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


     

Published online 2 February 2006
Published in Soil Sci Soc Am J 70:359-366 (2006)
DOI: 10.2136/sssaj2005.0061
© 2006 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 ISI Web of Science (2)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Corre, M. D.
Right arrow Articles by Veldkamp, E.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Corre, M. D.
Right arrow Articles by Veldkamp, E.
Agricola
Right arrow Articles by Corre, M. D.
Right arrow Articles by Veldkamp, E.
Related Collections
Right arrow Agricultural Systems
Right arrow Agroforestry
Right arrow Soil Biochemistry
Right arrow Soil Biology

Soil Biology and Biochemistry

Soil Nitrogen Cycling following Montane Forest Conversion in Central Sulawesi, Indonesia

Marife D. Corre*, Georg Dechert and Edzo Veldkamp

Institute of Soil Science and Forest Nutrition, Univ. of Goettingen, Buesgenweg 2, Goettingen 37077, Germany

* Corresponding author (mcorre{at}gwdg.de)


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
The lower montane forest zone of Indonesia is undergoing rapid conversion of indigenous forests to agriculture. In this tropical region, however, the effects of forest conversion on soil N processes have not been investigated. Corn (Zea mays L.) and cacao (Theobroma cacao L.)–coffee (Coffea canephora Pierre ex Froehner) agroforestry are the main land use types in cleared lower montane forests in Central Sulawesi, Indonesia. Our main objective was to compare the soil N dynamics under agroforest systems and corn cultivation with indigenous forest. We measured the gross rates of N transformation processes using 15N pool dilution. The agroforest systems and indigenous forests had higher gross N mineralization rates and faster turnover rates of NH4+ and microbial N pools than the long-term cultivated corn sites. Faster soil N turnover rates in agroforest systems suggest a more dynamic soil N cycling. Leguminous shade trees, which are important components of these agroforest systems, may have influenced the fast microbial N cycling through release of N-rich root exudates and plant residues. Our results show that compared with corn, agroforestry is a better option in terms of sustainability in the N-supplying capacity of the soil. In addition, we measured higher 15NH4+ recoveries than 15NO3– recoveries after 15 min of 15N addition in all our sites. Our measured rates of gross nitrification were very low to negligible, due to rapid disappearance of added 15NO3–. Such fast reaction of NO3– warrants further investigation, especially in tropical areas where 15N studies are very few.

Abbreviations: ISSFN, Institute of Soil Science and Forest Nutrition • MBN, microbial biomass nitrogen • MRT, mean residence time


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
ASIA'S TROPICAL FORESTS comprise 20% of the world's tropical forest, and Indonesia has 43% of the tropical rain forest in South East Asia (FAO, 2001). However, the lower montane forest zone of Indonesia is undergoing rapid conversion of indigenous forests to slash-and-burn agriculture, and annual deforestation rates have increased dramatically in the last decade (Van Rheenen et al., 2004). Slash-and-burn agriculture depends largely on the nutrients stored in the soil and aboveground biomass, which are released after clearing and burning of vegetation. Because of the export of nutrients through harvest and leaching and the absence of fertilizer input, this agricultural system may become easily depleted in nutrients after years of continuous cultivation (Nye and Greenland, 1965; Palm et al., 1996). The consequences of forest conversion on soil N processes are poorly understood for tropical montane forests, particularly in Southeast Asia.

From studies conducted in Latin America, tropical montane forests support lower rates of net N mineralization and net nitrification than tropical lowland forests (Vitousek and Matson, 1988; Marrs et al., 1988; Rhoades and Coleman, 1999; Cavelier et al., 2000), implying that N is likely in short supply in many montane forests. Conversion of montane forests to pasture further decreases net rates of soil N cycling (Rhoades and Coleman, 1999). On the other hand, conversion of lowland forests to pastures may exhibit elevated net rates of soil N cycling in the immediate period following clearing (Matson et al., 1987; Montagnini and Buschbacher, 1989; Neill et al., 1999), which could be responsible for high N losses (Veldkamp et al., 1999). Only pastures established ≥3 yr after forest clearing show lower net rates of soil N cycling than the reference forests (Reiners et al., 1994; Neill et al., 1997).

