| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Dep. of Food Production, Faculty of Agriculture and Natural Sciences, The Univ. of the West Indies, St. Augustine Campus, Trinidad
* Corresponding author (gouv{at}carib-link.net)
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONREFERENCES |
|---|
525 µg NH4-N and was more dependent on digest volume than on its N content. Fractionation of 15N was effectively nonexistent within the diffusion periods used except only for immediately (a few hours) into the diffusion; suggesting that the need to optimize diffusion is not that critical to the accuracy of the 15N results provided a sufficient quantity of N is obtained for the CF-IRMS. Additionally, diffusion times were slightly affected by soil type, indicating a mild matrix effect probably because of mineralogical variation among soils. These diffusion methods therefore proved to be very accurate and reliable for practical use as a preconcentration step prior to 15N analysis of the NH4f fraction in soils.
Abbreviations: ANCA, automated N and C analyzer • BC, Bejucal clay • CEC, cation-exchange capacity • CF-IRMS, continuous flow–isotope ratio mass spectrometry • CV, coefficient of variation • NH4e, exchangeable ammonium • NH4f, fixed ammonium • PFS, Piarco fine sand • REL, River Estate loam
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONREFERENCES |
|---|
For a long time, the method of Silva and Bremner (1966), which includes a KOBr pretreatment to get rid of N associated with the organic fraction prior to destruction of the minerals with a HF–HCl mixture to liberate the NH4f, has been the widely accepted method for determining NH4f in soils. However, recent findings of Paramasivam and Breitenbeck (2000) showed that a significant proportion of the recently fixed fraction of the total NH4f, which is not so strongly held and which is located near the edges of clay minerals, is also removed by the KOBr washing. Therefore, this suggests that while their method, which excludes the KOBr pretreatment and uses a membrane diffusion technique to quantify the NH4f liberated by the HF digestion, is able to quantify the total NH4f in soils by difference with NH4 that is extracted with 2 M KCl, the Silva and Bremner (1966) method quantifies only that fraction of NH4f that is strongly fixed and not removed by KOBr treatment. This highlights the need to redefine NH4f by recognizing two forms; one that is strongly fixed and determined by the method of Silva and Bremner (1966) and the other that is not so strongly fixed and is otherwise removed by the one washing with KOBr. The latter could therefore only be calculated based on the difference between the results of NH4f determinations by the two methods. The behavior of these two forms is likely to differ particularly in relation to their release characteristics and potential availability for plant utilization. The continued use of the Silva and Bremner (1966) method is therefore critical towards the better understanding of NH4f behavior in soils. The fact that there is very little difference between the two methods for the NH4f measured in soils that do not contain recently NH4f (Paramasivam and Breitenbeck, 2000), suggests that the not so strongly fixed fraction somehow disappears with time. This further supports the need to distinguish the Silva and Bremner (1966) NH4f from that of Paramasivam and Breitenbeck (2000) to help trace the fate over time of this recently NH4f that is not so strongly held. Hereon after, NH4f would refer to that fraction which is strongly held and quantified by the method of Silva and Bremner (1966).
Following the release of NH4f via destruction of the clay minerals, distillation procedures have been widely used to quantify the amounts of NH4 present in the extract (Mulvaney, 1996). In instances where 15N analysis is to be conducted on this fraction by CF-IRMS, a preconcentration step is necessary because the N concentration in the extract is normally too low for direct injection into the CF-IRMS system (Barrie et al., 1995). A widely used preconcentration technique is microdiffusion, in which the NH4 extracts are incubated under alkaline conditions in a closed vessel allowing the transfer of the NH4 via NH3 formation and diffusion into a receiving absorbent (Brooks et al., 1989; Kelley et al., 1991; Liu and Mulvaney, 1992; Mulvaney, 1996).
Microdiffusion has been present for quite awhile as an alternative for the quantitative determination and isotope ratio analysis of inorganic N in soils (Brooks et al., 1989) with significant modifications having been made within the last decade. Recorded improvements (Khan et al., 1997; Mulvaney et al., 1997; Khan et al., 1998) have allowed for greater accuracy, precision, and sensitivity, as well as, speed through the successful utilization of mason jars for the quantitative determination of the available inorganic N fractions in soils. Diffusion methods have also been effective in quantifying inorganic N in a complex organic matrix (Mulvaney and Khan, 1999) with negligible interference from labile organic compounds. Recently, Khan et al. (2000) reported on a modification that allows direct diffusion of NH4-N from the soil solution matrix without any prior extraction.
