andisol pastures? Correlations of Cdf with Alo, Sio, Alp, the Alp/Alo ratio, and the difference Alo – Alp strongly suggest that in these soils, C is stabilized principally by Al–humus complexes (r = 0.78, P = 0.01, Pearson correlation) and allophane (Shoji et al., 1993). For Andisols, this has been demonstrated before in field studies (e.g., Veldkamp, 1994; Torn et al., 1997; Powers, 2001; Percival et al., 2000; Powers and Schlesinger, 2002), where correlation analysis was used to show that complexation is an important mechanism of soil C stabilization in soils derived from volcanic ash. Also, in laboratory experiments, Boudot (1992) showed that Al–humus complexes were more effective in reducing biodegradation of organic C than allophane. In Inceptisol pastures, forest-derived C was correlated with clay and silt, suggesting that sorption of organic matter to clay minerals was the stabilizing mechanism. This soil stabilization mechanism has been also shown by Hassink (1997). Our results showed that the interaction mechanism of soil C stabilization plays an important role in soil C permanence, but differ per soil type. They also show that in a period of 30 yr, stabilization of soil C by Al–humus complexes in Andisols is larger than stabilization by clay minerals in Inceptisols (Fig. 2a, 2b).

As C derived from forest in the secondary forest soils is a mixture of old soil C (from the original forest) and recent soil C (from the secondary forest), correlations can be expected both with mechanisms that stabilize soil C and with indicators of forest productivity. The Cdf in Andisols was significantly correlated with biomass, soil C/N ratio, and elevation, which are all related to forest productivity. In contrast, in Inceptisols Cdf was correlated with clay content (which refers to the stabilization mechanism) and soil C/N ratio (which is related to forest productivity). However, because of the mixture of recent and stabilized Cdf in secondary forests, these correlations should be interpreted with care.

In secondary forests, correlations of Cdp with soil characteristics are more indicative, as we know that this C was incorporated relatively recently. Correlations can indicate whether this soil C is stabilized as well. In Andisols, Cdp correlated with clay content, but as the determination of clay in these soils is problematic (because of problems to disperse the clay, Mizota and van Reeuwijk, 1989), we suspect that this is an artifact. In Inceptisols under secondary forest, Cdp was positively correlated with Alo, and negatively correlated with slope. The negative correlation with slope is probably a reflection of erosion or other slope processes. The correlation with Alo (accompanied by a near-significant correlation with Sio), may indicate the stabilizing process for Cdp in these soils. Acid oxalate extracts poorly crystalline and amorphous oxides. This suggests that these components may play a role in the stabilization of recently incorporated C.

With our approach we could show that interactions played a role in stabilization of soil C both in Andisols and in Inceptisols. However, as we did not measure any indicators of accessibility (e.g., aggregation stability or size) nor of recalcitrance (e.g., humus chemistry), it is not surprising that we did not find any evidence that these mechanisms were important in these soils.

Soil Carbon Stocks and Dynamics in a Forest-to-Pasture and Pasture-to-Forest Sequence
Comparison of the deforestation and reforestation sequences of the two soil groups revealed quite contrasting soil C dynamics following land use changes. In Andisols, the total soil C stock sharply decreased following forest clearing and pasture establishment, while in Inceptisols, this was not the case. Decreases in total soil organic C following pasture establishment in Andisols have been found in other studies (e.g., Veldkamp, 1994; Rhoades et al., 2000), but increases also have been reported (e.g., Osher et al., 2003). As in our study, the amount of soil C derived from pastures did not differ between the two soils (in both cases, Cdp after 30 yr of pasture was about 20 Mg ha–1), the difference in total soil C dynamics is caused by the larger amount of soil C which decomposed in the Andisols compared with the Inceptisols (Fig. 2). However, this larger stock of decomposable soil C in Andisols is not the only reason why the total soil C stocks are higher. After 30 yr of pasture, the amount of soil C derived from the original forest is still much higher in the Andisols (about 50 Mg ha–1) than in the Inceptisols (about 35 Mg ha–1). The higher soil C stock in the Andisols compared with the Inceptisols was caused by a combination of a larger stable soil C pool (stabilized by complexation with a turnover time of >30 yr) and a larger decomposable soil C pool (turnover time <30 yr). The larger stable soil C pool in Andisols, combined with stronger correlations with the stabilizing mechanism (Table 2), suggests that the mechanism of C stabilization in Andisols was more effective than the mechanism of C stabilization in Inceptisols.