While net mineralization and net nitrification provide an index of plant available N, they do not reflect the total amount of N cycling between organic and mineral N pools. For example, rates of gross N mineralization and gross nitrifications were similar in lowland forest and in pastures up to 10 yr old and only declined in pastures older than 10 yr (Neill et al., 1999). These findings indicate that soil N turnover eventually slows in old pastures but not as quickly as net mineralization and net nitrification alone would suggest.

Our main objective was to compare the soil N dynamics under major land use systems with indigenous forest as the reference. This would lead us to understand the processes determining the N-supplying capacity of the soil. We measured gross rates of N transformations in lower montane forests and in the two dominant land use types, agroforest and corn, in cleared areas in Central Sulawesi, Indonesia. In these unfertilized systems, external sources of N mainly come from atmospheric deposition (2.7 kg N ha–1 yr–1 from bulk precipitation; Dechert et al., 2006), except for the agroforest systems that have additional N from leguminous shade trees. A considerable portion of available N for plant and microbial use must be provided by the microbially mediated N processes in the soil. Our study is the first to our knowledge to investigate how gross rates of soil N cycling change as lower montane forest is converted to agricultural use in South East Asia. Our results provide a comparative estimate of the N-supplying capacity of the soil that could serve as basis to assess the sustainability of these land use systems.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Site Description
Our research area was around the margin of the lower montane forest in the Lore Lindu National Park in Central Sulawesi, Indonesia. The mountains surrounding the national park are still widely forested, but the forest margins are currently undergoing rapid conversion into agricultural lands. Forests are cleared by manual cutting of undergrowth and trees, leaving them to dry for some time and, after removing valuable woods, the rest of the biomass is burned on site. These cleared lands are mainly utilized for corn cultivation, agroforestry, and vegetable production (including legumes and starch-tuber crops). Corn is cultivated mainly as monoculture in continuous cropping systems without fallow periods. Planting, weeding, and harvesting are all done manually. Most farmers achieve 2 to 3 harvests per year. The agroforestry systems are mostly mixed stands of cacao and coffee with leguminous shade trees (Gliricidia sepium Jacq., Erythrina fusca Lour., and Erythrina subumbrans Merr.). After agroforestry establishment, the soil undergoes only minimal cultivation disturbance, mainly through manual weeding; the shade trees and crop trees are pruned regularly.

Our measurements were conducted on the two main agricultural systems, corn and agroforestry, and on forest sites as the reference system. We selected these land use systems in three locations, which have different soils but under similar climatic conditions (Table 1). The area has average annual (1984–1999) rainfall of 1590 mm yr–1 and average (2002) daily air temperature of 21°C. In Locations 1 and 2 all three land use systems were sampled, while in Location 3 agroforestry was absent and only corn and forest sites were sampled. The three land use systems in each location were selected at <50 m apart such that they were located on the same soil type. Forest sites in all three locations had been minimally disturbed, with manual logging of some individual trees, but the undergrowth was largely intact. Plant survey in these forests revealed low density and low species number of N-fixing legume species (Kessler et al., 2005). All agroforest and corn sites were established from a previously cleared forest, except the agroforest site in Location 2 that was previously under 2 to 3 yr of corn cultivation following forest clearing. It must be mentioned that the percentage clay in Location 3 was higher in the forest site than in the corn site (Table 2), which might reflect the selective removal of clay as a result of erosion on the cultivated area.


View this table:
[in this window]
[in a new window]
  		Table 1. Site description.

 

View this table:
[in this window]
[in a new window]
  		Table 2. Soil characteristics (means ± 1 SE; n = 4) measured at 0- to 5-cm depth.

 
Sampling Design
Soil sampling was conducted in May 2002. Before taking samples from the forest and agroforestry sites, the fresh to slightly decomposed litter was removed. This was done to have comparable soil sampling depth with the corn sites where litter layer was absent. The litter layer was about 2 to 3 cm in the agroforest and forest sites. Soil sampling at these sites was done between trees. Along a 60-m long transect across each site, four sampling points were selected at a distance of 10 m apart. At each sampling point, four intact cores of topsoil (0–5 cm) were taken with stainless steel cores of 8-cm diam. The four core samples at each sampling point were taken within a 0.3 m x 0.3 m area. Additional soil samples were taken at each sampling point for analysis of initial mineral N content and other supporting soil parameters. The samples were transported immediately to the laboratory where they were allowed to acclimatize for 48 h in the dark at 24°C. The average soil temperature in May was 24°C, and hence measurement of soil N cycling rates was conducted in similar temperature.

