| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Agriculture and Agri-Food Canada, 2560 Hochelaga Blvd., Québec, QC, G1V 2J3, Canada
* Corresponding author (rochettep{at}agr.gc.ca).
 |
ABSTRACT |
|---|
 TOPNOTES ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
Abbreviations: NFT-NSS, non-flow-through, non-steady-state
|
INTRODUCTION |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
There are few reports of soil N2O fluxes before 1980 (Bremner and Blackmer, 1978; Matthias et al., 1980), but their number has increased considerably since 1990 (Stehfest and Bouwman, 2006). Soil N2O flux can be measured using micrometeorological (Wagner-Riddle et al., 1997), soil profile (Rolston, 1986), or steady-state chamber techniques (Denmead, 1979). Nearly all fluxes under field conditions reported in the literature, however, were obtained using non-flow-through, non-steady-state (NFT-NSS) chambers (Bouwman et al., 2002). These chamber measurements were often used for relative flux comparisons between situations (e.g., treatments) within a given study. The absolute values, however, were also used to estimate mean N2O emission rates from agricultural soils (Freibauer, 2003; Bouwman et al., 2002; Gregorich et al., 2005) and to develop the default soil N2O emission factors of the Intergovernmental Panel on Climate Change that are currently used in many countries to calculate greenhouse gas inventories (Eggleston et al., 2006). Moreover, mathematical models that are used to predict N2O emissions at a national scale (Li et al., 1996; Brown et al., 2002; Del Grosso et al., 2005) were calibrated using NFT-NSS chamber data. Therefore, biases in the accuracy of chamber N2O data would also result in similar errors in soil N2O emission inventories.
Chambers are an intrusive gas flux measuring method and their deployment on the soil surface often modifies the flux that it is intended to measure. Consequently, several precautions need to be taken to avoid biased flux estimates when using chambers. The NFT-NSS chamber techniques were developed almost 100 yr ago and were first used for measuring soil respiration rates (Lundegardh, 1926). Very few methodological changes were made to these chambers until Matthias et al. (1980) and Hutchinson and Mosier (1981) proposed several improvements to chamber design and deployment methods. More recently, reviews of chamber methodology (Mosier et al., 1990; Livingston and Hutchinson, 1995; Holland et al., 1999; Smith and Conen, 2004; Rochette and Hutchinson, 2005) and standard protocols (Hutchinson and Livingston, 2002; Rochette and Bertrand, 2007) have summarized the various improvements that were made over time to the NFT-NSS chamber methodology. Several factors, however, including the absence of an absolute reference for the gas source, have impeded efforts for accepting a standard methodology. Accordingly, considerable variation can be observed in the chamber methodology used to measure soil N2O fluxes in recent literature. For example, chamber deployment time ranged from 10 (Colbourn et al., 1984) to 1080 min (Abbasi and Adams, 2000) and the number of air samples taken during deployment varied from one (Baggs et al., 2000; Larsson et al., 1998) to 14 (Jorgensen et al., 1998). Several methodological options such as the type of mathematical model to fit chamber gas concentration with time (Healy et al., 1996; Pedersen et al., 2001) or the use of preinserted bases (Matthias et al., 1980) have been shown to influence NFT-NSS flux estimates. Therefore variations in methodology may result in variable and biased flux estimates.
The objectives of this study were: (i) to determine the criteria for assessing the quality of soil N2O flux measurements made using NFT-NSS chambers; (ii) to apply these criteria to evaluate chamber methodologies used in the scientific literature; and (iii) to propose a minimum set of criteria for NFT-NSS chamber methodology for the measurement of soil N2O flux.
|
MATERIALS AND METHODS |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
|
Insulation
Deployment of a chamber changes the energy balance of the enclosed soil surface, which in turn alters the soil and headspace temperatures. Changes in soil temperature may affect N2O production and flux rates while changes in headspace temperature influence gas concentration determination through pressure effects. Insulation of the chamber is recommended to inhibit energy exchange with the atmosphere and minimize temperature variations during deployment.
Vent
Atmospheric turbulence and barometric pressure fluctuations above the soil surface can affect the soil gas flux rate (Kimball, 1983). A properly designed venting tube transmits changes in external atmospheric pressure to the chamber headspace, thereby minimizing suppression by the chamber of their effect on soil gas fluxes (Hutchinson and Mosier, 1981; Hutchinson and Livingston, 2001). The venting tube also overcomes the effects of chamber volume and pressure changes during chamber deployment and headspace air sampling.