Quantitative inorganic-N analysis by microdiffusion requires the diffusion period to be sufficiently long to ensure that transfer of NH3, via liberation and subsequent absorption, is virtually complete. Lory and Russelle (1994) stated that because of mass difference between 14NH4 and 15NH4, incomplete diffusion can result in a disproportionately lower transfer of the heavier molecule, leading to an underestimation of the contribution by the labeled pool of N. However, Stark and Hart (1996) indicated that diffusions of NH4-N are subject to very little error from isotopic fractionation and need not be complete; a view supported by Khan et al. (1998). This, however, is true providing NH4 is the only N fraction being analyzed for 15N, since if NO3 in the extract is also to be measured either separately or in combination with NH4, the diffusion of NH4 would be required to be complete, otherwise erroneous results would be obtained. The minimal period for complete diffusion is, in turn, determined by a combination of factors, most important of which are the design of the diffusion unit, the capacity and dynamics of NH3 absorption by the trap, incubation temperature, sample volume, the quantity of N to be diffused, and the sample's level of salinity. While many of these factors have been studied to determine their effects on the incubation period for complete diffusion of extractable ammonium (NH4e), NO3, and NO2 in KCl digests (Mulvaney et al., 1997), this has not been done for NH4f. Nitrogen-15 analysis of NH4f has been achieved through transfer of the NH4f into H3BO3 via distillation, which is then converted to N2 gas via the Rittenberg method. This can be time-consuming, requiring technical skill, and is subject to cross contamination if many samples are being distilled and not suitable for automated CF-IRMS systems such as an automated N and C analyzer–mass spectrometer (ANCA-MS) (Brooks et al., 1989; Kelley et al., 1991).
Since the HF acid medium used to liberate NH4f is a good solvent for silicates, the incubation conditions for this type of digest would be expected to differ from that for the KCl extracts of available soil N, and definitely should be confined to nonsiliceous diffusion jars made of polypropylene or other synthetic plastics. This paper therefore presents evidence based on successful investigations into the application of microdiffusion techniques towards the preconcentration of NH4f prior to 15N analysis by CF-IRMS. Note that the information being presented pertain to that fraction of NH4f determined by the method of Silva and Bremner (1966) and does not apply to that determined by the method of Paramasivam and Breitenbeck (2000). The scope for research guiding any further modifications to the microdiffusion technique being presented therefore exists for NH4f obtained by the latter.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONREFERENCES |
|---|
|
For the experiments that followed a sufficient volume (2 L) of soil digests for NH4f was obtained by using the digestion procedure of Silva and Bremner (1966). In the procedure, the soil (air dried and ground to <100 µm) was initially pretreated with KOBr to rid it of organic N compounds and not so strongly held NH4f, including exchangeable NH4, which was followed by washing with a dilute salt solution, and finally, treatment with 5 M HF:1 M HCl to liberate the fixed NH4.
Reagents
Boric Acid Indicator Solution (20 g H3BO3 L-1) (Mulvaney, 1996)
To 9 L deionized water in a 10-L Pyrex bottle (Corning Glassworks, Scence Products, Corning, NJ) (marked to indicate a volume of 10 L were added 200 g reagent grade H3BO3. The H3BO3 was dissolved by vigorous stirring with a motorized stirrer. Then 200 mL of indicator solution was added (prepared by dissolving 0.247 g of bromocresol green and 0.165 g of methyl red in 250 mL of ethanol), and the volume was brought to 10 L with deionized water. With continuous stirring, the pH of this solution was adjusted to between 4.8 and 5.0 by cautiously adding single NaOH pellets.