Gross Rates of Soil Nitrogen Cycling
We used the 15N pool dilution technique to estimate gross rates of soil N cycling (Davidson et al., 1991; Hart et al., 1994b). For each sampling point, two cores were injected with (15NH4)2SO4 solution for measurements of gross rates of N mineralization and NH4+ consumption, and another two cores with K15NO3 solution for gross rates of nitrification and NO3– consumption measurements. Each core received five 1-mL injections of the solutions containing 30 µg N mL–1 with 98% 15N enrichment. This was equivalent to a rate of 0.7 to 1.4 µg 15N g–1. Injection was done using a side-port needle in five injection points (1 mL each) per core, leaving columns of the solution in the core.

Two cores (one injected with 15NH4 and one injected with 15NO3) were immediately extracted with 0.5 M K2SO4 by mixing the soil well in a plastic bag and taking a subsample for mineral N extraction (approx. 5:1 ratio of solution to dry mass soil). The time elapsed between injection and extraction was 15 min (T0 cores). Not all of the added 15NH4+ and 15NO3 were recovered in the labeled pools at T0. The T0 cores were then used to correct for the reactions that occur immediately after injection of 15NH4+ and 15NO3–, as recommended by Davidson et al. (1991) and Hart et al. (1994b). The other pair of the labeled cores was incubated for 48 h in the dark at 24°C (T1 cores), and then extracted with 0.5 M K2SO4. In each core sample, gravimetric moisture content was also measured to express gross rates on a soil dry mass basis.

Extraction was done by shaking the samples for 1 h and filtering the extracts through prewashed (0.5 M K2SO4) filter papers. Extracts were immediately frozen until analysis. For analysis of NH4+ and NO3– contents, frozen extracts were packed in ice and were flown to Germany, where analysis was conducted in the Institute of Soil Science and Forest Nutrition (ISSFN) using continuous flow injection colorimetry (Cenco/Skalar Instruments, Breda, Netherlands). Ammonium was determined using the Berthelot reaction method (Skalar Method 155–000, 1992) and NO3– was measured using the Copper-Cadmium reduction method (Skalar Method 461–000, 1992). Detection limits were 150 µg L–1 for both NO3––N and NH4+–N.

For 15N analysis from the extracts, the same diffusion procedure described in detail by Corre et al. (2003) and Corre and Lamersdorf (2004) was followed; this was adapted from Stark and Hart (1996). Diffusion was performed in Indonesia and only the diffused samples ready for 15N analysis were brought to ISSFN, Germany. For diffusion, 50 mL of extract was placed in a 150-mL glass bottle. We used 5-cm wide Teflon tape to encase the acidified filter discs (2 discs of 7 mm diam. cut from glass fiber filter paper and acidified with 20 µL of 2.5 M KHSO4 solution). This wide Teflon tape also covered the mouth of the diffusion bottle and this helped fastening the lid tightly. For the 15NH4+–labeled samples, diffusion was performed only for the NH4+ pool. MgO was added to the extract to convert NH4+ to NH3, and the acid trap was immediately placed on the mouth of the bottle and the lid fastened. Diffusion proceeded for 6 d. For the 15NO3––labeled samples, diffusion was performed only for the NO3– pool. The bottles were first left open after adding MgO for 6 d to get rid of NH4+, followed by 6 d of diffusion after adding Devarda's alloy to convert NO3– to NH4+ and eventually to NH3. Nitrogen-15 was analyzed using EA-IRMS (Finigan MAT, Bremen, Germany). Gross rates of N mineralization, NH4+ consumption, nitrification, and NO3– consumption were calculated using the modified calculation procedure of Davidson et al. (1991) from the Kirkham and Bartholomew (1954) model.

The mean residence time (MRT) of the NH4+ and microbial N pools were calculated. Mean residence time indicates the average length of time an N atom resides in a given pool; a lower MRT indicates a faster pool turnover rate and hence a more dynamic pool (Hart et al., 1994a). The calculation of MRT (N pool ÷ flux rate; e.g., NH4+ pool MRT = NH4+ pool÷ gross N mineralization rate) assumed that the NH4+ and microbial N pools were at steady state and that the fluxes were equal to gross rates of N mineralization and N immobilization, respectively.