Quality Control of Air Samples
Efficiency in air sample handling may vary between sampling dates. Whenever chambers are used, samples should be taken from a source of known gas concentration (N2O or other reference gas) in the field following the same procedure as for chamber air samples. These samples should be stored and analyzed in the same way as the unknown samples to assess sample handling efficiency.
Nonlinear Model Considered for Determination of the Rate of Nitrous Oxide Change
Gas diffusion theory predicts that an increase in headspace N2O concentration during chamber deployment has an immediate negative impact on the rate of N2O flux at the soil surface. Therefore, nonlinear models fitted to the change in headspace gas concentration with time often yield less-biased estimates of the rate of change in headspace N2O concentration (dC/dt) than linear models (Healy et al., 1996). Linear fit of headspace N2O concentration with time should be used to determine dC/dt only when quadratic effects are not significant.
Testing for Zero Flux
Soil N2O fluxes are often small and fluxes should be reported only if they are greater than the minimum detectable flux. This minimum flux is determined by statistically testing if the slope of dC/dt is significantly different from zero (Livingston and Hutchinson, 1995).
Headspace Temperature Correction
Air temperature during chamber deployment in the field is rarely the same as laboratory temperature at the time of air sample analysis. Temperature differences between air sampling and analysis bias determination of the gas concentration. Corrections based on the perfect gas law are required (Rochette and Hutchinson, 2005).
Sample Taken at Deployment Time Zero
The key to minimizing biases induced by the presence of the chamber on soil N2O fluxes is to project a dC/dt representative of the predeployment conditions. This requires that chamber headspace N2O concentration at deployment time zero be determined as accurately as possible. It is therefore recommended that the first air sample be taken inside the chamber immediately after being placed on the collar rather than assuming ambient air concentration.
Type of Air Sample Container
Air samples can be stored in various types of containers. Plastic syringes are leaky and cannot preserve the integrity of air samples even during short periods of time (Rochette and Bertrand, 2003). Glass syringes offer a better seal but are considered inferior to rigid glass or aluminum containers. Among fixed-volume containers, Exetainers (Labco, High Wycombe, UK) are a recommended option because their efficiency has been extensively documented (Laughlin and Stevens, 2003; Rochette and Bertrand, 2003).
Positive Pressure during Storage and Handling of Air Samples
Pressurizing air samples in fixed-volume containers allows the detection of leaky containers but most importantly avoids contamination when air subsamples are taken from the container for analysis.
Numerical Characteristics
Height of Chamber
Chamber height affects the quality of chamber measurement in several ways. High chambers have large minimum detectable flux and their performance can be reduced by poor headspace air mixing. Inversely, the deployment of short chambers has greater impacts on environmental conditions because the inertia of chamber headspace conditions (temperature, humidity, and gas concentration) decreases with decreasing chamber height (Hutchinson and Livingston, 2002). The optimum chamber height is therefore a compromise between achieving a reasonable minimum detectable flux while minimizing changes in the headspace environment. As a result, it is intimately linked to deployment duration and is expressed as the ratio of chamber height to deployment duration. We estimated that a flux that could not be measured using 20-cm-high chambers during a 30-min deployment time was insignificant in most situations. Accordingly, chamber height was estimated as very good for values 
40 cm h–1.
Chamber Base Insertion
A good seal between the chamber and soil requires insertion of the chamber walls or base into the soil. Leakage or contamination can occur by lateral diffusion of N2O beneath the base in response to deformation of the vertical N2O concentration gradient in the soil. The required depth of chamber base insertion is related to the duration of chamber deployment and this characteristic was estimated as the insertion depth/deployment duration ratio. Model simulations suggest that insertion depths required to limit biases in gas flux estimates to <1% are 2.5 cm for 10-min deployments and 13 cm for 60-min deployments in a soil having air-filled porosity of 0.3 m3 m–3 (Hutchinson and Livingston, 2001). Accordingly, chamber base insertion was estimated as very good for insertion depth/deployment duration ratios 12 cm h–1. We assumed that the chamber base insertion in paddy rice and natural wet ecosystems was very good regardless of insertion depth because the saturated soil conditions in these ecosystems result in very low soil gas diffusivity and lateral leaks below the chamber base (Hutchinson and Livingston, 2001).