Potassium Hypobromite Solution (Mulvaney, 1996)
To 200 mL of deionized water in a 250-mL Erlenmeyer flask containing a Teflon-coated magnetic stirring bar, 22 g of KOH were added. Then the flask was immersed in a container of crushed ice, and the solution was allowed to cool to 5°C. With constant stirring, 6 mL of Br2 was added dropwise from a burette at a rate of 0.5 mL min-1. The solution was stored in a refrigerator.
 |
All safety precautions associated with the use of these chemicals were applied, such as the use of laboratory coats, eye protectors, gloves, and fumehoods.
Apparati
Diffusion Unit
The unit used consisted of a 140-mL specimen polypropylene container, normally used for urine samples and described by Brooks et al. (1989); this can be obtained from any dispenser of medical or laboratory supplies. Modifications were made enabling the support of a small polypropylene petri dish, capable of holding 5 mL of a H3BO3–indicator solution, from the cover utilizing a screw-down mounting base (screw, nuts, and washers). These modifications were applied directly to the polypropylene screw-on cap and were similar to that described and demonstrated by Mulvaney et al. (1997). Caution was taken to ensure that the cap did not distort during tightening onto the container, in which case an airtight seal may not be obtained. Modifications were also made to other diffusion units for 15N preconcentration, through the inclusion of two stainless steel wire devices (similar in size and design to that described and shown by Khan et al., 1998) to the screw-down mounting base off the inside of the lid. Each of the two wires supported a 7-mm diam. disc of Whatman GF-D filter paper (Whatman International Ltd., Maidstone, UK), which was acidified with 10 µl of 2.5 M KHSO4 to serve as the 15NH3-trap.
In setting up the units, a specified volume of 10 M KOH was first added to each container; the volume dependent on the volume of 5 M HF:1 M HCl digest/solution to be subsequently added (normally a ratio of 1:1 is adequate). The acid digest/solution was then cautiously added below the surface of the KOH ensuring that minimal mixing occurred during this process. The container was then sealed with the lid assembly and swirled gently; care being taken to eliminate any possibility of the KOH getting into contact with the petri dish containing the H3BO3-indicator solution, since this can lead to overestimation of diffused NH3-N, or with the acidified disks which could reduce their capacity to trap NH3. Because of the evolution of water vapor on addition of the 5 M HF:1 M HCl digest/solution to the 10 M KOH, the internal pressure rise may result in slight distortion of the cap if it is not fastened properly. Although the diffusion units were sealed while slightly warm there was no deformation of the unit upon cooling, and the airtight seal was maintained.
Continuous Flow–Isotope Ratio Mass Spectrometer
The system used was an automated N analyzer (ANCA; Model NA 1500, Carlo Erba Strumentazione, Rodano, Italy) interfaced with a mass spectrometer (MS; Model Tracermass–Stable Isotope Analyzer, Europa Scientific, Cheshire, UK) (Barrie et al., 1995).
Microburette
Manual titrations were carried out using a 10-mL microburette graduated at 0.02-mL intervals and equipped with a three-way stopcock for rapid refilling from a reservoir.
Wrist Hand Shaker
A heavy-duty shaker (Burrell Model 75, Burrell Corp., Pittsburg, PA) equipped with 12 holding hands.
Distillation Unit
Kjectec 1002 distilling unit (Itecator, Tecator AB, Hoganas, Sweden) with a timer was used.
Preliminary Study with Ammonium-Nitrogen Solutions
Native concentrations of NH4f-N in Caribbean soils range from as low as 10 mg kg-1 in sandy soils to >500 mg kg-1 in soils containing NH4 trapping clay minerals (Dalal, 1978; Ahmad et al., 1982). Accordingly, a range of quantities of N were used in this investigation to mimic this natural condition. The time-sequence recovery of NH4 from acid (5 M HF:1 M HCl) solutions of varying volumes (5–20 mL) and quantities of N (50–200 µg) were determined. All treatment combinations were replicated four times. Diffusion units were established as described above with boric acid traps and incubated at room temperature (25°C). At sampling, the lid assembly was removed and the petri dish carefully detached from it. The H3BO3 in each dish was brought up to 10 mL with deionized water and the trapped NH3 was determined by titration with standardized H2SO4 (of 0.0025 M) from a microburette. The quantity of NH4-N recovered was determined by the following equation,
 | [1] |
Study with Soil Digests Containing NH4f-N and Added NH4-N
A study was conducted with soil digests, their native NH4f-N concentrations being first confirmed by distillation (Bremner, 1965; Silva and Bremner, 1966) and the sampling time-sequence based on the results of the preliminary study with similar volumes, and to a lesser extent, on the N levels present. Five-, 10-, and 20-mL aliquots of soil digest/spike combinations were added to the specimen containers already containing a known volume of KOH and the units prepared with the boric acid traps and sealed as described previously. The N spikes ranged from 0 to 200 µg as NH4.