Other Supporting Soil Parameters
From each sampling point, additional soil samples from the same depth were taken and acclimatized at the same temperature and period as the cores used for soil N cycling measurements. These were used for measurement of initial NH4+ and NO3– concentrations (using the same extraction procedure mentioned above), and for microbial biomass N (MBN) determination. We used the fumigation-extraction procedure (Brookes et al., 1985; Davidson et al., 1989) for determining MBN. In this analysis, two 25-g fresh subsamples were taken; one pair was immediately extracted with 0.5 M K2SO4 (approx. 5:1 ratio of solution to dry mass soil) and the other pair was fumigated for 5 d and then extracted. Extractable organic N was determined using modified micro-Kjeldahl digestion (Davidson et al., 1989; Wyland et al., 1994; Stark and Hart, 1996). The Kjeldahl-N content was determined using the modified Berthelot reaction method (Skalar Method 475–000, 1992) and analyzed using continuous flow injection colorimetry. The difference in extracted Kjeldahl-N between the fumigated and unfumigated soils (N flush) is assumed to represent the N released from lysed soil microbes. The N flush was converted to MBN, using kN = 0.68 for 5-d fumigated samples (Shen et al., 1984; Brookes et al., 1985).

Other soil characteristics were determined at the start of the study and are reported in Table 2. Total organic C and N were measured from air-dried, ground samples using CNS Elemental Analyzer (Elementar Vario EL, Hanau, Germany). Bulk density was determined using soil core method, and soil pH was measured from a saturated paste mixture (1:1 ratio of soil to 1 M KCl). Base saturation was calculated as the percentage base cations of the cation exchange capacity (CEC); CEC was determined from air-dried, 2-mm sieved samples, percolated with 1 M NH4Cl, and the percolates analyzed for exchangeable cations using Flame-Atomic Absorption Spectrometer (Varian, Darmstadt, Germany).

Statistical Analyses
We tested the sampling points in each land use type and location for spatial independence using the data on gross N mineralization rates. This test was performed using the rank version of von Neumann's ratio test (Bartels, 1982). We found that our sampling points spaced at 10 m apart were spatially independent, and hence they were considered replicates in the succeeding analyses. Microbially mediated N processes were also reported earlier to show spatial independence at similar or even shorter distance: >1-m distance for N2O emission (Ambus and Christensen, 1994) and 10-m distance for gross rates of microbial N cycle (Corre et al., 2002). Tests for normality using Kolmogorov-Smirnov D statistic (Sokal and Rohlf, 1981) was first conducted for each of the measured parameters. The parameters were found to have normal distribution. Analyses were then performed using one-way ANOVA, with sites representing as treatments. Multiple comparisons of sites were conducted using a Least Significant Difference test at P ≤ 0.05. Means and standard errors were reported as measures of central tendency and dispersion, respectively.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Extractable NH4+ was generally higher than extractable NO3– on all sites, although a significant difference was only detected on the corn site of Location 1 and on all sites of location 3 (Fig. 1A ). The agroforest site in Location 1 showed similar NH4+ and NO3– concentrations with the forest site, while the long-term cultivated corn sites exhibited lower NO3– levels (Location 1) and NH4+ levels (Location 3) than the forest sites. In Location 2, where corn was established only 1 yr after forest clearing, the mineral N pools were comparable with the forest site. On the other hand, the agroforest site in Location 2, established after 2 yr of corn cultivation following forest clearing, showed lower NO3– concentrations than the forest site (Fig. 1A).


Figure 1
View larger version (26K):
[in this window]
[in a new window]
  		Fig. 1. (A) Initial NH4+ and NO3– pools, and (B) percentage of 15N recovery in the labeled pools 15 min (T0) after 15N injection in the intact cores. Means (bars for standard errors; n = 4) of either NH4+ or NO3– pools without letter or with the same letter indicate no significant difference among land use types at each location (One-way ANOVA, Least Significant Difference test at P ≤ 0.05). (n.d. = not detectable). * indicates significant difference between NH4+ and NO3– pools and between 15NH4+ and 15NO3– recoveries at each land use type and location (Independent t test at P ≤ 0.05).

