| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a Dep. Plant Sciences, Univ. of California, Davis, CA 95616
b Dep. of Land, Air and Water Resources, Univ. of California, Davis, CA 95616
* Corresponding author (moran{at}anl.gov)
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Abbreviations: MB, microbial biomass • mSOM, mineral-associated soil organic matter • OM, organic matter • POM, particulate organic matter • SE, standard error • SOM, soil organic matter
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Plant residue-N is the main source of N in stable SOM (Angers et al., 1997; Bird et al., 2003), especially in natural systems. But, it is not well understood how N from residue versus mineral sources not associated with C, such as fertilizers and deposition, are transformed into stable SOM when both are present. Past research suggests that there is an opportunity for mineral-N to be preferentially sequestered in stable SOM over residue-N. Bird et al. (2003) reported divergent pathways of residue C and N during the formation of SOM suggesting an uncoupling of processes that affect the fate of plant residue C and N. Application of 15N-labeled fertilizer or maize residue together resulted in more than twice as much 15N recovered from fertilizer than from residue in the SOM (Dourado-Neto et al., 2001).
At the least, residue-N will affect mineral-N transformations and vice versa in the soil through the microbial biomass (MB). Inorganic-N availability may be regulated by organic-N availability through the regulation of proteolytic enzymes (Smith et al., 1989). Azam et al. (1985) found that in the presence of legume residues, more ammonium sulfate-N (19%) was transformed into humic compounds than applied without legume residue. The authors attributed this to greater microbial activity when plant residue was added. Microbial preference for the N source that is immediately available and can be assimilated without prior chemical transformation may lead to greater uptake of mineral-N than residue-N. Since microbial by-products are thought to be a major contributor to stable SOM formation (Nelson et al., 1979; Stevenson, 1994), mineral-N may be preferentially stabilized over residue-N.
Changes in the POM and mobile humic acid fractions can show shifts in C dynamics under different treatments. The POM fraction has been shown to be a more sensitive indicator of the effects of mineral-N additions on soil C and N than total SOM (Malhi et al., 2002). Recent research has also shown that the N content of NaOH-extractable mobile humic fractions derived from fertilizer-N are affected by crop residue management practices (Bird et al., 2003; Devevre and Horwath, 2001). The POM fraction was shown to cycle the greatest amount of fertilizer-N followed by humic and fulvic acids, and then humin (Bird et al., 2002). Since humic fractions are transformed during NaOH-extraction from soil, these pools are not a direct measure of original SOM. Humic fractions are chemically extracted without any relationship to soil structure or biological activity. Nevertheless, NaOH-extractable humic fractions are useful to characterize OM pools representing a range of turnover rates and recalcitrance.
The objectives of this study were to determine: (i) the role of mineral-N as an N source for stable SOM-N in the presence of residue-N; and (ii) whether mineral-N inputs could potentially increase the movement of residue-C into more stable fractions of SOM-C. We hypothesized that mineral-N contributes more to stable SOM formation than residue-N. To test this hypothesis, a 90-d incubation study using 13C15N-labeled rice residue and 15N-(NH4)2SO4 was conducted.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The top 10 cm of soil was sampled at 20 random positions throughout the field and air-dried to facilitate sieving to 4 mm; roots and large pieces of plant material were removed. The soil was hand-sorted to further remove all OM >2 mm. The bulk soil contained 16 mg C g–1 soil and 1.3 mg N g–1 soil. Texture was 430 mg sand g–1 soil and 420 mg clay g–1 soil, determined by the hydrometer method (Gee and Bauder, 1986). Before the experiment, initial soil samples analyzed had 0.2 mg of mineral-N per gram of soil.
Plant Residue Labeling
Seeds of a medium-grain, short-season rice (Oryza sativa L.) cultivar (M-103) were obtained from the California Cooperative Rice Research Foundation. A dilute soil potting media composed of rice soil and sand (1:3) was used to maximize uptake of 15N fertilizer. Half of the pots were left unlabeled to produce non–15N and 13C-labeled rice residues grown under similar conditions (see below). Plants were fertilized every 2 to 3 wk with half-strength Hoagland solution (Hoagland and Arnon, 1950) with double the recommended concentration of Zn and Fe and zero N. Nitrogen was applied with a separate solution of (NH4)2SO4 with labeled plants receiving 5.0 atom % 15N-(NH4)2SO4 and unlabeled plants receiving natural isotope abundance (NH4)2SO4. We followed the method of Bird et al. (2003) to ensure uniform labeling through frequent applications with reassimilation of respired 13C the following day. Plants were labeled with 13C-CO2 in a climate-controlled chamber (Bird et al., 2003) every 2 to 3 wk for the first 4 mo and once a week during the final 2 mo of growth. Then, senescing plants received a final pulse of 13CO2 and were immediately moved to a dark drying oven at 30°C. Grain was removed and stover cut just above the root crown. The stover was further dried, milled to 2-mm pieces and sieved such that residue pieces were between 0.5 and 2 mm. Subsamples were ball-milled and analyzed for C and N content, atom %15N and 
13C by an automated N/C analyzer-isotope ratio mass spectrometer (ANCA-IRMS, Europa Scientific Integra, UK) at the UC Davis Stable Isotope Facility.