In the secondary forest soils, the opposite trends could be observed compared with the pastures. The increase in total soil C of Andisols can only be explained by the strong increase in forest derived C and, according to the correlations, this may be explained by a high leaf litter production (Brown and Lugo, 1990) and/or high root biomass production (Berish and Ewel, 1988). In the Andisols, pasture-derived soil C practically disappeared after >15 yr of secondary forest. This strongly suggests that the soil C that originated from the pasture was not stabilized in the Andisols. However, this result should be interpreted with care. While it was possible to get information about the age of the present land use, most owners could not provide information on how long pastures had existed before they were converted into secondary forests. We can therefore not exclude that, in some of the sites, the contribution of Cdp was already low at the time when the pasture was converted into a secondary forest. In the Inceptisols, total soil C and Cdf did not significantly increase (although there was a tendency to increase, and the average soil C stocks under secondary forest were higher than the soil C stocks under pastures). The stronger increase in Andisols compared with Inceptisols may be caused by higher biomass production in Andisols which are located in an area with higher precipitation. In contrast to the Andisols, considerable amounts of Cdp were observed in Inceptisols under secondary forest of ≥15 yr. This suggests that soil C that was incorporated in the pastures was more stable in Inceptisols than in Andisols.


   CONCLUSIONS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
REFERENCES
 
In this study we have shown that soil C stocks in Andisols and Inceptisols react differently on land use changes. This difference was probably caused by a combination of different factors: (i) Forest productivity, where litter input in the secondary forests on Andisols was probably higher than secondary forests on Inceptisols (López et al., 2002). (ii) In Andisols, more soil C was stabilized than in Inceptisols, and the processes of stabilization were different. We found indications that relatively recently incorporated soil C was stabilized in Inceptisols through interactions with poorly crystalline Al (hydr-) oxides, while in Andisols this was not the case. In neither Inceptisols nor Andisols did we find indications that the processes that are responsible for the long-term (>30 yr) stabilization of soil C played a role in the stabilization of recently incorporated soil C. The reason may be that the soil C turnover in these stabilized pools is so slow that in the few decades that these land use changes had taken place, no significant changes have occurred.


   ACKNOWLEDGMENTS
 
This study was financed by the Tropical Ecology Support Program (TOEB) of the German Technical Cooperation (GTZ) and the German Federal Ministry of Education and Research through the BIOTEAM program. We thank the TOEB staff, especially Elisabeth Mausolf, for their continuous support. We acknowledge the many landowners who allowed us to collect field data and we thank Wolfgang Lutz of GTZ-Ecuador for his institutional support.

Received for publication November 9, 2004.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 




This article has been cited by other articles:


Home page
 Soil Sci.Home page
E. A. Paul, S. J. Morris, R. T. Conant, and A. F. Plante
Does the Acid Hydrolysis-Incubation Method Measure Meaningful Soil Organic Carbon Pools?
Soil Sci. Soc. Am. J., April 19, 2006; 70(3): 1023 - 1035.
[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 (1)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by López-Ulloa, M.
Right arrow Articles by de Koning, G. H. J.
Right arrow Search for Related Content
PubMed
Right arrow Articles by López-Ulloa, M.
Right arrow Articles by de Koning, G. H. J.
GeoRef
Right arrow GeoRef Citation
Agricola
Right arrow Articles by López-Ulloa, M.
Right arrow Articles by de Koning, G. H. J.
Related Collections
Right arrow Soil Conservation
Right arrow Tropical Soil Management
Right arrow Soil Mineralogy
Right arrow Tropical Soils
Right arrow Soil Organic Matter

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
The SCI Journals Agronomy Journal Crop Science
Journal of Natural Resources
and Life Sciences Education		 Vadose Zone Journal
Journal of Plant Registrations Journal of
Environmental Quality		 The Plant Genome