 
On average (all land use types and locations), 37 ± 3% of the added 15NH4+ was recovered in the form added when the intact cores were extracted at T0 (Fig. 1B). There was no difference detected in 15NH4+ recoveries among land use types at each location. However, 15NH4+ recoveries differed among locations; higher 15NH4+ recoveries were observed in Location 2 (53 ± 4%) than Locations 1 and 3 (21 ± 2 and 36 ± 6%, respectively). We measured very low 15NO3– recoveries at T0 (Fig. 1B): 8 ± 2, 8 ± 1, and 6 ± 4% for Locations 1, 2, and 3-corn, respectively; for Location 3-forest neither added 15NO3– nor extant NO3– was detected. Two days after 15NO3– injection (T1), 15NO3– recoveries in NO3– pool were 16 ± 3, 13 ± 3, and 2 ± 2% for Locations 1, 2, and 3-corn, respectively; for Location 3-forest site NO3– and 15NO3– remained below detection limits. These T0 and T1 15NO3– recoveries did not differ (Paired-Samples t test at P > 0.05), indicating no detectable changes in the very low NO3– levels and 15NO3– atom percent excess within 2-d incubation period. As a result of these data, our estimate of gross nitrification and NO3– consumption rates (data not shown) were very low and were not significantly different from zero (One-sample t test at P > 0.05). Hence, we assumed that our measured gross NH4+ consumption rates were contributed largely by microbial immobilization and we calculated the MRT of microbial N pool by dividing microbial N levels with gross NH4+ consumption rates.

The agroforest sites in Locations 1 and 2 had comparable gross N mineralization rates with the reference forest sites while the corn sites in Locations 1 and 3 showed lower gross N mineralization rates than the forest sites (Fig. 2A ). Only the 1-yr-old corn cultivated site in Location 2 showed similar gross N mineralization rates as the agroforest and forest sites. There was no difference detected in gross NH4+ consumption rates among land use types at each location (Fig. 2A); however, these rates were correlated with gross N mineralization rates (r = 0.81; P = 0.00). Furthermore, the agroforest site in Location 1 tended to have a fast turnover rate of the NH4+ pool (i.e., short MRT), while the 3-yr-old corn cultivated site tended to have a slow NH4+ turnover rate (Fig. 2B). The 9-yr-old corn cultivated site in Location 3 showed the slowest NH4+ turnover rate. Again, only the 1-yr-old corn cultivated site in Location 2 tended to have fast NH4+ turnover rate, although this was not significantly different from the agroforest and forest sites (Fig. 2B).


Figure 2
View larger version (26K):
[in this window]
[in a new window]
  		Fig. 2. (A) Gross N mineralization rates and gross NH4+ consumption rates, and (B) mean residence time of NH4+ pool. Means (bars for standard errors; n = 4) with the same letters (lowercase for gross N mineralization rates and uppercase for gross NH4+ consumption rates) indicate no significant differences across all land use types and locations (One-way ANOVA, Least Significant Difference test at P ≤ 0.05).

 
Specific gross N mineralization rates (calculated as gross N mineralization rate ÷ microbial N pool) are a measure of activity per unit microbial biomass. Gross N mineralization rates reflect both the substrate and microbial biomass (presumably active in mineralization, e.g., chemoheterotrophs). By accounting for the difference in microbial biomass size, the specific gross N mineralization rates reflect the quantity and quality of organic N available for mineralization. In Locations 1 and 2, the specific gross N mineralization rates were highest in the agroforest sites, followed by the forest sites, and lowest in the corn sites (Table 3). A similar trend was observed for the turnover rates of microbial N pool: fastest in the agroforest and slowest in the corn sites (Table 3). In Location 3, the forest site showed low specific gross N mineralization rate, despite having the highest microbial N pool (Table 3). Under 9 yr of continuous corn cultivation, there was a slight decrease in specific gross N mineralization rate and a significant decrease in MBN. This site showed the slowest turnover rates of microbial N pool (Table 3).


View this table:
[in this window]
[in a new window]
  		Table 3. Differences among land use types and locations (soil types) on microbial biomass activity. Means (± 1 SE; n = 4) with the same letters indicate no significant differences across all land use types and locations (One-way ANOVA, Least Significant Difference test at P ≤ 0.05).