Residue and Mineral-Nitrogen Incubation
The experimental design was a completely randomized design with four main treatments and two controls (four replications). Soil was amended with the four treatments of: (1) 15N-labeled mineral-N only [(NH4)2SO4, at 5.0 atom % 15N]; (2) 13C15N-labeled residue only; (3) natural isotope abundance mineral-N [(NH4)2SO4), at 0.37 atom % 15N] together with 13C15N-labeled residue; and (4) 15N-labeled mineral-N together with unlabeled residue. A control soil amended with only unlabeled residue was included to determine whether there were any differences in the decomposition of the 13C15N-labeled residue or the unlabeled residue. A second control was soil only. Treatments amended with residues received 0.2 ± 0.0001 g rice straw per 40 g soil (0.5 mg N and 72.1 mg C) and those amended with mineral-N received 3.4 mg N per 40 g soil.
For 1 wk before the start of the incubation, the soil was pre-incubated at 40% water holding capacity at 25°C. On Day 0, five 40-g samples of soil were used to determine the initial, pre-incubation values for 13C and 15N, mineral-N, and SOM fractions. The remaining samples were treated with their respective amendments. Soils were amended with residues in specimen cups, followed by additions of either deionized water or mineral-N solution to establish a final moisture content of 55% water holding capacity. The specimen cup was placed in a 970-mL mason jar with 2 mL of water in the bottom to prevent desiccation. The jar was then sealed with a modified Mason jar lid containing a gas sampling septum. Mason jars were incubated under standard conditions in the dark at 25°C for 90 d.
Gas Sampling
On Days 1, 2, 4, 6, 15, 30, 60, and 90, headspace CO2 and 13CO2 were determined. Three empty mason jars provided the blank (background) CO2 concentration and C-isotopic composition. 12 mL of headspace gas was transferred into an evacuated container (Exetainer, Labco, UK), and subsequently analyzed by a trace gas-isotope ratio mass spectrometer (TGII-IRMS, PDZ Europa, UK). After each gas sampling, mason jars were opened and allowed to come into equilibrium with ambient levels of CO2.
Residue-derived respiration was calculated using an isotope-mixing model:
 |
t is the PDB of the measured total respiration; res is the PDB of the added labeled residue (3461 ± 0.7); bl is the PDB measured from the blanks; and soil is the PDB of CO2 derived from the soil in the soil-only control treatments, calculated from the equation:
 |
Soil Organic Matter Fractionation
After 90 d, soil (
15 g) from each treatment replicate was air-dried before SOM fractionation. First, a physical separation was undertaken to separate undecomposed OM (>500 µm, the size of the smallest residue pieces added at the beginning of the incubation), POM (53–500 µm), and mineral-associated SOM (mSOM < 53 µm). The soil was shaken with sodium hexametaphospate (1:3 soil/0.5% HMP) on a reciprocal shaker for 14 h, then washed through two stacked sieves (500 and 53 µm) with deionized water. The >500-µm fraction and the POM fraction collected on each sieve were rinsed into aluminum pans and dried at 105°C for 48 h to a constant weight. The remaining <53 µm that passed through both sieves was rinsed into a centrifuge bottle and saved for chemical fractionation.
Chemical fractionation of the <53-µm physical fraction consisted of alkali extraction of fulvic and humic acids from the humin, followed by acid precipitation of the humic acid from the fulvic acid (adapted from Bird et al., 2002). Concentrated NaOH was added to the suspended <53-µm physical fraction to achieve 200 mL of a 0.4 M NaOH solution, followed by purging with He. Samples were then shaken for 12 h on a reciprocal shaker, centrifuged (15200 xg, 25°C, 15 min.), the dark-colored supernatant (humic and fulvic acids) was decanted off and saved, followed by the addition of 100 mL of 0.4 M NaOH, and purging with He. This step was repeated (7 times) until the supernatant was colorless. The alkali insoluble residue (humin) was neutralized with concentrated HCl and oven-dried. Humic acids were precipitated from the supernatant (fulvic acids) through the addition of concentrated HCl until pH 1.5 and centrifuged (25600 xg, 10°C, 15 min.). All fractions were ground to fine powder, and subsamples were packed in tin capsules and analyzed for C and N content, and 13C and 15N isotopic composition by ANCA-IRMS.