Chamber Area/Perimeter Ratio
The relative error in flux estimates associated with a poor chamber seal increases with decreasing chamber diameter. This is because the gas flux is proportional to the source area while the gas leak is proportional to the perimeter of the chamber (Healy et al., 1996). Therefore, when the diameter of a cylindrical chamber is increased, the soil gas flux into the chamber is increased at a greater rate than the leakage below the base or at the joint between the chamber and the base. Based on model predictions (Healy et al., 1996), we estimated that this characteristic was very good for values 10 cm (i.e, a cylindrical chamber with a diameter of 40 cm).
Duration of Deployment
Most problems related to the presence of chambers on the soil surface, such as changes in soil and air temperature and humidity, gas leakage, and negative feedback on the N2O flux, increase with deployment time. Accordingly, air samples should be taken in as short a time as possible to observe a measurable increase in N2O headspace concentration. We estimated that deployment durations 40 min are likely to result in significant negative impacts on chamber conditions without associated gain in chamber performance in most situations.
Number of Air Samples Taken
The quality of soil N2O flux estimates using NFT-NSS chambers are intimately linked to that of dC/dt. The accuracy of the determination of dC/dt increases with the number of air samples taken during deployment. Also, the use of fitting models (Pedersen et al., 2001) and the determination of the uncertainty in dC/dt require that more than two air samples be taken. It is therefore recommended that a minimum of three, but preferably four or more discrete air samples be taken during the deployment period.
Duration of Air Sample Storage
All gas containers leak and the contamination of air samples increases with storage duration and varies with the type of gas container. Significant contamination occurs within a few hours in plastic syringes but contamination can be <1.5% per week in pressurized Exetainers (Rochette and Bertrand, 2003). There is little information on the leakage from other containers such as glass syringes and various rigid vials but their performance is probably intermediate between plastic syringes and Exetainers.
Factors that Influence Measurement
The quality of flux measurements made using a NFT-NSS chamber is related to the quality of the characteristics described above. The importance and manner in which each characteristic affects the quality of the measurement are not equal, however, and it is the interaction between characteristics that often determines their impact on the measurement. The quality of the N2O flux measurement was therefore evaluated based on factors that we identified as having a direct influence on the measurement. Factors are indices integrating the impact of several characteristics on four given aspects of NFT-NSS methodology: the design of the chamber, the seal on the soil surface, the air sample handling and storage, and the determination of dC/dt.
Chamber Design
This factor describes how the chamber is designed to minimize soil disturbance, optimize headspace volume, and avoid pressure or temperature variations during deployment. The importance of reducing soil disturbance during measurement by using a semipermanent base installed in the soil ahead of time is given more importance than the other components. A high score for this factor indicates minimum disturbance of the soil and headspace environment by the chamber.
Seal on Soil Surface
This factor describes how effective the design and deployment of the chamber is at reducing the risk of a gas leak under the bottom edges of the chamber. The risk of leakage under the lower edge of the chamber or base will increase with shallow insertions in the soil, small area/perimeter ratios, shorter chambers, and longer deployment durations. The chamber insertion characteristic was given more importance in the equation and a high score for this factor indicates a tall and wide chamber inserted deeply into the soil for a short deployment time. Chambers used in ecosystems with saturated soil conditions such as paddy rice and natural wet ecosystems were given very good scores based on the very low risk of gas leakage in these systems.
Air Sample Handling and Storage
This factor describes the integrity and reliability of the samples taken. It is an evaluation of the risk of sample contamination based on how the air samples are handled and stored. The type of vial is given greater importance than the handling characteristics. A high score for this factor indicates that the air sample was stored in a reliable container, that the duration of sample storage was sufficiently short to prevent significant leaks, and that a quality control sample was tested to ensure replicable methodology.
Determination of the Rate of Change of Nitrous Oxide Concentration
This factor describes the accuracy of the determination of the rate of change in headspace concentration at deployment time zero. The rate of change in headspace N2O concentration can be better estimated with a greater number of data points and with the appropriate mathematical model. Therefore the number of air samples taken and consideration of a nonlinear equation were given greater importance. A high score for this factor indicates that more than three samples were taken, including one sample taken from the headspace at the beginning of the deployment, that the duration of deployment was short, and that the nonlinear model was tested.