 

   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
CONCLUSIONS
 REFERENCES
 
Land-use Change Effects on Soil Nitrogen Cycling
Data on gross rates of soil N cycling in tropical land use systems are very few, and methodological differences make comparisons difficult. Garcia-Montiel and Binkley (1998) measured gross N transformation rates from tropical Eucalyptus saligna Sm. (non N2–fixer) and Albizia falcataria L. Fosberg (N2–fixer) plots in Hawaii, and rates were calculated without correction of the initial (T0) 15NH4+ and 15NO3– recoveries. Such calculation method could lead to overestimation of gross rates (Davidson et al., 1991). Our measured rates of gross N transformations were lower than those of Garcia-Montiel and Binkley (1998). However, our values were comparable with the rates measured by Hall and Matson (1999) from an N-limited, tropical montane forest in Hawaii and by Neill et al. (1999) from a tropical lowland forest converted to pastures of different ages in Brazilian Amazon; these studies and our study calculated the gross rates of N transformations with correction of the T0 recoveries.

The close correlation of N mineralization and NH4+ consumption rates in our study indicates that microbial assimilation of NH4+ was highly dependent on NH4+ availability (gross N mineralization). This may signify high competition for available N between microorganisms and plants. Such tightly coupled NH4+ transformation processes provide strong evidence that these land use systems are N limited. Hall and Matson (1999, 2003) also showed very low to undetectable rates of gross N transformations from an N-limited, tropical montane forest. The few data available regarding net N mineralization rates in tropical montane forests (Marrs et al., 1988; Rhoades and Coleman, 1999; Cavelier et al., 2000) also showed that the rates were lower than in tropical lowland forest, with the latter also having lower N immobilization by microbes than the former (Vitousek and Matson, 1988).

The establishment of the corn and agroforest systems in our study both involved forest clearing and biomass burning. This type of land conversion often leads to N losses. However, the agroforest system showed higher NH4+ cycling rates, which was comparable with the forest, than the long-term corn cultivation. The difference in NH4+ cycling activity among these land use systems is probably related to the difference in quality and quantity of available organic matter. The N-fixing shade trees in the agroforestry systems provide additional N to the system through release of N-rich root exudates and plant residues (Beer et al., 1998; Schroth et al., 2001). The fast NH4+ turnover rates and high specific gross N mineralization rates in the agroforest system suggest a high quality substrate is available and supports high microbial activity. In the case of corn, continuous cultivation, and N export by harvest without external N input appears to have depleted the levels of available organic N that could be potentially mineralized. An ancillary chronosequence study in the same research area showed declining soil C and N stocks under continuous corn cultivation, but stable C and N stocks under the agroforestry system (Dechert et al., 2004). Only the 1-yr cultivated corn site showed NH4+ cycling rates similar to the agroforest and forest sites, presumably this was because available organic matter was not yet as depleted as continuously cultivated sites.

In addition, the agroforest system also showed lower MBN but faster turnover rate of microbial N pool than the corn cultivation. Hart et al. (1994a) showed previously that fast turnover rates of N pools was a response to an increase in C availability. The fast turnover rates of microbial N in agroforest systems may be due to an increase in available organic matter. Our results also suggest that microbial biomass alone was not a good indicator of rates of soil N cycling under these land use systems, and a decline of microbial biomass after forest conversion does not necessarily result in reduced microbial activity.

Implications of Rapid Reaction of 15NO3– Added to Soils
We observed very low 15NO3– recoveries in the NO3– pool at 5 min after 15NO3– addition. This fast reaction of NO3– is typically attributed to abiotic immobilization (Berntson and Aber, 2000). Recovery of 15NO3– at T0 did not differ with that at T1, signifying that this initial reaction was completed within a short period and/or the immobilized 15NO3– was not released back to the NO3– pool within the 2-d incubation period. We were unable to detect significant NO3– transformation activity. Hall and Matson (2003) reported similar results from an N-limited upland forest in Hawaii, which also showed absent to negligible NO3– pools. The fast reaction of NO3– in our sites possibly imposed additional competition for the fate of NO3–, and could in part explain the dominance of NH4+ over NO3– concentrations in all our sites. The low levels of mineral N with high NH4+/NO3– ratios in our study sites were similar to the N-limited, tropical montane forests in Costa Rica (Marrs et al., 1988), Hawaii (Hall and Matson, 1999, 2003) and Colombia (Cavelier et al., 2000), and were comparable to N-limited > 10-yr-old pastures established from cleared tropical lowland forests (Neill et al., 1999; Veldkamp et al., 1999).