The containers were left to incubate at room temperature (25°C) and after the specified time periods the quantity of NH4-N recovered was determined as previously described.
Study with Soil Digests Containing NH4f-N and Added 15NH4-N
In addition to their use for quantitative determinations, the diffusion procedure also prepares samples for isotope ratio analysis of the 15NH4f-N using a CF-IRMS technique (Brooks et al., 1989; Kelley et al., 1991; Khan et al., 1998). The diffusion units were modified by substituting the boric acid trap with two acidified (10 µL, 2.5 M KHSO4) 7 mm diam. disks of Whatman GF-D filter paper (Whatman International Ltd., Maidstone, UK) as described above. The 15NH3 liberated is trapped on the Whatman GF-D filter paper disks, which are then prepared for direct analysis. This technique is fast and comparable with alternatives such as the Rittenberg technique as described by Mulvaney et al. (1997), and is not prone to sample cross contamination as with other techniques such as oven drying (MacKown et al., 1987).
In this study, two sets of digests were prepared for each of the three soil types across a range of volumes similar to that used previously. One set was spiked with 100 µg of 15NH4 as 15(NH4)2SO4 having 10.6% 15N atom abundance, and the other with a similar quantity of unlabelled NH4. The specified volume of digests was dispensed into the units containing 10 M KOH as described before, then the disks were quickly acidified and the lid fastened tightly onto the container with a final gentle swirling to mix the contents. The units were incubated at room temperature (25°C) for varying periods (dictated by the results of the previous study), at which time, assessments on 15N atom excess and N recovery in the disks were determined for the labeled and unlabeled treatments, respectively. For 15N analysis, the lid was removed from the container in an NH3-free atmosphere, and the wire and disks were placed overnight on a wooden rack in a dessicator over concentrated H2SO4. After drying, both disks from each container were carefully packed into Sn capsules and stored for subsequent analysis via the CF-IRMS system (Brooks et al., 1989; Khan et al., 1998). Concurrently, the N recovered in the discs obtained from similar but unlabeled NH4 treatments were determined by distillation following dissolution (with shaking) of the NH4 from the disks in a measured volume of water for 10 min.
Evaluation of Methodology
Five milliliters of the 20 g L-1 H3BO3-indicator solution was able to trap as much as 1.5 mg of NH4-N in our laboratory, which was well above the maximum quantity used in the recovery studies. Prior to titrimetric determination of the absorbed NH3-N, the solution was diluted to 10 mL with deionized water. This compensated for evaporative loss of water during incubation (especially over long periods) and increased the sensitivity of the titration (Mulvaney et al., 1997). Slight difficulty was experienced in detecting the endpoint of the titration accurately, which was related to the method of titration (manual; 10-mL microburette) with dependence on a color change rather than, as with some automatic titrators, a pH endpoint. This contributed to increasing the coefficient of variation (CV) values. The CVs were also generally higher for recoveries measured earlier in the incubations when diffusions had not yet been completed and small differences among the replicates were observed.
Evaluation of the accuracy and precision in the microdiffusion procedures for quantitative determination of NH4-N in soil digests was accomplished through assessing the recoveries and accompanying CVs from replicate diffusions of NH4-N in H3BO3 and on the Whatman GF-D filter (Whatman International Ltd., Maidstone, UK) paper disks. Mean observed (O) percentage of 15N atom abundance values (n = 4) from analyses by ANCA-MS were compared with an expected (E) value calculated using the equation,
 | [2] |
Quality Control
All diffusion units were cleaned thoroughly of all traces of alkali before reuse, by washing them under tap water and then rinsing them with deionized water. The units used in the procedures described for isotope ratio analysis have the capacity to retain 15NH3 (Saghir et al., 1993), and hence, special techniques as described by Mulvaney et al. (1997) were employed in cleaning the components to prevent cross contamination. The units were immersed in 0.2 M H2SO4 overnight, then rinsed under tap water and finally rinsed with deionized water before being oven dried prior to storage. Additionally, the recommendations by Lory and Russelle (1994) for a quality control approach to minimize diffusion errors associated with discrimination against the heavier 15N isotope were adopted in this study and have worked well.