Five soil samples taken on Day 0 were fractionated to determine initial (pre-amendment) SOM content and distribution in the various pools. These samples had 0.6 ± 0.1 mg undecomposed (>500 µg)-C per g of soil, 3.9 ± 0.2 mg POM-C per g of soil, 1.5 ± 0.1 mg humic acid-C per g of soil, 2.7 ± 0.2 mg fulvic acid-C per g of soil, and 6.4 ± 0.3 mg humin-C per g of soil.
Since the fulvic acid fraction collected also contained inorganic N extracted from the soil, a 10-mL aliquot of the fulvic acid fraction collected for each sample was dialyzed to remove Na and inorganic N using 500 (molecular weight cutoff, MWCO) cellulose-ester membrane (Spectrum Industries, Inc., Rancho Dominguez, CA) in deionized water. The remaining solution was freeze-dried and analyzed by ANCA-IRMS to determine the percentage of N and 15N composition of the fulvic acid and the inorganic N/15N content was calculated by difference.
The fraction (f) of C derived from residue in any OM fraction was calculated by:
 |
end is PDB of the treatment C; initial is the PDB of the initial fraction; and input is the PDB of the labeled residue. The amount of residue-derived C in the OM fraction is the product of f x [total fraction C]. The residue or mineral derived N was first calculated as percentage of N derived from labeled source (percent Ndfl):
 |
The original OM also changed and redistributed among the various pools during the period of the incubation as its own decomposition continued and was impacted by treatment effects. Some initial OM was also respired, defined earlier as soil-derived respiration, which we measured throughout the incubation. At the end of 90 d, the net C and N that moved into or out of designated fractions that was not derived from inputs was calculated as C or N transfer:
 |
Statistical Analysis
All statistical analyses were performed using the SAS system version 8 (SAS Institute Inc., NC). Data was analyzed using ANOVA (the GLM procedure) with residue, mineral-N input, or combinations as treatments. Significant differences (P < 0.05) and trends (P < 0.1 and P < 0.2) among treatments and controls were determined using Tukey's Studentized Range Test.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
|
|
|
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Jenkinson and Rayner (1985) suggested that added mineral-N might increase the decomposition rate of straw by satisfying N requirements of microorganisms. There is also evidence that in the early stages of plant decomposition, mineral-N inputs stimulate hydrolysis of soluble C compounds and nonlignified holocellulose (Berg and Matzner, 1997). However, our study shows this effect is lost after 30 to 60 d (Fig. 1) when the difference in residue-derived CO2 mineralization begins to narrow. At this point, the addition of mineral-N may have rapidly exhausted all labile C. At the end of the incubation, mineral-N levels are still high in these treatments (Fig. 3). Many studies have also shown that mineral-N inhibits the later-stage lignin decomposers (Henriksen and Breland, 1999; Carreiro et al., 2000; Saiya-Cork et al., 2002), and our treatments may have been similarly affected, since rice straw is between 15 and 20% lignin.
Residue-Carbon in Soil Organic Matter Fractions
Mineral-N inputs lead to a faster formation of recalcitrant material (Ågren et al., 2001) possibly from greater microbial turnover (Devevre and Horwath, 2001). A faster decomposition at early stages of the incubation in the mineral-N input treatment led to a greater movement of residue-C into the more decomposed and stable humin fraction at the end of 90 d (Fig. 2). The increase in new residue-C incorporated into the humin with mineral-N input, though small, shows a shift in the equilibrium of the stable humin fraction toward a new C accumulation. These results indicate potential changes in soil C storage that could become more important over time because small short-term differences in decomposition with mineral-N input can add up to large differences in long-term storage of SOM (Ågren et al., 2001). With continuous mineral-N and residue additions, humin C may be maintained or increase with time.
Mineral-N may also change the loss of C from the humin pool. Neff et al. (2002) found no significant net change of soil C due to chronic N inputs in a natural ecosystem, but they observed that mineral-N input changed C cycling among the various stable and dynamic pools. Malhi et al. (2002) found the POM fraction to change most with mineral-N addition to bromegrass-amended soil, but total organic C and N did not change nearly as much. Our data is similar in that our treatments show no significant differences in the total C in SOM fractions. However, while total C pools were not affected by treatments after a 90-d incubation, new C accumulation did show significant differences in the humin and humic acid fractions, indicating new C inputs were maintaining fractions.