Evaluation of Chamber Methodology
Data Set
The characteristics and factors described above were used to evaluate the NFT-NSS chamber methodology reported in 356 articles published in peer-reviewed journals. This data set was compiled from previous N2O flux literature reviews (Stehfest and Bouwman, 2006; Jungkunst et al., 2006; Lu et al., 2006) to which we added studies from a survey of recent literature. Only studies using NFT-NSS chamber methods to measure soil N2O flux were included. The few studies using micrometeorological, laboratory, mesocosm, and other chamber techniques were excluded from the data set.
To minimize bias in our selection, we collected studies from the earliest record (1978) to February 2007, all continents excluding Antarctica, and a range of ecosystems. The data set was divided into five time intervals, six continents, and seven ecosystems. Five-year time intervals were used with the exception of the earliest (1978–1989) and latest (2005–2007) intervals. The ecosystems were selected based on those sharing similar climate and soil properties. They were: (i) annual field crops, (ii) paddy rice, (iii) perennial crops, (iv) natural grasslands and savannahs, (v) temperate and boreal forests, (vi) tropical forests, and (vii) natural wet ecosystems such as mangroves and marshes. Arid and polar ecosystems were excluded because there were no studies from these regions. We attempted to collect a minimum of 20 studies from each time interval, continent, and ecosystem; however, for certain situations this was not possible. Studies from Africa (n = 9), grasslands and savannahs (n = 11), and natural wet ecosystems (n = 4) were rare but were nonetheless included in the data set. If a study was conducted in more than one ecosystem, it was recorded more than once; therefore, from the 356 studies reviewed in the data set, there were 416 records in different ecosystems. The same applied to studies in more than one country, although all studies occurring in more than one country were also on the same continent, therefore there are also 356 records across all continents.
Scoring of Characteristics and Factors
From each study in the data set, we extracted information about each characteristic and evaluated each record for their contribution to the quality of the N2O flux measurement. Each record was given a score from 0 to 3 and the quality described as very poor (0), poor (1), good (2), or very good (3). The quality of each value was determined based on previous assessments of NFT-NSS techniques (Hutchinson and Livingston, 1993, 2002; Livingston and Hutchinson, 1995; Holland et al., 1999; Davidson et al., 2002; Smith and Conen 2004; Rochette and Hutchinson, 2005; Rochette and Bertrand, 2007). If no information about a characteristic was given for the study, then we assumed that it was absent in the case of binary characteristics or that no information was provided in the case of numerical characteristics. In situations where a characteristic did not apply, "not applicable" was recorded and omitted from further data manipulation. A full list of the characteristics and scores for each value is given in Table 1.
We evaluated the quality of each factor as the weighted sum of the score of all component characteristics, as shown in Eq. [1–4]
. The importance of each characteristic to the impact of each factor was determined by the weighting of each characteristic in the equation. "Primary" characteristics of greater importance were given higher weighting than "secondary" characteristics of lower importance:
 | [1] |
 | [2] |
Air sample handling and storage:
 | [3] |
 | [4] |
The lack of information for certain characteristics created some problems during the evaluation of each factor. We considered the chamber base insertion (d) and the number of air samples taken (n) to be essential for making an evaluation of the factors so that if no information was given for these primary characteristics then no score could be calculated for S or F (Eq. [2] and [4]). If no information was given for the secondary characteristics (t, h, nl, and t0), then a default value of 1 (poor) was given to numerical characteristics (t, h, and t0) and an "absence" or zero score (very poor) given for testing the nonlinear model (nl). In Eq. [1] and [3], we assumed that there were no essential characteristics without which we could not make an evaluation of the factors. Therefore, if no information was given for any of the characteristics in Eq. [1] and [3], then the characteristic was omitted from the factor calculation and the denominator adjusted to the number of terms left in the equation. The quality for each factor was described using the following scoring categories: very poor (0–0.74), poor (0.75–1.49), good (1.50–2.24), and very good (2.25–3.0). Additionally, undetermined factors were scored as poor.
Confidence in Chamber Emission Measurements
The overall confidence in the flux measurement was based on the weakest factors in each study, and therefore decreased when one or more factors scored poor or very poor. If the quality of a measurement was compromised by one factor scoring poorly, the loss in confidence could not be offset by good or very good scores for the other factors. For example, if the air sample leaked from a plastic syringe, an example of very poor sample handling, then the confidence in that measurement was low regardless of whether the other factors were very good.