Statistical Analysis
All experiments were conducted with four replicates, and the data were analyzed using Minitab statistical Software (Minitab, 2000). Analyses of means were conducted for significant differences with a significance level of P < 0.05.
|
RESULTS AND DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONREFERENCES |
|---|
|
|
In situations where laboratories service large sample numbers for NH4f on a regular basis, a number of opportunities exist for shortening and refining the optimum diffusion times required. Khan et al. (1998) indicated that the larger surface area/volume ratio of NH4 digests that can be achieved with the larger mason jars resulted in shorter diffusion periods, and therefore, highlights the potential benefits from using larger jars made of polypropylene for NH4f diffusions. Additionally, the use of these larger jars would accommodate larger volumes of digest, which is of particular relevance to soils low in NH4f. The use of accelerated means of diffusion could also be considered. The fact that shorter diffusion times were observed with higher temperatures and with orbital shaking for mason jar microdiffusions of extractable soil N (Khan et al., 1997; Mulvaney et al., 1997; Khan et al., 1998; Khan et al., 2000) could suggest their use with the polypropylene units as well but with caution. Much greater care must be taken with the use of higher incubation temperatures because of the risks of the polypropylene containers expanding and being deformed at sustained elevated temperatures. Furthermore, to pin down, the minimal period to a more accurate time, plus or minus just a few hours, investigations and research would require multiple sampling times of shorter intervals than that used in these studies, particularly around the period when diffusion is almost complete.
Diffusion of Soil Digests Containing NH4f-N and Added15NH4-N
The results of a series of investigations to compare the quantitative recovery of NH4f and evaluate the 15NH4f preconcentration technique to determine the minimal diffusion periods necessary for obtaining in excess of 96% of the expected 15N atom excess are given in Tables 4 and 5, respectively. Even with low quantitative recoveries of NH4f-N (Table 4), the data indicate the absence of any significant isotopic discrimination and fractionation (discrimination against the heavier 15N isotope during processes such as evaporation) in the diffusion process as evident by >96% recovery of expected 15N atom % (Table 5). This is supported by Khan et al. (1998), who obtained only an error of <2% (between corrected and expected 15N atom %) for incomplete diffusions (1–30 h) of extractable NH4-N. Since the CF-IRMS system requires about 100 µg N for optimization, it is necessary for this amount to be diffused to obtain the most accurate 15N analysis on that system, and accordingly, for the diffusion times to be known for achieving this. The data again show that the optimum diffusion times were influenced by solution volumes rather than by the N concentration or by matrix effects arising from differences among soils. It was observed, however, that this modified system would require incubation for slightly longer periods than with the H3BO3 solution (Table 4). This is probably because of the slower rate of absorption of the NH3 by the Whatman GF-D filter paper (Whatman International Ltd., Maidstone, UK) trap, which has a much smaller exposed surface area for absorption. However, the exact length of time could not be ascertained from the data given except to suggest that it may be a matter of just a few hours more since the differences in recoveries between the two trapping systems range from about 4 to 12%.
|
|
|
CONCLUSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONREFERENCES |
|---|
Further investigations with larger plastic (polyethylene or polypropylene) jars and other means of accelerating the diffusion process would be useful to shorten the incubation times. Additionally, modifications to this technique would need to be explored for their suitability using NH4f digests obtained from the method of Paramasivam and Breitenbeck (2000), which produces a matrix with organic components unlike that produced by the method of Silva and Bremner (1966).
However, this technique, as described herein, can presently be very useful for researchers interested in tracing the peculiar dynamics of fixed NH+4 in soils.
|
ACKNOWLEDGMENTS |
|---|
Received for publication January 10, 2001.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONREFERENCES |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 | |||