Mineral-Nitrogen vs. Residue-Nitrogen in Organic Matter Fractions
The greatest mineral-N recovery was in the inorganic N fraction (Fig. 3) with small amounts being sequestered in mSOM. Thus, a majority of it remains in a dynamic state that under field conditions could be lost (leached or denitrified) or taken up by plants. This was also indicated by our lower recovery of 15N in our mineral-N treatments, especially mineral-N alone. For the mineral-N treatment, the absence of a C-source limits transformation of the N into more stable forms. The greater recovery of mineral-N in the humin and POM fractions shows a larger sequestration of mineral-N into POM and mSOM fractions when a C source is present. Mineral-N that moved into the POM and >500-µm fractions (Fig. 3) may be attributed to fungal growth in these fractions that immobilized some of the added mineral-N and translocated it into the larger OM fractions to aid decomposition (Frey et al., 2003; Salas et al., 2003).
Due to the multiple steps involved in fractionations such as the present one, values within 10% are usually well within the error acceptable for recoveries. The lower recoveries were for treatments with mineral–15N. This should not affect our conclusions, since the major reason for losses of N is soluble N lost during rinsing of the physical fractions. In a sense this can be considered similar to leaching losses of N under field conditions. While they are not totally equivalent, greater losses seen during the fractionation of certain treatments may indicate that these treatments would likely show relatively similar losses in the field. In a sense, this fractionation made the incubations, traditionally seen as a closed system, open systems (as field systems are). Therefore, inorganic N values in Fig. 3, 4, and 5 are underestimations of inorganic N present in the incubated soil after 90 d, but this does not affect the interpretation of N present in the OM fractions.
Normalized for N addition (Fig. 4), a greater amount of residue-N was recovered in all of the OM fractions, indicating a preference for residue-N assimilation into SOM. This is in contrast to our hypothesis that mineral-N is the primary contributor to SOM formation. The mechanism for why residue-N would be more likely to contribute to stable SOM may be that it is preferentially assimilated by microbes with residue-C into stable byproducts, or that chemical stabilization mechanisms favor residue-N from decomposition products. Stable SOM can also be formed when lignin reacts with mineral-N or residue decomposition products and undergoes other condensation reactions in the first step of humification (Nömmik and Vahtras, 1982; Berg and Matzner, 1997). However, this process requires high pH, which our soil did not have.
There was a trend (P < 0.2) of less residue-N recovery in mSOM fractions with mineral-N addition, and a trend (P < 0.2) of more mineral-N recovery in the mSOM fractions with residue addition (Table 1). This suggests that mineral-N is replacing some of the residue-N that is sequestered when these two inputs are added together. However, the source of the N notwithstanding, an added C source results in greater transformation of all forms of added-N into the humin overall (Fig. 5). In this incubation we added 6.8 times more mineral-N than residue-N, to mimic amendment additions under typical agronomic conditions. When standardized on an N-added basis, more residue-N is transformed into humin than mineral-N when both are added together (Fig. 5). Therefore, there was no preference for mineral-N to be sequestered in stable SOM but residue addition did promote the transformation of mineral-N into humin. There is a net benefit to N sequestration when residue and mineral-N inputs are added together. Residue and mineral-N treatment result in the recovery of more new N in humin than if either was amended alone.
Further understanding this interaction between mineral-N and residue inputs to soil in the long term could have additional benefits beyond C-sequestration. This understanding could also lead to better agronomic management to mitigate fertilizer-N losses, since residue addition with mineral-N promotes N transformation into more stable pools of SOM. While the influence of mineral-N on C sequestration is still under investigation, which led to the present study, the mitigation of N-losses with residue amendments is not a completely new concept in agronomy. Azam et al. (1985) found that plants took up less fertilizer-N when legume residues were added, and more fertilizer-N was retained in the soil and transformed into MB and humus components. Vanlauwe et al. (2001) proposed to combine the addition of OM and fertilizer for its possible positive interactions and added benefits such as a delay in fertilizer-N loss to deeper layers and greater recovery of fertilizer-N in subsequent crops. Thus, this may be a common phenomenon where mineral-N and residue additions enhance SOM formation more than adding one component alone. Besides the N-retention benefits, OM and mineral-N additions together improve soil-water conditions beyond what either amendment can accomplish alone (Vanlauwe et al., 2001).
|
ACKNOWLEDGMENTS |
|---|
Received for publication September 7, 2004.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  C. Bosshard, E. Frossard, D. Dubois, P. Mader, I. Manolov, and A. Oberson Incorporation of Nitrogen-15-Labeled Amendments into Physically Separated Soil Organic Matter Fractions Soil Sci. Soc. Am. J., May 29, 2008; 72(4): 949 - 959. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 | |||