We did not ascertain which soil N pools the unrecovered portion of the added 15NO3– was converted at T0. Recent studies from temperate forest ecosystems reported significant fast abiotic NO3– immobilization (Berntson and Aber, 2000; Dail et al., 2001; Corre et al., 2003; Fitzhugh et al., 2003; Corre and Lamersdorf, 2004). Reports of 15N recoveries in different organic N pools 10 to 15 min after 15NO3– addition varied: Dail et al. (2001) reported 30 to 55% recovery in the dissolved organic N and 5 to 8% in the insoluble organic N fractions, Fitzhugh et al. (2003) reported 7% in the insoluble organic N fraction, and Corre and Lamersdorf (2004) reported 0.6 to 1.2% in the dissolved organic N and 49 to 79% in the insoluble organic N fractions. Recently, this fast NO3– reaction is considered as a possible mechanism contributing to N retention in temperate ecosystems (Davidson et al., 2003). The importance of abiotic N immobilization in the tropical ecosystems certainly warrants further investigation.

Our 15NH4+ recoveries at T0 were similar with those of Hall and Matson (1999) measured from an N-limited, tropical montane forests in Hawaii. The fast reactions of added 15NH4+ are usually attributed to abiotic NH4+ immobilization [e.g., physical condensation reactions with phenolic compounds (Nömmik and Vahtras, 1982), and fixation on clay minerals (Davidson et al., 1991)]. The differences in 15NH4+ recoveries among locations could possibly be due to the differences in clay mineralogy inherent to the differences in soil parent materials of these locations. Furthermore, the higher recoveries of 15NH4+ than of 15NO3– at T0 signify a potential for higher abiotic N immobilization for NO3– than for NH4+ in these sites.


   CONCLUSIONS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
REFERENCES
 
For the sites included in our study, our results show that agroforestry was a better option compared with corn in terms of sustainability in the N-supplying capacity of the soil. Agroforest systems and indigenous forests showed higher soil N cycling rates than the long-term corn cultivation. Leguminous shade trees, which are important components of these agroforest systems, may have influenced the fast microbial N cycling by providing high quantity and quality of available organic matter. Our results also show that microbial biomass was not predictive of the rates of soil N cycling, and hence microbial biomass cannot be used as an indicator of the N-supplying capacity in these land use types. The results of our study provide vital information for resource management-oriented projects in Central Sulawesi, where ongoing conversion of rainforest to unfertilized agricultural systems occurs apace. The common practice of corn cultivation, which fuels further forest clearing when the cultivated areas become depleted in nutrients, can be avoided when agroforestry programs in the presently cleared areas are supported. Sustainability of agroforest system, in terms of the N-supplying capacity of the soil, may lessen further forest clearing in the long run. However, the overall effect of agroforestry on soil fertility needs to be further evaluated for other nutrients, e.g., phosphorus. Our study and that of Hall and Matson (1999) are the only two so far to indicate the potential importance of abiotic N immobilization, especially of NO3–, in tropical ecosystems. The significance of this process certainly warrants further investigation, for example, the role it will play when increased external N inputs occur in these N-limited systems.


   ACKNOWLEDGMENTS
 
This study was a subproject of the research program "Stability of Rainforest Margins" (STORMA), which is comanaged by the Universities of Goettingen and Kassel in Germany and the Universities of Bogor and Palu in Indonesia. The program was funded by the German Research Foundation (Deutsche Forschungsgemeinschaft). We thank the laboratory staff of the Institute of Soil Science and Forest Nutrition, University of Goettingen for their technical support. The thorough reviews of Dr. Helga Van Miegroet and the three anonymous reviewers are highly appreciated.

Received for publication February 28, 2005.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 





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 ISI Web of Science (2)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Corre, M. D.
Right arrow Articles by Veldkamp, E.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Corre, M. D.
Right arrow Articles by Veldkamp, E.
Agricola
Right arrow Articles by Corre, M. D.
Right arrow Articles by Veldkamp, E.
Related Collections
Right arrow Agricultural Systems
Right arrow Agroforestry
Right arrow Soil Biochemistry
Right arrow Soil Biology

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