The confidence was estimated from very low to high based on the following criteria:
|
RESULTS |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
17% of the studies, and insulation, venting, and temperature corrections were reported in 33 to 49% of the studies (Table 2). The type of chamber, pressurized air samples, and the sample taken at time zero were the binary characteristics most often reported (>62% of the studies). Reporting of most binary characteristics increased with time except for the use of quality control air samples and of an air sample at deployment time zero. Surprisingly, the frequency of the report of insulation and venting after 2005 was lower than the average across all time periods.
|
|
|
|
|
|
|
|
|
The quality of air sample handling and storage factor (H) improved with time and the number of studies with good or very good sample handling increased from 32% of studies from before 1990 to 55% after 2005 (Fig. 3c; Table 4). This largely reflected the reduced use of plastic syringes in favor of glass vials, and increased use of pressurized samples in fixed-volume containers. Nonetheless, it is of concern that 45% of studies after 2005 had very poor or poor air sample handling (Fig. 3c). Similarly, very poor or poor sample handling was very common in studies from South America (81%), Africa (78%), grasslands and savannahs (91%), and all forest ecosystems (71–74%) (Fig. 4c and 5c). Poor scores for air sample handling in these regions were probably because several of these studies were led by scientists from other continents and had associated long delays before sample analysis. Studies from Oceania had the best sample handling, with 61% of studies rating good or very good.
Unlike the seal factor, the number of studies not providing enough information to calculate the dC/dt factor (F) decreased from 18% of studies before 1990 to 5% after 2005. This factor was not calculated, however, for 21 to 22% of studies from Asia and paddy rice ecosystems (Fig. 3d, 4d, and 5d). The mean score of all studies was 1.5, and 52% were estimated good or very good, yet there were still up to 38% of studies with very poor or poor methods of determining dC/dt after 2005. Studies from North and South America scored particularly well (Table 5), and only 19% of studies rated very poor or poor compared with the mean of other continents (53%) (Fig. 4d). Studies from forest and natural wet ecosystems had higher proportion of studies with good or very good methods (63–83%) compared with the mean of other ecosystems (46%) (Fig. 5d, Table 6 ).
|
0 (z) are characteristics that were not explicitly included in the calculation of factors even though they may contribute to the quality of the flux measurement. Although only 33% of studies reported temperature corrections, its impact may be minimal. Flux rates are estimated to differ by a maximum of only about 7% if there is a 20°C difference between field and laboratory temperatures. Similarly, only 5% of studies reported testing to determine if the reported flux is significantly greater or less than zero. This may not have a major effect on mean fluxes reported during a given study because this action would only affect observations with very low flux rates. Many small negative fluxes reported in the literature may not be significantly different from zero, however, and therefore not make robust estimates of soil N2O sinks (Chapuis-Lardy et al., 2006).
Confidence in Flux Data
As the number of very poor or poor characteristics in a study increases, the confidence in flux measurement decreases. It is of concern that only 40% of all studies had medium or high confidence (Table 7
). The confidence in flux measurements increased with time, however, as the proportion of studies with high or medium confidence increased from a low of 29% of studies in 1990 to 50% after 2005 (Table 7). Similarly, the proportion of studies with very low confidence decreased from a high of 46 to 9% of studies in the same period. There was lower confidence in measurements from Africa, with 89% of studies with a very low or low confidence compared with 61% of studies from other continents. About 70% of studies from tropical forest, perennial, and grasslands and savannah ecosystems had very low or low confidence compared with about 25 to 60% of studies from other ecosystems.
|
|
DISCUSSION |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
The design of chambers was generally good across all studies, although increased reporting of insulation and venting is necessary. Important improvements are needed for several other aspects of the NFT-NSS chamber methodology. Better seal on the soil surface should be achieved by increasing the insertion depth of the base, reducing deployment duration, and avoiding chamber heights of <10 cm. Quality control gas samples are necessary to evaluate the risk of leaks or contamination in the handling process. The integrity and reliability of the air sample handling and storage process could also be significantly improved for the majority of studies by avoiding plastic syringes, decreasing the duration of sample storage, and using quality control samples. Plastic syringes should be avoided as storage vessels because N2O losses and adsorption on the inner walls are large (Rochette and Bertrand, 2003). Minimizing the delay between the collection and analysis of the air sample is important as losses of N2O can be as large as 16% after 24 h from plastic syringes and 0.2% per day from glass vials (Rochette and Bertrand, 2003). A more accurate description of the change in chamber headspace N2O concentration would improve estimates of dC/dt at deployment time zero. Although the increase in chamber headspace N2O concentration often appears linear, the nonlinear model is most often a better fit and the flux can be >20% greater when using a nonlinear than a linear model (Matthias et al., 1978; Healy et al., 1996; Rochette and Bertrand, 2007).
Use of poor chamber methodologies increases uncertainty in estimated soil surface N2O emissions (i.e., reduces precision). It would also be useful to determine if there is a systematic positive or negative bias in the loss of flux measurement precision. An approximate estimate of the error was made based on previous assessments found in the scientific literature. Characteristics contributing to a poor seal such as a shallow insertion, low chamber height, and long sampling duration increase the risk of lateral diffusion of N2O beneath the lower edge of the chamber, which can lead to underestimating the flux by >50% (Healy et al., 1996; Hutchinson and Livingston, 2001). Similarly, poor air sample storage, due to a combination of poor-quality vials, long storage duration, or contamination of unpressurized samples, increases leaks from samples and can cause the flux to be underestimated by up to 25% (Rochette and Bertrand, 2003). Furthermore, use of a linear model to fit nonlinear variations in chamber headspace N2O concentration often yields estimates of dC/dt at time zero that are as much as 20% lower than nonlinear models (Healy et al., 1996). In rare situations, a poor methodology can be compensated by a good method. For example, the increase in N2O leaks from the chamber due to a poor seal causes a nonlinear increase in chamber N2O concentration. This poor methodology can therefore be partially compensated by using a nonlinear model to fit the flux curve. This situation is relatively uncommon, however, as only 31% of studies with a very poor or poor score on the seal factor tested the nonlinear model. Finally, the use of push-in chambers is the only characteristic that may result in an overestimation of flux estimates by decreasing soil resistance to gas diffusion; however, this effect is marginal since only 10% of studies reviewed here used push-in chambers.
Of further concern with the low confidence of flux measurements are that the errors are additive and that most of them underestimate real N2O fluxes. Therefore, we estimated that the range of cumulative underestimation of N2O fluxes would be approximately from 0 to 10% for studies with high confidence, from 0 to 30% for studies with medium confidence, from 10 to 50% for studies with low confidence, and from 20 to 60% for studies with very low confidence.
|
CONCLUSIONS |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
Considering the contributions that have been made since 1990 concerning the optimal design and methodology of NFT-NSS chambers (Hutchinson and Livingston, 1993, 2002; Livingston and Hutchinson, 1995; Davidson et al., 2002; Smith and Conen, 2004; Rochette and Hutchinson 2005), it is regrettable that the quality of flux measurements remains poor or very poor for about 50% of recently published studies (2005–2007). Some improvements to the design and methodology of NFT-NSS chambers have been made, notably in handling and storage of air samples; however, the adoption of better methods and more complete reporting are required for many characteristics such as testing of the nonlinear model, quality control of air samples, and elimination of plastic syringes as storage vessels.
This review indicates that improving the reliability of soil N2O flux data will require a greater effort by scientists to apply and report more rigorous methodological standards and greater vigilance by reviewers and scientific editors. We acknowledge that N2O flux measurements in some situations may require exceptional chamber design or methodology. In the interest of improving the quality and confidence of future N2O flux data, however, we recommend adoption of a minimum set of criteria for reliable soil N2O flux measurement using NFT-NSS chamber methods. More specifically, we propose that NFT-NSS chamber methodologies should for most situations:
|
NOTES |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
Received for publication June 12, 2007.
|
REFERENCES |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
. Kasimir-Klemedtsson, and L. Klemedtsson. 1998. Ammonia and nitrous oxide emissions from grass and alfalfa mulches. Nutr. Cycling Agroecosyst. 51:41–46.[CrossRef]This article has been cited by other articles:
|
|
 |
|
  F. Alluvione, A. D. Halvorson, and S. J. Del Grosso Nitrogen, Tillage, and Crop Rotation Effects on Carbon Dioxide and Methane Fluxes from Irrigated Cropping Systems J. Environ. Qual., August 24, 2009; 38(5): 2023 - 2033. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. T. Venterea, K. A. Spokas, and J. M. Baker Accuracy and Precision Analysis of Chamber-Based Nitrous Oxide Gas Flux Estimates Soil Sci. Soc. Am. J., May 13, 2009; 73(4): 1087 - 1093. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. K. Shrestha, R. Lal, and C. Penrose Greenhouse Gas Emissions and Global Warming Potential of Reclaimed Forest and Grassland Soils J. Environ. Qual., February 6, 2009; 38(2): 426 - 436. [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 | |||
